Hospital Medicine

Anticoagulation is commonly used in the management of atrial fibrillation to reduce the risk of ischemic stroke. Warfarin and other anticoagulants increase the risk of hemorrhagic complications, including upper gastrointestinal bleeding (UGIB). Following UGIB, management of anticoagulation is highly variable. Many patients permanently discontinue anticoagulation, while others continue without interruption.^1-4 Among patients who resume warfarin, different cohorts have measured median times to resumption ranging from four days to 50 days.^1-3 Outcomes data are sparse, and clinical guidelines offer little direction.⁵

Following UGIB, the balance between the risks and benefits of anticoagulation changes over time. Rebleeding risk is highest immediately after the event and declines quickly; therefore, rapid resumption of anticoagulation causes patient harm.³ Meanwhile, the risk of stroke remains constant, and delay in resumption of anticoagulation is associated with increased risk of stroke and death.¹ At some point in time following the initial UGIB, the expected harm from bleeding would equal the expected harm from stroke. This time point would represent the optimal time to restart anticoagulation.

Trial data are unlikely to identify the optimal time for restarting anticoagulation. A randomized trial comparing discrete reinitiation times (eg, two weeks vs six weeks) may easily miss the optimal timing. Moreover, because the daily probability of thromboembolic events is low, large numbers of patients would be required to power such a study. In addition, a number of oral anticoagulants are now approved for prevention of thromboembolic stroke in atrial fibrillation, and each drug may have different optimal timing.

In contrast to randomized trials that would be impracticable for addressing this clinical issue, microsimulation modeling can provide granular information regarding the optimal time to restart anticoagulation. Herein, we set out to estimate the expected benefit of reinitiation of warfarin, the most commonly used oral anticoagulant,⁶ or apixaban, the direct oral anticoagulant with the most favorable risk profile,⁷ as a function of days after UGIB.

We previously described a microsimulation model of anticoagulation among patients with nonvalvular atrial fibrillation (NVAF; hereafter, we refer to this model as the Personalized Anticoagulation Decision-Making Assistance model, or PADMA).^8,9 For this study, we extended this model to incorporate the probability of rebleeding following UGIB and include apixaban as an alternative to warfarin. This model begins with a synthetic population following UGIB, the members of which are at varying risk for thromboembolism, recurrent UGIB, and other hemorrhages. For each patient, the model simulates a number of possible events (eg, thromboembolic stroke, intracranial hemorrhage, rebleeding, and other major extracranial hemorrhages) on each day of an acute period of 90 days after hemostasis. Patients who survive until the end of the acute period enter a simulation with annual, rather than daily, cycles. Our model then estimates total quality-adjusted life-years (QALYs) for each patient, discounted to the present. We report the average discounted QALYs produced by the model for the same population if all individuals in our input population were to resume either warfarin or apixaban on a specific day. Input parameters and ranges are summarized in Table 1, a simplified schematic of our model is shown in the Supplemental Appendix, and additional details regarding model structure and assumptions can be found in earlier work.^8,9 We simulated from a health system perspective over a lifelong time horizon. All analyses were performed in version 14 of Stata (StataCorp, LLC, College Station, Texas).

Synthetic Population

To generate a population reflective of the comorbidities and age distribution of the US population with NVAF, we merged relevant variables from the National Health and Nutrition Examination Survey (NHANES; 2011-2012), using multiple imputation to correct for missing variables.¹⁰ We then bootstrapped to national population estimates by age and sex to arrive at a hypothetical population of the United States.¹¹ Because NHANES does not include atrial fibrillation, we applied sex- and age-specific prevalence rates from the AnTicoagulation and Risk Factors In Atrial Fibrillation study.¹² We then calculated commonly used risk scores (CHA₂DS₂-Vasc and HAS-BLED) for each patient and limited the population to patients with a CHA₂DS₂-Vasc score of one or greater.^13,14 The population resuming apixaban was further limited to patients whose creatinine clearance was 25 mL/min or greater in keeping with the entry criteria in the phase 3 clinical trial on which the medication’s approval was based.¹⁵

To estimate patient-specific probability of rebleeding, we generated a Rockall score for each patient.¹⁶ Although the discrimination of the Rockall score is limited for individual patients, as with all other tools used to predict rebleeding following UGIB, the Rockall score has demonstrated reasonable calibration across a threefold risk gradient.^17-19 International consensus guidelines recommend the Rockall score as one of two risk prediction tools for clinical use in the management of patients with UGIB.²⁰ In addition, because the Rockall score includes some demographic components (five of a possible 11 points), our estimates of rebleeding risk are covariant with other patient-specific risks. We assumed that the endoscopic components of the Rockall score were present in our cohort at the same frequency as in the original derivation and are independent of known patient risk factors.¹⁶ For example, 441 out of 4,025 patients in the original Rockall derivation cohort presented with a systolic blood pressure less than 100 mm Hg. We assumed that an independent and random 10.96% of the cohort would present with shock, which confers two points in the Rockall score.

The population was replicated 60 times, with identical copies of the population resuming anticoagulation on each of days 1-60 (where day zero represents hemostasis). Intermediate data regarding our simulated population can be found in the Supplemental Appendix and in prior work.

Event Type, Severity, and Mortality

Each patient in our simulation could sustain several discrete and independent events: ischemic stroke, intracranial hemorrhage, recurrent UGIB, or extracranial major hemorrhage other than recurrent UGIB. As in prior analyses using the PADMA model, we did not consider minor hemorrhagic events.⁸

The probability of each event was conditional on the corresponding risk scoring system. Patient-specific probability of ischemic stroke was conditional on CHA₂DS₂-Vasc score.^21,22 Patient-specific probability of intracranial hemorrhage was conditional on HAS-BLED score, with the proportions of intracranial hemorrhage of each considered subtype (intracerebral, subarachnoid, or subdural) bootstrapped from previously-published data.^21-24 Patient-specific probability of rebleeding was conditional on Rockall score from the combined Rockall and Vreeburg validation cohorts.¹⁷ Patient-specific probability of extracranial major hemorrhage was conditional on HAS-BLED score.²¹ To avoid double-counting of UGIB, we subtracted the baseline risk of UGIB from the overall rate of extracranial major hemorrhages using previously-published data regarding relative frequency and a bootstrapping approach.²⁵

Probability of Rebleeding Over Time

To estimate the decrease in rebleeding risk over time, we searched the Medline database for systematic reviews of recurrent bleeding following UGIB using the strategy detailed in the Supplemental Appendix. Using the interval rates of rebleeding we identified, we calculated implied daily rates of rebleeding at the midpoint of each interval. For example, 39.5% of rebleeding events occurred within three days of hemostasis, implying a daily rate of approximately 13.2% on day two (32 of 81 events over a three-day period). We repeated this process to estimate daily rates at the midpoint of each reported time interval and fitted an exponential decay function.²⁶ Our exponential fitted these datapoints quite well, but we lacked sufficient data to test other survival functions (eg, Gompertz, lognormal, etc.). Our fitted exponential can be expressed as:

P _rebleeding = b ₀ *exp(b ₁ *day)

where b₀ = 0.1843 (SE: 0.0136) and b₁ = –0.1563 (SE: 0.0188). For example, a mean of 3.9% of rebleeding episodes will occur on day 10 (0.1843 *exp(–0.1563 *10)).

Relative Risks of Events with Anticoagulation

For patients resuming warfarin, the probabilities of each event were adjusted based on patient-specific daily INR. All INRs were assumed to be 1.0 until the day of warfarin reinitiation, after which interpolated trajectories of postinitiation INR measurements were sampled for each patient from an earlier study of clinical warfarin initiation.²⁷ Relative risks of ischemic stroke and hemorrhagic events were calculated based on each day’s INR.

For patients taking apixaban, we assumed that the medication would reach full therapeutic effect one day after reinitiation. Based on available evidence, we applied the relative risks of each event with apixaban compared with warfarin.²⁵

Future Disability and Mortality

Each event in our simulation resulted in hospitalization. Length of stay was sampled for each diagnosis.²⁸ The disutility of hospitalization was estimated based on length of stay.⁸ Inpatient mortality and future disability were predicted for each event as previously described.⁸ We assumed that recurrent episodes of UGIB conferred morbidity and mortality identical to extracranial major hemorrhages more broadly.^29,30

Disutilities

We used a multiplicative model for disutility with baseline utilities conditional on age and sex.³¹ Each day after resumption of anticoagulation carried a disutility of 0.012 for warfarin or 0.002 for apixaban, which we assumed to be equivalent to aspirin in disutility.³² Long-term disutility and life expectancy were conditional on modified Rankin Score (mRS).^33,34 We discounted all QALYs to day zero using standard exponential discounting and a discount rate centered at 3%. We then computed the average discounted QALYs among the cohort of patients that resumed anticoagulation on each day following the index UGIB.

Sensitivity Analyses and Metamodel

To assess sensitivity to continuously varying input parameters, such as discount rate, the proportion of extracranial major hemorrhages that are upper GI bleeds, and inpatient mortality from extracranial major hemorrhage, we constructed a metamodel (a regression model of our microsimulation results).³⁵ We tested for interactions among input parameters and dropped parameters that were not statistically significant predictors of discounted QALYs from our metamodel. We then tested for interactions between each parameter and day resuming anticoagulation to determine which factors may impact the optimal day of reinitiation. Finally, we used predicted marginal effects from our metamodel to assess the change in optimal day across the ranges of each input parameter when other parameters were held at their medians.

Resuming warfarin on day zero produced the fewest QALYs. With delay in reinitiation of anticoagulation, expected QALYs increased, peaked, and then declined for all scenarios. In our base-case simulation of warfarin, peak utility was achieved by resumption 41 days after the index UGIB. Resumption between days 32 and 51 produced greater than 99.9% of peak utility. In our base-case simulation of apixaban, peak utility was achieved by resumption 32 days after the index UGIB. Resumption between days 21 and 47 produced greater than 99.9% of peak utility. Results for warfarin and apixaban are shown in Figures 1 and 2, respectively.

The optimal day of warfarin reinitiation was most sensitive to CHA₂DS₂-Vasc scores and varied by around 11 days between a CHA₂DS₂-Vasc score of one and a CHA₂DS₂-Vasc score of six (the 5^th and 95^th percentiles, respectively) when all other parameters are held at their medians. Results were comparatively insensitive to rebleeding risk. Varying Rockall score from two to seven (the 5^th and 95^th percentiles, respectively) added three days to optimal warfarin resumption. Varying other parameters from the 5^th to the 95^th percentile (including HAS-BLED score, sex, age, and discount rate) changed expected QALYs but did not change the optimal day of reinitiation of warfarin. Optimal day of reinitiation for warfarin stratified by CHA₂DS₂-Vasc score is shown in Table 2.

Anticoagulation is frequently prescribed for patients with NVAF, and hemorrhagic complications are common. Although anticoagulants are withheld following hemorrhages, scant evidence to inform the optimal timing of reinitiation is available. In this microsimulation analysis, we found that the optimal time to reinitiate anticoagulation following UGIB is around 41 days for warfarin and around 32 days for apixaban. We have further demonstrated that the optimal timing of reinitiation can vary by nearly two weeks, depending on a patient’s underlying risk of stroke, and that early reinitiation is more sensitive to rebleeding risk than late reinitiation.

Prior work has shown that early reinitiation of anticoagulation leads to higher rates of recurrent hemorrhage while failure to reinitiate anticoagulation is associated with higher rates of stroke and mortality.^1-4,36 Our results add to the literature in a number of important ways. First, our model not only confirms that anticoagulation should be restarted but also suggests when this action should be taken. The competing risks of bleeding and stroke have left clinicians with little guidance; we have quantified the clinical reasoning required for the decision to resume anticoagulation. Second, by including the disutility of hospitalization and long-term disability, our model more accurately represents the complex tradeoffs between recurrent hemorrhage and (potentially disabling) stroke than would a comparison of event rates. Third, our model is conditional upon patient risk factors, allowing clinicians to personalize the timing of anticoagulation resumption. Theory would suggest that patients at higher risk of ischemic stroke benefit from earlier resumption of anticoagulation, while patients at higher risk of hemorrhage benefit from delayed reinitiation. We have quantified the extent to which patient-specific risks should change timing. Fourth, we offer a means of improving expected health outcomes that requires little more than appropriate scheduling. Current practice regarding resuming anticoagulation is widely variable. Many patients never resume warfarin, and those that do resume do so after highly varied periods of time.^1-5,36 We offer a means of standardizing clinical practice and improving expected patient outcomes.

Disclosures

The authors have nothing to disclose.

Funding

The authors received no specific funding for this work.

Anticoagulation is commonly used in the management of atrial fibrillation to reduce the risk of ischemic stroke. Warfarin and other anticoagulants increase the risk of hemorrhagic complications, including upper gastrointestinal bleeding (UGIB). Following UGIB, management of anticoagulation is highly variable. Many patients permanently discontinue anticoagulation, while others continue without interruption.^1-4 Among patients who resume warfarin, different cohorts have measured median times to resumption ranging from four days to 50 days.^1-3 Outcomes data are sparse, and clinical guidelines offer little direction.⁵

Following UGIB, the balance between the risks and benefits of anticoagulation changes over time. Rebleeding risk is highest immediately after the event and declines quickly; therefore, rapid resumption of anticoagulation causes patient harm.³ Meanwhile, the risk of stroke remains constant, and delay in resumption of anticoagulation is associated with increased risk of stroke and death.¹ At some point in time following the initial UGIB, the expected harm from bleeding would equal the expected harm from stroke. This time point would represent the optimal time to restart anticoagulation.

Trial data are unlikely to identify the optimal time for restarting anticoagulation. A randomized trial comparing discrete reinitiation times (eg, two weeks vs six weeks) may easily miss the optimal timing. Moreover, because the daily probability of thromboembolic events is low, large numbers of patients would be required to power such a study. In addition, a number of oral anticoagulants are now approved for prevention of thromboembolic stroke in atrial fibrillation, and each drug may have different optimal timing.

In contrast to randomized trials that would be impracticable for addressing this clinical issue, microsimulation modeling can provide granular information regarding the optimal time to restart anticoagulation. Herein, we set out to estimate the expected benefit of reinitiation of warfarin, the most commonly used oral anticoagulant,⁶ or apixaban, the direct oral anticoagulant with the most favorable risk profile,⁷ as a function of days after UGIB.

We previously described a microsimulation model of anticoagulation among patients with nonvalvular atrial fibrillation (NVAF; hereafter, we refer to this model as the Personalized Anticoagulation Decision-Making Assistance model, or PADMA).^8,9 For this study, we extended this model to incorporate the probability of rebleeding following UGIB and include apixaban as an alternative to warfarin. This model begins with a synthetic population following UGIB, the members of which are at varying risk for thromboembolism, recurrent UGIB, and other hemorrhages. For each patient, the model simulates a number of possible events (eg, thromboembolic stroke, intracranial hemorrhage, rebleeding, and other major extracranial hemorrhages) on each day of an acute period of 90 days after hemostasis. Patients who survive until the end of the acute period enter a simulation with annual, rather than daily, cycles. Our model then estimates total quality-adjusted life-years (QALYs) for each patient, discounted to the present. We report the average discounted QALYs produced by the model for the same population if all individuals in our input population were to resume either warfarin or apixaban on a specific day. Input parameters and ranges are summarized in Table 1, a simplified schematic of our model is shown in the Supplemental Appendix, and additional details regarding model structure and assumptions can be found in earlier work.^8,9 We simulated from a health system perspective over a lifelong time horizon. All analyses were performed in version 14 of Stata (StataCorp, LLC, College Station, Texas).

Synthetic Population

To generate a population reflective of the comorbidities and age distribution of the US population with NVAF, we merged relevant variables from the National Health and Nutrition Examination Survey (NHANES; 2011-2012), using multiple imputation to correct for missing variables.¹⁰ We then bootstrapped to national population estimates by age and sex to arrive at a hypothetical population of the United States.¹¹ Because NHANES does not include atrial fibrillation, we applied sex- and age-specific prevalence rates from the AnTicoagulation and Risk Factors In Atrial Fibrillation study.¹² We then calculated commonly used risk scores (CHA₂DS₂-Vasc and HAS-BLED) for each patient and limited the population to patients with a CHA₂DS₂-Vasc score of one or greater.^13,14 The population resuming apixaban was further limited to patients whose creatinine clearance was 25 mL/min or greater in keeping with the entry criteria in the phase 3 clinical trial on which the medication’s approval was based.¹⁵

To estimate patient-specific probability of rebleeding, we generated a Rockall score for each patient.¹⁶ Although the discrimination of the Rockall score is limited for individual patients, as with all other tools used to predict rebleeding following UGIB, the Rockall score has demonstrated reasonable calibration across a threefold risk gradient.^17-19 International consensus guidelines recommend the Rockall score as one of two risk prediction tools for clinical use in the management of patients with UGIB.²⁰ In addition, because the Rockall score includes some demographic components (five of a possible 11 points), our estimates of rebleeding risk are covariant with other patient-specific risks. We assumed that the endoscopic components of the Rockall score were present in our cohort at the same frequency as in the original derivation and are independent of known patient risk factors.¹⁶ For example, 441 out of 4,025 patients in the original Rockall derivation cohort presented with a systolic blood pressure less than 100 mm Hg. We assumed that an independent and random 10.96% of the cohort would present with shock, which confers two points in the Rockall score.

The population was replicated 60 times, with identical copies of the population resuming anticoagulation on each of days 1-60 (where day zero represents hemostasis). Intermediate data regarding our simulated population can be found in the Supplemental Appendix and in prior work.

Event Type, Severity, and Mortality

Each patient in our simulation could sustain several discrete and independent events: ischemic stroke, intracranial hemorrhage, recurrent UGIB, or extracranial major hemorrhage other than recurrent UGIB. As in prior analyses using the PADMA model, we did not consider minor hemorrhagic events.⁸

The probability of each event was conditional on the corresponding risk scoring system. Patient-specific probability of ischemic stroke was conditional on CHA₂DS₂-Vasc score.^21,22 Patient-specific probability of intracranial hemorrhage was conditional on HAS-BLED score, with the proportions of intracranial hemorrhage of each considered subtype (intracerebral, subarachnoid, or subdural) bootstrapped from previously-published data.^21-24 Patient-specific probability of rebleeding was conditional on Rockall score from the combined Rockall and Vreeburg validation cohorts.¹⁷ Patient-specific probability of extracranial major hemorrhage was conditional on HAS-BLED score.²¹ To avoid double-counting of UGIB, we subtracted the baseline risk of UGIB from the overall rate of extracranial major hemorrhages using previously-published data regarding relative frequency and a bootstrapping approach.²⁵

Probability of Rebleeding Over Time

To estimate the decrease in rebleeding risk over time, we searched the Medline database for systematic reviews of recurrent bleeding following UGIB using the strategy detailed in the Supplemental Appendix. Using the interval rates of rebleeding we identified, we calculated implied daily rates of rebleeding at the midpoint of each interval. For example, 39.5% of rebleeding events occurred within three days of hemostasis, implying a daily rate of approximately 13.2% on day two (32 of 81 events over a three-day period). We repeated this process to estimate daily rates at the midpoint of each reported time interval and fitted an exponential decay function.²⁶ Our exponential fitted these datapoints quite well, but we lacked sufficient data to test other survival functions (eg, Gompertz, lognormal, etc.). Our fitted exponential can be expressed as:

P _rebleeding = b ₀ *exp(b ₁ *day)

where b₀ = 0.1843 (SE: 0.0136) and b₁ = –0.1563 (SE: 0.0188). For example, a mean of 3.9% of rebleeding episodes will occur on day 10 (0.1843 *exp(–0.1563 *10)).

Relative Risks of Events with Anticoagulation

For patients resuming warfarin, the probabilities of each event were adjusted based on patient-specific daily INR. All INRs were assumed to be 1.0 until the day of warfarin reinitiation, after which interpolated trajectories of postinitiation INR measurements were sampled for each patient from an earlier study of clinical warfarin initiation.²⁷ Relative risks of ischemic stroke and hemorrhagic events were calculated based on each day’s INR.

For patients taking apixaban, we assumed that the medication would reach full therapeutic effect one day after reinitiation. Based on available evidence, we applied the relative risks of each event with apixaban compared with warfarin.²⁵

Future Disability and Mortality

Each event in our simulation resulted in hospitalization. Length of stay was sampled for each diagnosis.²⁸ The disutility of hospitalization was estimated based on length of stay.⁸ Inpatient mortality and future disability were predicted for each event as previously described.⁸ We assumed that recurrent episodes of UGIB conferred morbidity and mortality identical to extracranial major hemorrhages more broadly.^29,30

Disutilities

We used a multiplicative model for disutility with baseline utilities conditional on age and sex.³¹ Each day after resumption of anticoagulation carried a disutility of 0.012 for warfarin or 0.002 for apixaban, which we assumed to be equivalent to aspirin in disutility.³² Long-term disutility and life expectancy were conditional on modified Rankin Score (mRS).^33,34 We discounted all QALYs to day zero using standard exponential discounting and a discount rate centered at 3%. We then computed the average discounted QALYs among the cohort of patients that resumed anticoagulation on each day following the index UGIB.

Sensitivity Analyses and Metamodel

To assess sensitivity to continuously varying input parameters, such as discount rate, the proportion of extracranial major hemorrhages that are upper GI bleeds, and inpatient mortality from extracranial major hemorrhage, we constructed a metamodel (a regression model of our microsimulation results).³⁵ We tested for interactions among input parameters and dropped parameters that were not statistically significant predictors of discounted QALYs from our metamodel. We then tested for interactions between each parameter and day resuming anticoagulation to determine which factors may impact the optimal day of reinitiation. Finally, we used predicted marginal effects from our metamodel to assess the change in optimal day across the ranges of each input parameter when other parameters were held at their medians.

Resuming warfarin on day zero produced the fewest QALYs. With delay in reinitiation of anticoagulation, expected QALYs increased, peaked, and then declined for all scenarios. In our base-case simulation of warfarin, peak utility was achieved by resumption 41 days after the index UGIB. Resumption between days 32 and 51 produced greater than 99.9% of peak utility. In our base-case simulation of apixaban, peak utility was achieved by resumption 32 days after the index UGIB. Resumption between days 21 and 47 produced greater than 99.9% of peak utility. Results for warfarin and apixaban are shown in Figures 1 and 2, respectively.

The optimal day of warfarin reinitiation was most sensitive to CHA₂DS₂-Vasc scores and varied by around 11 days between a CHA₂DS₂-Vasc score of one and a CHA₂DS₂-Vasc score of six (the 5^th and 95^th percentiles, respectively) when all other parameters are held at their medians. Results were comparatively insensitive to rebleeding risk. Varying Rockall score from two to seven (the 5^th and 95^th percentiles, respectively) added three days to optimal warfarin resumption. Varying other parameters from the 5^th to the 95^th percentile (including HAS-BLED score, sex, age, and discount rate) changed expected QALYs but did not change the optimal day of reinitiation of warfarin. Optimal day of reinitiation for warfarin stratified by CHA₂DS₂-Vasc score is shown in Table 2.

Anticoagulation is frequently prescribed for patients with NVAF, and hemorrhagic complications are common. Although anticoagulants are withheld following hemorrhages, scant evidence to inform the optimal timing of reinitiation is available. In this microsimulation analysis, we found that the optimal time to reinitiate anticoagulation following UGIB is around 41 days for warfarin and around 32 days for apixaban. We have further demonstrated that the optimal timing of reinitiation can vary by nearly two weeks, depending on a patient’s underlying risk of stroke, and that early reinitiation is more sensitive to rebleeding risk than late reinitiation.

Prior work has shown that early reinitiation of anticoagulation leads to higher rates of recurrent hemorrhage while failure to reinitiate anticoagulation is associated with higher rates of stroke and mortality.^1-4,36 Our results add to the literature in a number of important ways. First, our model not only confirms that anticoagulation should be restarted but also suggests when this action should be taken. The competing risks of bleeding and stroke have left clinicians with little guidance; we have quantified the clinical reasoning required for the decision to resume anticoagulation. Second, by including the disutility of hospitalization and long-term disability, our model more accurately represents the complex tradeoffs between recurrent hemorrhage and (potentially disabling) stroke than would a comparison of event rates. Third, our model is conditional upon patient risk factors, allowing clinicians to personalize the timing of anticoagulation resumption. Theory would suggest that patients at higher risk of ischemic stroke benefit from earlier resumption of anticoagulation, while patients at higher risk of hemorrhage benefit from delayed reinitiation. We have quantified the extent to which patient-specific risks should change timing. Fourth, we offer a means of improving expected health outcomes that requires little more than appropriate scheduling. Current practice regarding resuming anticoagulation is widely variable. Many patients never resume warfarin, and those that do resume do so after highly varied periods of time.^1-5,36 We offer a means of standardizing clinical practice and improving expected patient outcomes.

Disclosures

The authors have nothing to disclose.

Funding

The authors received no specific funding for this work.

Anticoagulation is commonly used in the management of atrial fibrillation to reduce the risk of ischemic stroke. Warfarin and other anticoagulants increase the risk of hemorrhagic complications, including upper gastrointestinal bleeding (UGIB). Following UGIB, management of anticoagulation is highly variable. Many patients permanently discontinue anticoagulation, while others continue without interruption.^1-4 Among patients who resume warfarin, different cohorts have measured median times to resumption ranging from four days to 50 days.^1-3 Outcomes data are sparse, and clinical guidelines offer little direction.⁵

Following UGIB, the balance between the risks and benefits of anticoagulation changes over time. Rebleeding risk is highest immediately after the event and declines quickly; therefore, rapid resumption of anticoagulation causes patient harm.³ Meanwhile, the risk of stroke remains constant, and delay in resumption of anticoagulation is associated with increased risk of stroke and death.¹ At some point in time following the initial UGIB, the expected harm from bleeding would equal the expected harm from stroke. This time point would represent the optimal time to restart anticoagulation.

Trial data are unlikely to identify the optimal time for restarting anticoagulation. A randomized trial comparing discrete reinitiation times (eg, two weeks vs six weeks) may easily miss the optimal timing. Moreover, because the daily probability of thromboembolic events is low, large numbers of patients would be required to power such a study. In addition, a number of oral anticoagulants are now approved for prevention of thromboembolic stroke in atrial fibrillation, and each drug may have different optimal timing.

In contrast to randomized trials that would be impracticable for addressing this clinical issue, microsimulation modeling can provide granular information regarding the optimal time to restart anticoagulation. Herein, we set out to estimate the expected benefit of reinitiation of warfarin, the most commonly used oral anticoagulant,⁶ or apixaban, the direct oral anticoagulant with the most favorable risk profile,⁷ as a function of days after UGIB.

We previously described a microsimulation model of anticoagulation among patients with nonvalvular atrial fibrillation (NVAF; hereafter, we refer to this model as the Personalized Anticoagulation Decision-Making Assistance model, or PADMA).^8,9 For this study, we extended this model to incorporate the probability of rebleeding following UGIB and include apixaban as an alternative to warfarin. This model begins with a synthetic population following UGIB, the members of which are at varying risk for thromboembolism, recurrent UGIB, and other hemorrhages. For each patient, the model simulates a number of possible events (eg, thromboembolic stroke, intracranial hemorrhage, rebleeding, and other major extracranial hemorrhages) on each day of an acute period of 90 days after hemostasis. Patients who survive until the end of the acute period enter a simulation with annual, rather than daily, cycles. Our model then estimates total quality-adjusted life-years (QALYs) for each patient, discounted to the present. We report the average discounted QALYs produced by the model for the same population if all individuals in our input population were to resume either warfarin or apixaban on a specific day. Input parameters and ranges are summarized in Table 1, a simplified schematic of our model is shown in the Supplemental Appendix, and additional details regarding model structure and assumptions can be found in earlier work.^8,9 We simulated from a health system perspective over a lifelong time horizon. All analyses were performed in version 14 of Stata (StataCorp, LLC, College Station, Texas).

Synthetic Population

To generate a population reflective of the comorbidities and age distribution of the US population with NVAF, we merged relevant variables from the National Health and Nutrition Examination Survey (NHANES; 2011-2012), using multiple imputation to correct for missing variables.¹⁰ We then bootstrapped to national population estimates by age and sex to arrive at a hypothetical population of the United States.¹¹ Because NHANES does not include atrial fibrillation, we applied sex- and age-specific prevalence rates from the AnTicoagulation and Risk Factors In Atrial Fibrillation study.¹² We then calculated commonly used risk scores (CHA₂DS₂-Vasc and HAS-BLED) for each patient and limited the population to patients with a CHA₂DS₂-Vasc score of one or greater.^13,14 The population resuming apixaban was further limited to patients whose creatinine clearance was 25 mL/min or greater in keeping with the entry criteria in the phase 3 clinical trial on which the medication’s approval was based.¹⁵

To estimate patient-specific probability of rebleeding, we generated a Rockall score for each patient.¹⁶ Although the discrimination of the Rockall score is limited for individual patients, as with all other tools used to predict rebleeding following UGIB, the Rockall score has demonstrated reasonable calibration across a threefold risk gradient.^17-19 International consensus guidelines recommend the Rockall score as one of two risk prediction tools for clinical use in the management of patients with UGIB.²⁰ In addition, because the Rockall score includes some demographic components (five of a possible 11 points), our estimates of rebleeding risk are covariant with other patient-specific risks. We assumed that the endoscopic components of the Rockall score were present in our cohort at the same frequency as in the original derivation and are independent of known patient risk factors.¹⁶ For example, 441 out of 4,025 patients in the original Rockall derivation cohort presented with a systolic blood pressure less than 100 mm Hg. We assumed that an independent and random 10.96% of the cohort would present with shock, which confers two points in the Rockall score.

The population was replicated 60 times, with identical copies of the population resuming anticoagulation on each of days 1-60 (where day zero represents hemostasis). Intermediate data regarding our simulated population can be found in the Supplemental Appendix and in prior work.

Event Type, Severity, and Mortality

Each patient in our simulation could sustain several discrete and independent events: ischemic stroke, intracranial hemorrhage, recurrent UGIB, or extracranial major hemorrhage other than recurrent UGIB. As in prior analyses using the PADMA model, we did not consider minor hemorrhagic events.⁸

The probability of each event was conditional on the corresponding risk scoring system. Patient-specific probability of ischemic stroke was conditional on CHA₂DS₂-Vasc score.^21,22 Patient-specific probability of intracranial hemorrhage was conditional on HAS-BLED score, with the proportions of intracranial hemorrhage of each considered subtype (intracerebral, subarachnoid, or subdural) bootstrapped from previously-published data.^21-24 Patient-specific probability of rebleeding was conditional on Rockall score from the combined Rockall and Vreeburg validation cohorts.¹⁷ Patient-specific probability of extracranial major hemorrhage was conditional on HAS-BLED score.²¹ To avoid double-counting of UGIB, we subtracted the baseline risk of UGIB from the overall rate of extracranial major hemorrhages using previously-published data regarding relative frequency and a bootstrapping approach.²⁵

Probability of Rebleeding Over Time

To estimate the decrease in rebleeding risk over time, we searched the Medline database for systematic reviews of recurrent bleeding following UGIB using the strategy detailed in the Supplemental Appendix. Using the interval rates of rebleeding we identified, we calculated implied daily rates of rebleeding at the midpoint of each interval. For example, 39.5% of rebleeding events occurred within three days of hemostasis, implying a daily rate of approximately 13.2% on day two (32 of 81 events over a three-day period). We repeated this process to estimate daily rates at the midpoint of each reported time interval and fitted an exponential decay function.²⁶ Our exponential fitted these datapoints quite well, but we lacked sufficient data to test other survival functions (eg, Gompertz, lognormal, etc.). Our fitted exponential can be expressed as:

P _rebleeding = b ₀ *exp(b ₁ *day)

where b₀ = 0.1843 (SE: 0.0136) and b₁ = –0.1563 (SE: 0.0188). For example, a mean of 3.9% of rebleeding episodes will occur on day 10 (0.1843 *exp(–0.1563 *10)).

Relative Risks of Events with Anticoagulation

For patients resuming warfarin, the probabilities of each event were adjusted based on patient-specific daily INR. All INRs were assumed to be 1.0 until the day of warfarin reinitiation, after which interpolated trajectories of postinitiation INR measurements were sampled for each patient from an earlier study of clinical warfarin initiation.²⁷ Relative risks of ischemic stroke and hemorrhagic events were calculated based on each day’s INR.

For patients taking apixaban, we assumed that the medication would reach full therapeutic effect one day after reinitiation. Based on available evidence, we applied the relative risks of each event with apixaban compared with warfarin.²⁵

Future Disability and Mortality

Each event in our simulation resulted in hospitalization. Length of stay was sampled for each diagnosis.²⁸ The disutility of hospitalization was estimated based on length of stay.⁸ Inpatient mortality and future disability were predicted for each event as previously described.⁸ We assumed that recurrent episodes of UGIB conferred morbidity and mortality identical to extracranial major hemorrhages more broadly.^29,30

Disutilities

We used a multiplicative model for disutility with baseline utilities conditional on age and sex.³¹ Each day after resumption of anticoagulation carried a disutility of 0.012 for warfarin or 0.002 for apixaban, which we assumed to be equivalent to aspirin in disutility.³² Long-term disutility and life expectancy were conditional on modified Rankin Score (mRS).^33,34 We discounted all QALYs to day zero using standard exponential discounting and a discount rate centered at 3%. We then computed the average discounted QALYs among the cohort of patients that resumed anticoagulation on each day following the index UGIB.

Sensitivity Analyses and Metamodel

To assess sensitivity to continuously varying input parameters, such as discount rate, the proportion of extracranial major hemorrhages that are upper GI bleeds, and inpatient mortality from extracranial major hemorrhage, we constructed a metamodel (a regression model of our microsimulation results).³⁵ We tested for interactions among input parameters and dropped parameters that were not statistically significant predictors of discounted QALYs from our metamodel. We then tested for interactions between each parameter and day resuming anticoagulation to determine which factors may impact the optimal day of reinitiation. Finally, we used predicted marginal effects from our metamodel to assess the change in optimal day across the ranges of each input parameter when other parameters were held at their medians.

Resuming warfarin on day zero produced the fewest QALYs. With delay in reinitiation of anticoagulation, expected QALYs increased, peaked, and then declined for all scenarios. In our base-case simulation of warfarin, peak utility was achieved by resumption 41 days after the index UGIB. Resumption between days 32 and 51 produced greater than 99.9% of peak utility. In our base-case simulation of apixaban, peak utility was achieved by resumption 32 days after the index UGIB. Resumption between days 21 and 47 produced greater than 99.9% of peak utility. Results for warfarin and apixaban are shown in Figures 1 and 2, respectively.

The optimal day of warfarin reinitiation was most sensitive to CHA₂DS₂-Vasc scores and varied by around 11 days between a CHA₂DS₂-Vasc score of one and a CHA₂DS₂-Vasc score of six (the 5^th and 95^th percentiles, respectively) when all other parameters are held at their medians. Results were comparatively insensitive to rebleeding risk. Varying Rockall score from two to seven (the 5^th and 95^th percentiles, respectively) added three days to optimal warfarin resumption. Varying other parameters from the 5^th to the 95^th percentile (including HAS-BLED score, sex, age, and discount rate) changed expected QALYs but did not change the optimal day of reinitiation of warfarin. Optimal day of reinitiation for warfarin stratified by CHA₂DS₂-Vasc score is shown in Table 2.

Anticoagulation is frequently prescribed for patients with NVAF, and hemorrhagic complications are common. Although anticoagulants are withheld following hemorrhages, scant evidence to inform the optimal timing of reinitiation is available. In this microsimulation analysis, we found that the optimal time to reinitiate anticoagulation following UGIB is around 41 days for warfarin and around 32 days for apixaban. We have further demonstrated that the optimal timing of reinitiation can vary by nearly two weeks, depending on a patient’s underlying risk of stroke, and that early reinitiation is more sensitive to rebleeding risk than late reinitiation.

Prior work has shown that early reinitiation of anticoagulation leads to higher rates of recurrent hemorrhage while failure to reinitiate anticoagulation is associated with higher rates of stroke and mortality.^1-4,36 Our results add to the literature in a number of important ways. First, our model not only confirms that anticoagulation should be restarted but also suggests when this action should be taken. The competing risks of bleeding and stroke have left clinicians with little guidance; we have quantified the clinical reasoning required for the decision to resume anticoagulation. Second, by including the disutility of hospitalization and long-term disability, our model more accurately represents the complex tradeoffs between recurrent hemorrhage and (potentially disabling) stroke than would a comparison of event rates. Third, our model is conditional upon patient risk factors, allowing clinicians to personalize the timing of anticoagulation resumption. Theory would suggest that patients at higher risk of ischemic stroke benefit from earlier resumption of anticoagulation, while patients at higher risk of hemorrhage benefit from delayed reinitiation. We have quantified the extent to which patient-specific risks should change timing. Fourth, we offer a means of improving expected health outcomes that requires little more than appropriate scheduling. Current practice regarding resuming anticoagulation is widely variable. Many patients never resume warfarin, and those that do resume do so after highly varied periods of time.^1-5,36 We offer a means of standardizing clinical practice and improving expected patient outcomes.

Disclosures

The authors have nothing to disclose.

Funding

The authors received no specific funding for this work.

Historically, academic medicine faculty were predominantly physician-scientists.¹ During the past decade, the number of clinician-educators and nontenured clinicians has grown.² Many academically oriented clinical faculty at our institution would like to participate in and learn how to conduct quality scholarship. While institutional requirements vary, scholarly work is often required for promotion,³ and faculty may also desire to support the scholarly work of residents. Moreover, a core program component of the Accreditation Council of Graduate Medical Education standards requires faculty to “maintain an environment of inquiry and scholarship with an active research component.”⁴ Yet clinical faculty often find academic projects to be challenging. Similar to residents, clinical academic faculty frequently lack formal training in health services research or quality improvement science, have insufficient mentorship, and typically have limited uncommitted time and resources.⁵

One approach to this problem has been to pair junior clinicians with traditional physician scientists as mentors.^6,7 This type of mentorship for clinical faculty is increasingly difficult to access because of growing pressure on physician-scientist faculty to conduct their own research, seek extramural funding, meet clinical expectations, and mentor fellows and faculty in their own disciplines.⁸ Moreover, senior research faculty may not be prepared or have the time to teach junior faculty how to deal with common stumbling blocks (eg, institutional review board [IRB] applications, statistically testable hypothesis development, and statistical analysis).^8,9 Seminars or works-in-progress sessions are another strategy to bolster scholarly work, but the experience at our institution is that such sessions are often not relevant at the time of delivery and can be intimidating to clinical faculty who lack extensive knowledge about research methods and prior research experience.

Another approach to supporting the research efforts of academic clinicians is to fund a consulting statistician. However, without sufficient content expertise, statisticians may be frustrated in their efforts to assist clinicians who struggle to formulate a testable question or to work directly with data collected. Statisticians may be inexperienced in writing IRB applications or implementing protocols in a clinical or educational setting. Furthermore, statistical consultations are often limited in scope¹⁰ and, in our setting, rarely produce a durable improvement in the research skills of the faculty member or the enduring partnership required to complete a longer-term project. Because of these shortcomings, we have found that purely statistical support resources are often underutilized and ineffective.

Other models to facilitate scholarship have been employed, but few focus on facilitating scholarship of clinical faculty. One strategy involved supporting hospitalist’s academic productivity by reducing hospitalists’ full-time equivalent (FTE) and providing mentorship.¹¹ For many, this approach is likely cost-prohibitive. Others have focused primarily on resident and fellow scholarships.^5,6

In this report, we describe an educational innovation to educate and support the scholarly work of academic hospitalists and internists by using an academic research coach. We recruited a health researcher with extensive experience in research methods and strong interpersonal skills with the ability to explain and teach research concepts in an accessible manner. We sought an individual who would provide high-yield single consultations, join project teams to provide ongoing mentorship from conception to completion, and consequently, bolster scholarly productivity and learning among nonresearch clinicians in our Division. We anticipated that providing support for multiple aspects of a project would be more likely to help faculty overcome barriers to research and disseminate their project results as scholarly output.

The coach initiative was implemented in the Division of General Internal Medicine at the University of Washington. The Division has over 200 members (60 hospitalists), including clinical instructors and acting instructors, who have not yet been appointed to the regular faculty (clinician-educators and physician scientists), and full-time clinical faculty. Division members staff clinical services at four area hospitals and 10 affiliated internal medicine and specialty clinics. Eligible clients were all Division members, although the focus of the initial program targeted hospitalists at our three primary teaching hospitals. Fellows, residents, students, and faculty from within and outside the Division were welcome to participate in a project involving coaching as long as a Division faculty member was engaged in the project.

Program Description

The overall goal of the coach initiative was to support the scholarly work of primarily clinical Division members. Given our focus was on clinical faculty with little training on research methodology, we did not expect the coach to secure grant funding for the position. Instead, we aimed to increase the quality and quantity of scholarship through publications, abstracts, and small grants. We defined scholarly work broadly: clinical research, quality improvement, medical education research, and other forms of scientific inquiry or synthesis. The coach was established as a 0.50 FTE position with a 12-month annually renewable appointment. The role was deemed that of a coach instead of a mentor because the coach was available to all Division members and involved task-oriented consultations with check-ins to facilitate projects, rather than a deeper more developmental relationship that typically exists with mentoring. The Division leadership identified support for scholarly activity as a high priority and mentorship as an unmet need based on faculty feedback. Clinical revenue supported the position.

Necessary qualifications, determined prior to hiring, included a PhD in health services or related field (eg, epidemiology) or a master’s degree with five years of experience in project management, clinical research, and study design. The position also called for expertise in articulating research questions, selecting study designs, navigating the IRB approval process, collecting/managing data, analyzing statistics, and mentoring and teaching clinical faculty in their scholarly endeavors. A track record in generating academic output (manuscripts and abstracts at regional/national meetings) was required. We circulated a description of the position to Division faculty and to leadership in our School of Public Health.

Based on these criteria, an inaugural coach was hired (author C.M.M.). The coach had a PhD in epidemiology, 10 years of research experience, 16 publications, and had recently finished a National Institutes of Health (NIH) career development award. At the time of hiring, she was a Clinical Assistant Professor in the School of Dentistry, which provided additional FTE. She had no extramural funding but was applying for NIH-level grants and had received several small grants.

To ensure uptake of the coach’s services, we realized that it was necessary to delineate the scope of services available, clarify availability of the coach, and define expectations regarding authorship. We used an iterative process that took into consideration the coach’s expertise, services most needed by the Division’s clinicians, and discussions with Division leadership and faculty at faculty meetings across hospitals and clinics. A range of services and authorship expectations were defined. Consensus was reached that the coach should be invited to coauthor projects where design, analysis, and/or substantial intellectual content was provided and for which authorship criteria were met.¹² Collegial reviews by the coach of already developed manuscripts and time-limited, low-intensity consultations that did not involve substantial intellectual contributions did not warrant authorship.¹² On this basis, we created and distributed a flyer to publicize these guidelines and invite Division members to contact the coach (Figure 1).

The coach attended Division, section, and clinical group meetings to publicize the initiative. The coach also individually met with faculty throughout the Division, explained her role, described services available, and answered questions. The marketing effort was continuous and calibrated with more or less exposure depending on existing projects and the coach’s availability. In addition, the coach coordinated with the director of the Division’s faculty development program to cohost works-in-progress seminars, identify coach clients to present at these meetings, and provide brief presentations on a basic research skill at meetings. Faculty built rapport with the coach through these activities and became more comfortable reaching out for assistance. Because of the large size of the Division, it was decided to roll out the initiative in a stepwise fashion, starting with hospitalists before expanding to the rest of the Division.

Most faculty contacted the coach by e-mail to request a consultation, at which time the coach requested that they complete a preconsultation handout (Figure 2). Initial coaching appointments lasted one hour and were in-person. Coaching entailed an in-depth analysis of the project plan and advice on how to move the project forward. The coach provided tailored scholarly project advice and expertise in research methods. After initial consultations, she would review grant proposals, IRB applications, manuscripts, case report forms, abstracts, and other products. Her efforts typically focused on improving the methods and scientific and technical writing. Assistance with statistical analysis was provided on a case-by-case basis to maintain broad availability. To address statistically complex questions, the coach had five hours of monthly access to a PhD biostatistician via an on-campus consulting service. Follow-up appointments were encouraged and provided as needed by e-mail, phone, or in-person. The coach conducted regular reach outs to facilitate projects. However, execution of the research was generally the responsibility of the faculty member.

Program Evaluation

To characterize the reach and scope of the program, the coach tracked the number of faculty supported, types of services provided, status of initiated projects, numbers of grants generated, and the dissemination of scholarly products including papers and abstracts. We used these metrics to create summary reports to identify successes and areas for improvement. Monthly meetings between the coach and Division leadership were used to fine-tune the approach.

We surveyed coach clients anonymously to assess their satisfaction with the coach initiative. Using Likert scale questions where 1 = completely disagree and 5 = completely agree, we asked (1) if they would recommend the coach to colleagues, (2) if their work was higher quality because of the coach, (3) if they were overall satisfied with the coach, (4) whether the Division should continue to support the coach, and (5) if the coach’s lack of clinical training negatively affected their experience. This work was considered a quality improvement initiative for which IRB approval was not required.

Over 18 months, the coach supported a 49 Division members including 30 hospitalists and 63 projects. Projects included a wide range of scholarship: medical education research, qualitative research, clinical quality improvement projects, observational studies, and a randomized clinical trial. Many clients (n = 16) used the coach for more than one project. The scope of work included limited support projects (identifying research resource and brainstorming project feasibility) lasting one to two sessions (n = 25), projects with a limited scope (collegial reviews of manuscripts and assistance with IRB submissions) but requiring more than two consultations (n = 24), and ongoing in-depth support projects (contributions on design, data collection, analysis, and manuscript writing) that required three consultations or more (n = 14). The majority of Division members (75%) supported did not have master’s level training in a health services-related area, six had NIH or other national-level funding, and two had small grants funded by local sources prior to providing support. The number of Division faculty on a given project ranged from one to four.

The coach directly supported 13 manuscripts with coach authorship, seven manuscripts without authorship, 11 abstracts, and four grant submissions (Appendix). The coach was a coauthor on all the abstracts and a coinvestigator on the grant applications. Of the 13 publications the coach coauthored, 11 publications have been accepted to peer-reviewed journals and two are currently in the submission process. The types of articles published included one medical evaluation report, one qualitative study, one randomized clinical trial, three quality assessment/improvement reports, and five epidemiologic studies. The types of abstracts included one qualitative report, one systematic review, one randomized clinical trial, two quality improvement projects, two epidemiologic studies, and four medical education projects. Three of four small grants submitted to local and national funders were funded.

The coach’s influence extended beyond the Division. Forty-eight university faculty, fellows, or students not affiliated with general internal medicine benefited from coach coaching: 26 were authors on papers and/or abstracts coauthored by the coach, 17 on manuscripts the coach reviewed without authorship, and five participated in consultations.

The coach found the experience rewarding. She enjoyed working on the methodologic aspects of projects and benefited from being included as coauthor on papers.

Twenty-nine of the 43 faculty (67%) still at the institution responded to the program assessment survey. Faculty strongly agreed that they would recommend the coach to colleagues (average ± standard deviation [SD]: 4.7 ± 0.5), that it improved the quality of their work (4.5 ± 0.9), that they were overall satisfied with the coaching (4.6 ± 0.7), and that the Division should continue to support the coach (4.9 ± 0.4). Faculty did not agree that the lack of clinical training of the coach was a barrier (2.0 ± 1.3).

The coach program was highly utilized, well regarded, and delivered substantial, tangible, and academic output. We anticipate the coach initiative will continue to be a valuable resource for our Division and could prove to be a valuable model for other institutions seeking to bolster the scholarly work of clinical academicians.

Several lessons emerged through the course of this project. First, we realized it is essential to select a coach who is both knowledgeable and approachable. We found that after meeting the coach, many faculty sought her help who otherwise would not have. An explicit, ongoing marketing strategy with regular contact with faculty at meetings was a key to receiving consult requests.

Second, the lack of a clinical background did not seem to hinder the coach’s ability to coach clinicians. The coach acknowledged her lack of clinical experience and relied on clients to explain the clinical context of projects. We also learned that the coach’s substantial experience with the logistics of research was invaluable. For example, the coach had substantial experience with the IRB process and her pre-reviews of IRB applications made for a short and relatively seamless experience navigating the IRB process. The coach also facilitated collaborations and leveraged existing resources at our institution. For example, for a qualitative research project, the coach helped identify a health services faculty member with this specific expertise, which led to a successful collaboration and publication. Although a more junior coach with less established qualifications may be helpful with research methods and with the research process, our endeavor suggests that having a more highly trained and experienced researcher was extremely valuable. Finally, we learned that for a Division of our size, the 0.50 FTE allotted to the coach is a minimum requirement. The coach spent approximately four hours a week on marketing, attending faculty meetings and conducting brief didactics, two hours per week on administration, and 14 hours per week on consultations. Faculty generally received support soon after their requests, but there were occasional wait times, which may have delayed some projects.

Academic leaders at our institution have noted the success of our coach initiative and have created a demand for coach services. We are exploring funding models that would allow for the expansion of coach services to other departments and divisions. We are in the initial stages of creating an Academic Scholarship Support Core under the supervision of the coach. Within this Core, we envision that various research support services will be triaged to staff with appropriate expertise; for example, a regulatory coordinator would review IRB applications while a master’s level statistician would conduct statistical analyses.

We have also transitioned to a new coach and have continued to experience success with the program. Our initial coach (author C.M.M.) obtained an NIH R01, a foundation grant, and took over a summer program that trains dental faculty in clinical research methods leaving insufficient time for coaching. Our new coach also has a PhD in epidemiology with NIH R01 funding but has more available FTE. Both of our coaches are graduates of our School of Public Health and institutions with such schools may have good access to the expertise needed. Nonclinical PhDs are often almost entirely reliant on grants, and some nongrant support is often attractive to these researchers. Additionally, PhDs who are junior or mid-career faculty that have the needed training are relatively affordable, particularly when the resource is made available to large number of faculty. For example, our first coach cost $48,000 a year for 50% FTE.

A limitation to our assessment of the coach initiative was the lack of pre- and postintervention metrics of scholarly productivity. We cannot definitively say that the Division’s scholarly output has increased because of the coach. Nevertheless, we are confident that the coach’s coaching has enhanced the scholarly work of individual clinicians and provided value to the Division as a whole. The coach program has been a success in our Division. Other institutions facing the challenge of supporting the research efforts of academic clinicians may consider this model as a worthy investment.

Disclosures

The authors have nothing to disclose.

Historically, academic medicine faculty were predominantly physician-scientists.¹ During the past decade, the number of clinician-educators and nontenured clinicians has grown.² Many academically oriented clinical faculty at our institution would like to participate in and learn how to conduct quality scholarship. While institutional requirements vary, scholarly work is often required for promotion,³ and faculty may also desire to support the scholarly work of residents. Moreover, a core program component of the Accreditation Council of Graduate Medical Education standards requires faculty to “maintain an environment of inquiry and scholarship with an active research component.”⁴ Yet clinical faculty often find academic projects to be challenging. Similar to residents, clinical academic faculty frequently lack formal training in health services research or quality improvement science, have insufficient mentorship, and typically have limited uncommitted time and resources.⁵

One approach to this problem has been to pair junior clinicians with traditional physician scientists as mentors.^6,7 This type of mentorship for clinical faculty is increasingly difficult to access because of growing pressure on physician-scientist faculty to conduct their own research, seek extramural funding, meet clinical expectations, and mentor fellows and faculty in their own disciplines.⁸ Moreover, senior research faculty may not be prepared or have the time to teach junior faculty how to deal with common stumbling blocks (eg, institutional review board [IRB] applications, statistically testable hypothesis development, and statistical analysis).^8,9 Seminars or works-in-progress sessions are another strategy to bolster scholarly work, but the experience at our institution is that such sessions are often not relevant at the time of delivery and can be intimidating to clinical faculty who lack extensive knowledge about research methods and prior research experience.

Another approach to supporting the research efforts of academic clinicians is to fund a consulting statistician. However, without sufficient content expertise, statisticians may be frustrated in their efforts to assist clinicians who struggle to formulate a testable question or to work directly with data collected. Statisticians may be inexperienced in writing IRB applications or implementing protocols in a clinical or educational setting. Furthermore, statistical consultations are often limited in scope¹⁰ and, in our setting, rarely produce a durable improvement in the research skills of the faculty member or the enduring partnership required to complete a longer-term project. Because of these shortcomings, we have found that purely statistical support resources are often underutilized and ineffective.

Other models to facilitate scholarship have been employed, but few focus on facilitating scholarship of clinical faculty. One strategy involved supporting hospitalist’s academic productivity by reducing hospitalists’ full-time equivalent (FTE) and providing mentorship.¹¹ For many, this approach is likely cost-prohibitive. Others have focused primarily on resident and fellow scholarships.^5,6

In this report, we describe an educational innovation to educate and support the scholarly work of academic hospitalists and internists by using an academic research coach. We recruited a health researcher with extensive experience in research methods and strong interpersonal skills with the ability to explain and teach research concepts in an accessible manner. We sought an individual who would provide high-yield single consultations, join project teams to provide ongoing mentorship from conception to completion, and consequently, bolster scholarly productivity and learning among nonresearch clinicians in our Division. We anticipated that providing support for multiple aspects of a project would be more likely to help faculty overcome barriers to research and disseminate their project results as scholarly output.

The coach initiative was implemented in the Division of General Internal Medicine at the University of Washington. The Division has over 200 members (60 hospitalists), including clinical instructors and acting instructors, who have not yet been appointed to the regular faculty (clinician-educators and physician scientists), and full-time clinical faculty. Division members staff clinical services at four area hospitals and 10 affiliated internal medicine and specialty clinics. Eligible clients were all Division members, although the focus of the initial program targeted hospitalists at our three primary teaching hospitals. Fellows, residents, students, and faculty from within and outside the Division were welcome to participate in a project involving coaching as long as a Division faculty member was engaged in the project.

Program Description

The overall goal of the coach initiative was to support the scholarly work of primarily clinical Division members. Given our focus was on clinical faculty with little training on research methodology, we did not expect the coach to secure grant funding for the position. Instead, we aimed to increase the quality and quantity of scholarship through publications, abstracts, and small grants. We defined scholarly work broadly: clinical research, quality improvement, medical education research, and other forms of scientific inquiry or synthesis. The coach was established as a 0.50 FTE position with a 12-month annually renewable appointment. The role was deemed that of a coach instead of a mentor because the coach was available to all Division members and involved task-oriented consultations with check-ins to facilitate projects, rather than a deeper more developmental relationship that typically exists with mentoring. The Division leadership identified support for scholarly activity as a high priority and mentorship as an unmet need based on faculty feedback. Clinical revenue supported the position.

Necessary qualifications, determined prior to hiring, included a PhD in health services or related field (eg, epidemiology) or a master’s degree with five years of experience in project management, clinical research, and study design. The position also called for expertise in articulating research questions, selecting study designs, navigating the IRB approval process, collecting/managing data, analyzing statistics, and mentoring and teaching clinical faculty in their scholarly endeavors. A track record in generating academic output (manuscripts and abstracts at regional/national meetings) was required. We circulated a description of the position to Division faculty and to leadership in our School of Public Health.

Based on these criteria, an inaugural coach was hired (author C.M.M.). The coach had a PhD in epidemiology, 10 years of research experience, 16 publications, and had recently finished a National Institutes of Health (NIH) career development award. At the time of hiring, she was a Clinical Assistant Professor in the School of Dentistry, which provided additional FTE. She had no extramural funding but was applying for NIH-level grants and had received several small grants.

To ensure uptake of the coach’s services, we realized that it was necessary to delineate the scope of services available, clarify availability of the coach, and define expectations regarding authorship. We used an iterative process that took into consideration the coach’s expertise, services most needed by the Division’s clinicians, and discussions with Division leadership and faculty at faculty meetings across hospitals and clinics. A range of services and authorship expectations were defined. Consensus was reached that the coach should be invited to coauthor projects where design, analysis, and/or substantial intellectual content was provided and for which authorship criteria were met.¹² Collegial reviews by the coach of already developed manuscripts and time-limited, low-intensity consultations that did not involve substantial intellectual contributions did not warrant authorship.¹² On this basis, we created and distributed a flyer to publicize these guidelines and invite Division members to contact the coach (Figure 1).

The coach attended Division, section, and clinical group meetings to publicize the initiative. The coach also individually met with faculty throughout the Division, explained her role, described services available, and answered questions. The marketing effort was continuous and calibrated with more or less exposure depending on existing projects and the coach’s availability. In addition, the coach coordinated with the director of the Division’s faculty development program to cohost works-in-progress seminars, identify coach clients to present at these meetings, and provide brief presentations on a basic research skill at meetings. Faculty built rapport with the coach through these activities and became more comfortable reaching out for assistance. Because of the large size of the Division, it was decided to roll out the initiative in a stepwise fashion, starting with hospitalists before expanding to the rest of the Division.

Most faculty contacted the coach by e-mail to request a consultation, at which time the coach requested that they complete a preconsultation handout (Figure 2). Initial coaching appointments lasted one hour and were in-person. Coaching entailed an in-depth analysis of the project plan and advice on how to move the project forward. The coach provided tailored scholarly project advice and expertise in research methods. After initial consultations, she would review grant proposals, IRB applications, manuscripts, case report forms, abstracts, and other products. Her efforts typically focused on improving the methods and scientific and technical writing. Assistance with statistical analysis was provided on a case-by-case basis to maintain broad availability. To address statistically complex questions, the coach had five hours of monthly access to a PhD biostatistician via an on-campus consulting service. Follow-up appointments were encouraged and provided as needed by e-mail, phone, or in-person. The coach conducted regular reach outs to facilitate projects. However, execution of the research was generally the responsibility of the faculty member.

Program Evaluation

To characterize the reach and scope of the program, the coach tracked the number of faculty supported, types of services provided, status of initiated projects, numbers of grants generated, and the dissemination of scholarly products including papers and abstracts. We used these metrics to create summary reports to identify successes and areas for improvement. Monthly meetings between the coach and Division leadership were used to fine-tune the approach.

We surveyed coach clients anonymously to assess their satisfaction with the coach initiative. Using Likert scale questions where 1 = completely disagree and 5 = completely agree, we asked (1) if they would recommend the coach to colleagues, (2) if their work was higher quality because of the coach, (3) if they were overall satisfied with the coach, (4) whether the Division should continue to support the coach, and (5) if the coach’s lack of clinical training negatively affected their experience. This work was considered a quality improvement initiative for which IRB approval was not required.

Over 18 months, the coach supported a 49 Division members including 30 hospitalists and 63 projects. Projects included a wide range of scholarship: medical education research, qualitative research, clinical quality improvement projects, observational studies, and a randomized clinical trial. Many clients (n = 16) used the coach for more than one project. The scope of work included limited support projects (identifying research resource and brainstorming project feasibility) lasting one to two sessions (n = 25), projects with a limited scope (collegial reviews of manuscripts and assistance with IRB submissions) but requiring more than two consultations (n = 24), and ongoing in-depth support projects (contributions on design, data collection, analysis, and manuscript writing) that required three consultations or more (n = 14). The majority of Division members (75%) supported did not have master’s level training in a health services-related area, six had NIH or other national-level funding, and two had small grants funded by local sources prior to providing support. The number of Division faculty on a given project ranged from one to four.

The coach directly supported 13 manuscripts with coach authorship, seven manuscripts without authorship, 11 abstracts, and four grant submissions (Appendix). The coach was a coauthor on all the abstracts and a coinvestigator on the grant applications. Of the 13 publications the coach coauthored, 11 publications have been accepted to peer-reviewed journals and two are currently in the submission process. The types of articles published included one medical evaluation report, one qualitative study, one randomized clinical trial, three quality assessment/improvement reports, and five epidemiologic studies. The types of abstracts included one qualitative report, one systematic review, one randomized clinical trial, two quality improvement projects, two epidemiologic studies, and four medical education projects. Three of four small grants submitted to local and national funders were funded.

The coach’s influence extended beyond the Division. Forty-eight university faculty, fellows, or students not affiliated with general internal medicine benefited from coach coaching: 26 were authors on papers and/or abstracts coauthored by the coach, 17 on manuscripts the coach reviewed without authorship, and five participated in consultations.

The coach found the experience rewarding. She enjoyed working on the methodologic aspects of projects and benefited from being included as coauthor on papers.

Twenty-nine of the 43 faculty (67%) still at the institution responded to the program assessment survey. Faculty strongly agreed that they would recommend the coach to colleagues (average ± standard deviation [SD]: 4.7 ± 0.5), that it improved the quality of their work (4.5 ± 0.9), that they were overall satisfied with the coaching (4.6 ± 0.7), and that the Division should continue to support the coach (4.9 ± 0.4). Faculty did not agree that the lack of clinical training of the coach was a barrier (2.0 ± 1.3).

The coach program was highly utilized, well regarded, and delivered substantial, tangible, and academic output. We anticipate the coach initiative will continue to be a valuable resource for our Division and could prove to be a valuable model for other institutions seeking to bolster the scholarly work of clinical academicians.

Several lessons emerged through the course of this project. First, we realized it is essential to select a coach who is both knowledgeable and approachable. We found that after meeting the coach, many faculty sought her help who otherwise would not have. An explicit, ongoing marketing strategy with regular contact with faculty at meetings was a key to receiving consult requests.

Second, the lack of a clinical background did not seem to hinder the coach’s ability to coach clinicians. The coach acknowledged her lack of clinical experience and relied on clients to explain the clinical context of projects. We also learned that the coach’s substantial experience with the logistics of research was invaluable. For example, the coach had substantial experience with the IRB process and her pre-reviews of IRB applications made for a short and relatively seamless experience navigating the IRB process. The coach also facilitated collaborations and leveraged existing resources at our institution. For example, for a qualitative research project, the coach helped identify a health services faculty member with this specific expertise, which led to a successful collaboration and publication. Although a more junior coach with less established qualifications may be helpful with research methods and with the research process, our endeavor suggests that having a more highly trained and experienced researcher was extremely valuable. Finally, we learned that for a Division of our size, the 0.50 FTE allotted to the coach is a minimum requirement. The coach spent approximately four hours a week on marketing, attending faculty meetings and conducting brief didactics, two hours per week on administration, and 14 hours per week on consultations. Faculty generally received support soon after their requests, but there were occasional wait times, which may have delayed some projects.

Academic leaders at our institution have noted the success of our coach initiative and have created a demand for coach services. We are exploring funding models that would allow for the expansion of coach services to other departments and divisions. We are in the initial stages of creating an Academic Scholarship Support Core under the supervision of the coach. Within this Core, we envision that various research support services will be triaged to staff with appropriate expertise; for example, a regulatory coordinator would review IRB applications while a master’s level statistician would conduct statistical analyses.

We have also transitioned to a new coach and have continued to experience success with the program. Our initial coach (author C.M.M.) obtained an NIH R01, a foundation grant, and took over a summer program that trains dental faculty in clinical research methods leaving insufficient time for coaching. Our new coach also has a PhD in epidemiology with NIH R01 funding but has more available FTE. Both of our coaches are graduates of our School of Public Health and institutions with such schools may have good access to the expertise needed. Nonclinical PhDs are often almost entirely reliant on grants, and some nongrant support is often attractive to these researchers. Additionally, PhDs who are junior or mid-career faculty that have the needed training are relatively affordable, particularly when the resource is made available to large number of faculty. For example, our first coach cost $48,000 a year for 50% FTE.

A limitation to our assessment of the coach initiative was the lack of pre- and postintervention metrics of scholarly productivity. We cannot definitively say that the Division’s scholarly output has increased because of the coach. Nevertheless, we are confident that the coach’s coaching has enhanced the scholarly work of individual clinicians and provided value to the Division as a whole. The coach program has been a success in our Division. Other institutions facing the challenge of supporting the research efforts of academic clinicians may consider this model as a worthy investment.

Disclosures

The authors have nothing to disclose.

Historically, academic medicine faculty were predominantly physician-scientists.¹ During the past decade, the number of clinician-educators and nontenured clinicians has grown.² Many academically oriented clinical faculty at our institution would like to participate in and learn how to conduct quality scholarship. While institutional requirements vary, scholarly work is often required for promotion,³ and faculty may also desire to support the scholarly work of residents. Moreover, a core program component of the Accreditation Council of Graduate Medical Education standards requires faculty to “maintain an environment of inquiry and scholarship with an active research component.”⁴ Yet clinical faculty often find academic projects to be challenging. Similar to residents, clinical academic faculty frequently lack formal training in health services research or quality improvement science, have insufficient mentorship, and typically have limited uncommitted time and resources.⁵

One approach to this problem has been to pair junior clinicians with traditional physician scientists as mentors.^6,7 This type of mentorship for clinical faculty is increasingly difficult to access because of growing pressure on physician-scientist faculty to conduct their own research, seek extramural funding, meet clinical expectations, and mentor fellows and faculty in their own disciplines.⁸ Moreover, senior research faculty may not be prepared or have the time to teach junior faculty how to deal with common stumbling blocks (eg, institutional review board [IRB] applications, statistically testable hypothesis development, and statistical analysis).^8,9 Seminars or works-in-progress sessions are another strategy to bolster scholarly work, but the experience at our institution is that such sessions are often not relevant at the time of delivery and can be intimidating to clinical faculty who lack extensive knowledge about research methods and prior research experience.

Another approach to supporting the research efforts of academic clinicians is to fund a consulting statistician. However, without sufficient content expertise, statisticians may be frustrated in their efforts to assist clinicians who struggle to formulate a testable question or to work directly with data collected. Statisticians may be inexperienced in writing IRB applications or implementing protocols in a clinical or educational setting. Furthermore, statistical consultations are often limited in scope¹⁰ and, in our setting, rarely produce a durable improvement in the research skills of the faculty member or the enduring partnership required to complete a longer-term project. Because of these shortcomings, we have found that purely statistical support resources are often underutilized and ineffective.

Other models to facilitate scholarship have been employed, but few focus on facilitating scholarship of clinical faculty. One strategy involved supporting hospitalist’s academic productivity by reducing hospitalists’ full-time equivalent (FTE) and providing mentorship.¹¹ For many, this approach is likely cost-prohibitive. Others have focused primarily on resident and fellow scholarships.^5,6

In this report, we describe an educational innovation to educate and support the scholarly work of academic hospitalists and internists by using an academic research coach. We recruited a health researcher with extensive experience in research methods and strong interpersonal skills with the ability to explain and teach research concepts in an accessible manner. We sought an individual who would provide high-yield single consultations, join project teams to provide ongoing mentorship from conception to completion, and consequently, bolster scholarly productivity and learning among nonresearch clinicians in our Division. We anticipated that providing support for multiple aspects of a project would be more likely to help faculty overcome barriers to research and disseminate their project results as scholarly output.

The coach initiative was implemented in the Division of General Internal Medicine at the University of Washington. The Division has over 200 members (60 hospitalists), including clinical instructors and acting instructors, who have not yet been appointed to the regular faculty (clinician-educators and physician scientists), and full-time clinical faculty. Division members staff clinical services at four area hospitals and 10 affiliated internal medicine and specialty clinics. Eligible clients were all Division members, although the focus of the initial program targeted hospitalists at our three primary teaching hospitals. Fellows, residents, students, and faculty from within and outside the Division were welcome to participate in a project involving coaching as long as a Division faculty member was engaged in the project.

Program Description

The overall goal of the coach initiative was to support the scholarly work of primarily clinical Division members. Given our focus was on clinical faculty with little training on research methodology, we did not expect the coach to secure grant funding for the position. Instead, we aimed to increase the quality and quantity of scholarship through publications, abstracts, and small grants. We defined scholarly work broadly: clinical research, quality improvement, medical education research, and other forms of scientific inquiry or synthesis. The coach was established as a 0.50 FTE position with a 12-month annually renewable appointment. The role was deemed that of a coach instead of a mentor because the coach was available to all Division members and involved task-oriented consultations with check-ins to facilitate projects, rather than a deeper more developmental relationship that typically exists with mentoring. The Division leadership identified support for scholarly activity as a high priority and mentorship as an unmet need based on faculty feedback. Clinical revenue supported the position.

Necessary qualifications, determined prior to hiring, included a PhD in health services or related field (eg, epidemiology) or a master’s degree with five years of experience in project management, clinical research, and study design. The position also called for expertise in articulating research questions, selecting study designs, navigating the IRB approval process, collecting/managing data, analyzing statistics, and mentoring and teaching clinical faculty in their scholarly endeavors. A track record in generating academic output (manuscripts and abstracts at regional/national meetings) was required. We circulated a description of the position to Division faculty and to leadership in our School of Public Health.

Based on these criteria, an inaugural coach was hired (author C.M.M.). The coach had a PhD in epidemiology, 10 years of research experience, 16 publications, and had recently finished a National Institutes of Health (NIH) career development award. At the time of hiring, she was a Clinical Assistant Professor in the School of Dentistry, which provided additional FTE. She had no extramural funding but was applying for NIH-level grants and had received several small grants.

To ensure uptake of the coach’s services, we realized that it was necessary to delineate the scope of services available, clarify availability of the coach, and define expectations regarding authorship. We used an iterative process that took into consideration the coach’s expertise, services most needed by the Division’s clinicians, and discussions with Division leadership and faculty at faculty meetings across hospitals and clinics. A range of services and authorship expectations were defined. Consensus was reached that the coach should be invited to coauthor projects where design, analysis, and/or substantial intellectual content was provided and for which authorship criteria were met.¹² Collegial reviews by the coach of already developed manuscripts and time-limited, low-intensity consultations that did not involve substantial intellectual contributions did not warrant authorship.¹² On this basis, we created and distributed a flyer to publicize these guidelines and invite Division members to contact the coach (Figure 1).

The coach attended Division, section, and clinical group meetings to publicize the initiative. The coach also individually met with faculty throughout the Division, explained her role, described services available, and answered questions. The marketing effort was continuous and calibrated with more or less exposure depending on existing projects and the coach’s availability. In addition, the coach coordinated with the director of the Division’s faculty development program to cohost works-in-progress seminars, identify coach clients to present at these meetings, and provide brief presentations on a basic research skill at meetings. Faculty built rapport with the coach through these activities and became more comfortable reaching out for assistance. Because of the large size of the Division, it was decided to roll out the initiative in a stepwise fashion, starting with hospitalists before expanding to the rest of the Division.

Most faculty contacted the coach by e-mail to request a consultation, at which time the coach requested that they complete a preconsultation handout (Figure 2). Initial coaching appointments lasted one hour and were in-person. Coaching entailed an in-depth analysis of the project plan and advice on how to move the project forward. The coach provided tailored scholarly project advice and expertise in research methods. After initial consultations, she would review grant proposals, IRB applications, manuscripts, case report forms, abstracts, and other products. Her efforts typically focused on improving the methods and scientific and technical writing. Assistance with statistical analysis was provided on a case-by-case basis to maintain broad availability. To address statistically complex questions, the coach had five hours of monthly access to a PhD biostatistician via an on-campus consulting service. Follow-up appointments were encouraged and provided as needed by e-mail, phone, or in-person. The coach conducted regular reach outs to facilitate projects. However, execution of the research was generally the responsibility of the faculty member.

Program Evaluation

To characterize the reach and scope of the program, the coach tracked the number of faculty supported, types of services provided, status of initiated projects, numbers of grants generated, and the dissemination of scholarly products including papers and abstracts. We used these metrics to create summary reports to identify successes and areas for improvement. Monthly meetings between the coach and Division leadership were used to fine-tune the approach.

We surveyed coach clients anonymously to assess their satisfaction with the coach initiative. Using Likert scale questions where 1 = completely disagree and 5 = completely agree, we asked (1) if they would recommend the coach to colleagues, (2) if their work was higher quality because of the coach, (3) if they were overall satisfied with the coach, (4) whether the Division should continue to support the coach, and (5) if the coach’s lack of clinical training negatively affected their experience. This work was considered a quality improvement initiative for which IRB approval was not required.

Over 18 months, the coach supported a 49 Division members including 30 hospitalists and 63 projects. Projects included a wide range of scholarship: medical education research, qualitative research, clinical quality improvement projects, observational studies, and a randomized clinical trial. Many clients (n = 16) used the coach for more than one project. The scope of work included limited support projects (identifying research resource and brainstorming project feasibility) lasting one to two sessions (n = 25), projects with a limited scope (collegial reviews of manuscripts and assistance with IRB submissions) but requiring more than two consultations (n = 24), and ongoing in-depth support projects (contributions on design, data collection, analysis, and manuscript writing) that required three consultations or more (n = 14). The majority of Division members (75%) supported did not have master’s level training in a health services-related area, six had NIH or other national-level funding, and two had small grants funded by local sources prior to providing support. The number of Division faculty on a given project ranged from one to four.

The coach directly supported 13 manuscripts with coach authorship, seven manuscripts without authorship, 11 abstracts, and four grant submissions (Appendix). The coach was a coauthor on all the abstracts and a coinvestigator on the grant applications. Of the 13 publications the coach coauthored, 11 publications have been accepted to peer-reviewed journals and two are currently in the submission process. The types of articles published included one medical evaluation report, one qualitative study, one randomized clinical trial, three quality assessment/improvement reports, and five epidemiologic studies. The types of abstracts included one qualitative report, one systematic review, one randomized clinical trial, two quality improvement projects, two epidemiologic studies, and four medical education projects. Three of four small grants submitted to local and national funders were funded.

The coach’s influence extended beyond the Division. Forty-eight university faculty, fellows, or students not affiliated with general internal medicine benefited from coach coaching: 26 were authors on papers and/or abstracts coauthored by the coach, 17 on manuscripts the coach reviewed without authorship, and five participated in consultations.

The coach found the experience rewarding. She enjoyed working on the methodologic aspects of projects and benefited from being included as coauthor on papers.

Twenty-nine of the 43 faculty (67%) still at the institution responded to the program assessment survey. Faculty strongly agreed that they would recommend the coach to colleagues (average ± standard deviation [SD]: 4.7 ± 0.5), that it improved the quality of their work (4.5 ± 0.9), that they were overall satisfied with the coaching (4.6 ± 0.7), and that the Division should continue to support the coach (4.9 ± 0.4). Faculty did not agree that the lack of clinical training of the coach was a barrier (2.0 ± 1.3).

The coach program was highly utilized, well regarded, and delivered substantial, tangible, and academic output. We anticipate the coach initiative will continue to be a valuable resource for our Division and could prove to be a valuable model for other institutions seeking to bolster the scholarly work of clinical academicians.

Several lessons emerged through the course of this project. First, we realized it is essential to select a coach who is both knowledgeable and approachable. We found that after meeting the coach, many faculty sought her help who otherwise would not have. An explicit, ongoing marketing strategy with regular contact with faculty at meetings was a key to receiving consult requests.

Second, the lack of a clinical background did not seem to hinder the coach’s ability to coach clinicians. The coach acknowledged her lack of clinical experience and relied on clients to explain the clinical context of projects. We also learned that the coach’s substantial experience with the logistics of research was invaluable. For example, the coach had substantial experience with the IRB process and her pre-reviews of IRB applications made for a short and relatively seamless experience navigating the IRB process. The coach also facilitated collaborations and leveraged existing resources at our institution. For example, for a qualitative research project, the coach helped identify a health services faculty member with this specific expertise, which led to a successful collaboration and publication. Although a more junior coach with less established qualifications may be helpful with research methods and with the research process, our endeavor suggests that having a more highly trained and experienced researcher was extremely valuable. Finally, we learned that for a Division of our size, the 0.50 FTE allotted to the coach is a minimum requirement. The coach spent approximately four hours a week on marketing, attending faculty meetings and conducting brief didactics, two hours per week on administration, and 14 hours per week on consultations. Faculty generally received support soon after their requests, but there were occasional wait times, which may have delayed some projects.

Academic leaders at our institution have noted the success of our coach initiative and have created a demand for coach services. We are exploring funding models that would allow for the expansion of coach services to other departments and divisions. We are in the initial stages of creating an Academic Scholarship Support Core under the supervision of the coach. Within this Core, we envision that various research support services will be triaged to staff with appropriate expertise; for example, a regulatory coordinator would review IRB applications while a master’s level statistician would conduct statistical analyses.

We have also transitioned to a new coach and have continued to experience success with the program. Our initial coach (author C.M.M.) obtained an NIH R01, a foundation grant, and took over a summer program that trains dental faculty in clinical research methods leaving insufficient time for coaching. Our new coach also has a PhD in epidemiology with NIH R01 funding but has more available FTE. Both of our coaches are graduates of our School of Public Health and institutions with such schools may have good access to the expertise needed. Nonclinical PhDs are often almost entirely reliant on grants, and some nongrant support is often attractive to these researchers. Additionally, PhDs who are junior or mid-career faculty that have the needed training are relatively affordable, particularly when the resource is made available to large number of faculty. For example, our first coach cost $48,000 a year for 50% FTE.

A limitation to our assessment of the coach initiative was the lack of pre- and postintervention metrics of scholarly productivity. We cannot definitively say that the Division’s scholarly output has increased because of the coach. Nevertheless, we are confident that the coach’s coaching has enhanced the scholarly work of individual clinicians and provided value to the Division as a whole. The coach program has been a success in our Division. Other institutions facing the challenge of supporting the research efforts of academic clinicians may consider this model as a worthy investment.

Disclosures

The authors have nothing to disclose.

Acute kidney injury (AKI) is increasingly common in the hospitalized patient^1,2 with recent adult and pediatric multinational studies reporting AKI rates of 57% and 27%, respectively.^3,4 The development of AKI is associated with significant adverse outcomes including an increased risk of mortality.^5-7 For those that survive, the history of AKI may contribute to a lifetime of impaired health with chronic kidney disease.^8,9 This is particularly concerning for pediatric patients as AKI may impact morbidity for many decades, influence available therapies for these morbidities, and ultimately contribute to a shortened lifespan.¹⁰

AKI in the hospitalized patient is no longer accepted as an unfortunate and unavoidable consequence of illness or the indicated therapy. Currently, there is strong interest in this hospital-acquired condition with global initiatives aimed at increased prevention and early detection and treatment of AKI.^11,12 To this objective, risk stratification tools or prediction models could assist clinicians in decision making. Numerous studies have tested AKI prediction models either in particular high-risk populations or based on associated comorbidities, biomarkers, and critical illness scores. These studies are predominantly in adult populations, and few have been externally validated.¹³ While associations between certain medications and AKI are well known, an AKI prediction model that is applicable to pediatric or adult populations and is based on medication exposure is difficult. However, there is a growing recognition of the potential to develop such a model using the electronic health record (EHR).¹⁴

In 2013, Seattle Children’s Hospital (SCH) implemented a nephrotoxin and AKI detection system to assist in clinical decision making within the EHR. This system instituted the automatic ordering of serum creatinines to screen for AKI when the provider ordered three or more medications that were suspected to be nephrotoxic. Other clinical factors such as the diagnoses or preexisting conditions were not considered in the decision-tool algorithm. This original algorithm (Algorithm 1) was later modified and the list of suspected nephrotoxins was expanded (Table 1) in order to align with a national pediatric AKI collaborative (Algorithm 2). However, it was unclear whether the algorithm modification would improve AKI detection.

Context

SCH is a tertiary care academic teaching hospital affiliated with the University of Washington School of Medicine, Harborview Medical Center, and the Seattle Cancer Care Alliance. The hospital has 371 licensed beds and approximately 18 medical subspecialty services.

Study Population

This was a retrospective cohort study examining all patients ages 0-21 years admitted to SCH between December 1, 2013 and November 30, 2015. The detection system was modified to align with the national pediatric AKI collaborative, Nephrotoxic Injury Negated by Just-in-Time Action (NINJA) in November 2014. Both acute care and intensive care patients were included (data not separated by location). Patients who had end-stage kidney disease and were receiving dialysis and patients who were evaluated in the emergency department without being admitted or admitted as observation status were excluded from analysis. Patients were also excluded if they did not have a baseline serum creatinine as defined below.

Study Measures

AKI is defined at SCH using the Kidney Disease: Improving Global Outcomes Stage 1 criteria as a guideline. The diagnosis of AKI is based on an increase in the baseline serum creatinine by 0.3 mg/dL or an increase in the serum creatinine by >1.5 times the baseline assuming the incoming creatinine is 0.5 mg/dL or higher. For our definition, the increase in serum creatinine needs to have occurred within a one-week timeframe and urine output is not a diagnostic criterion.¹⁵ Baseline serum creatinine is defined as the lowest serum creatinine in the previous six months. Forty medications were classified as nephrotoxins based on previous analysis¹⁶ and adapted for our institutional formulary.

Statistical Analysis

To evaluate the efficacy of our systems in detecting nephrotoxin-related AKI, the sensitivity and the specificity using both our original algorithm (Algorithm 1) and the modified algorithm (Algorithm 2) were generated on our complete data set. To test sensitivity, the proportion of AKI patients who would trigger alert using Algorithm 1 and then with Algorithm 2 was identified. Similarly, to test specificity, the proportion of non-AKI patients who did not trigger an alert by the surveillance systems was identified. The differences in sensitivity and specificity between the two algorithms were evaluated using two-sample tests of proportion.

The statistical method of Combinatorial Inference has been utilized in studies of cancer biology¹⁷ and in genomics.¹⁸ A variation of this approach was used in this study to identify the specific medication combinations most associated with AKI. First, all of the nephrotoxic medications and medication combinations that were prescribed during our study period were identified from a data set (ie, a training set) containing 75% of all encounters selected at random without replacement. Using this training set, the prevalence of each medication combination and the rate of AKI associated with each combination were identified. The predicted overall AKI risk of an individual medication is the average of all the AKI rates associated with each combination containing that specific medication. Also incorporated into the determination of the predicted AKI risk was the prevalence of that medication combination.

To test our model’s predictive capability, the algorithm was applied to the remaining 25% of the total patient data (ie, the test set). The predicted AKI risk was compared with the actual AKI rate in the test data set. Our model’s predictive capability was represented in a receiver operator characteristic (ROC) analysis. The goal was to achieve an area under the ROC curve (AUC) approaching one as this would reflect 100% sensitivity and 100% specificity, whereas an AUC of 0.5 would represent a random guess (50% chance of being correct).

Lastly, our final step was to use our model’s ROC curve to determine an optimal threshold of AKI risk for which to trigger an alert. This predicted risk threshold was based on our goal to increase our surveillance system’s sensitivity balanced with maintaining an acceptable specificity.

An a priori threshold of P = .05 was used to determine statistical significance of all results. Analyses were conducted in Stata 12.1 (StataCorp LP, College Station, Texas) and R 3.3.2 (R Foundation for Statistical Computing, Vienna, Austria). A sample data set containing replication code for our model can be found in an online repository (https://dataverse.harvard.edu/dataverse/chuan). This study was approved by the Seattle Children’s Institutional Review Board.

Sensitivity and Specificity

Of the patient encounters, 14,779 were eligible during the study period. The sensitivity of the system’s ability to identify nephrotoxin-related AKI decreased from 46.9% using Algorithm 1 to 43.3% using Algorithm 2, a change of 3.6% (P = .22). The specificity increased from 73.6% to 89.3%, a change of 15.7% (P < .001; Table 2).

Improvement of Our Nephrotoxin-Related AKI Detection System Using a Novel AKI Prediction Strategy

A total of 838 medication combinations were identified in our training set and the predicted AKI risk for every medication combination was determined. By comparing the predicted risk of AKI to the actual AKI occurrence, an ROC curve with an AUC of 0.756 (Figure) was generated. An increase in system sensitivity was prioritized when determining the optimal AKI risk at which the model would trigger an alert. Setting an alert threshold at a predicted AKI risk of >8%, our model performed with a sensitivity of 74% while decreasing the specificity to 70%.

Identification of High-Risk Nephrotoxic Medications and Medication Combinations

Approximately 200 medication combinations were associated with >8% AKI risk, our new AKI prediction model’s alert threshold. Medication combinations consisting of up to 11 concomitantly prescribed medications were present in our data set. However, many of these combinations were infrequently prescribed. Further analysis, conducted in order to increase the clinical relevance of our findings, identified 10 medications or medication combinations that were both associated with a predicted AKI risk of >8% and that were prescribed on average greater than twice a month (Table 3).

The nephrotoxin-related AKI detection system at SCH automatically places orders for serum creatinines on patients who have met criteria for concomitant nephrotoxin exposure. This has given us a robust database from which to develop our clinical decision-making tool. Both our original and updated systems were based on the absolute number of concomitant nephrotoxic medications prescribed.¹⁶ This is a reasonable approach given the complexity of building a surveillance system¹⁹ and resource limitations. However, a system based on observed rather than theoretical or in vitro data, adaptable to the institution and designed for ongoing refinement, would be more valuable.

The interest in AKI prediction tools continues to be high. Bedford et al. employed numerous variables and diagnostic codes to predict the development of AKI in adults during hospitalization. They were able to produce a prediction model with a reasonable fit (AUC 0.72) to identify patients at higher risk for AKI but were less successful in their attempts to predict progression to severe AKI.²⁰ Hodgson et al. recently developed an adult AKI prediction score (AUC 0.65-0.72) also based on numerous clinical factors that was able to positively impact inpatient mortality.²¹ To our knowledge, our model is unique in that it focuses on nephrotoxins using a predicted AKI risk algorithm based on observed AKI rates of previously ordered medications/medication combinations (two to 11 medications). Having a decision tool targeting medications gives the clinician guidance that can be used to make a specific intervention rather than identifying a patient at risk due to a diagnosis code or other difficult to modify factors.

There are abundant case studies and reports using logistic regression models identifying specific medications associated with AKI. Our choice of methodology was based on our assessment that logistic regression models would be inadequate for the development of a real-time clinical decision-making tool for several reasons. Using logistic regression to explore every medication combination based on our medication list would be challenging as there are approximately 5.5 × 10¹⁰ potential medication combinations. Additionally, logistic regression ignores any potential interactions between the medications. This is an important point as medication interactions can be synergistic, neutral, or antagonist. Consequently, the outcome generated from a set of combined variables may be different from one generated from the sum of each variable taken independently. Logistic regression also does not account for the potential prescribing trends among providers as it assumes that all medications or medication combinations are equally available at the same time. However, in practice, depending on numerous factors, such as hospital culture (eg, the presence of clinical standard work pathways), local bacterial resistance patterns, or medication shortages; certain medication combinations may occur more frequently while others not at all. Finally, logistic regression cannot account for the possibility of a medication combination occurring; therefore, logistic regression may identify a combination strongly associated with AKI that is rarely prescribed.

We theorized that AKI detection would improve with the Algorithm 2 modifications, including the expanded nephrotoxin list, which accompanied alignment with the national pediatric AKI collaborative, NINJA. The finding that our surveillance sensitivity did not improve with this system update supported our subsequent objective to develop a novel nephrotoxin-related AKI decision tool or detection system using our EHR data to identify which specific medications and/or medication combinations were associated with a higher rate of AKI. However, it should be noted that two factors related to measurement bias introduce limitations to our sensitivity and specificity analyses. First, regarding the presence of the alert system, our system will order serum creatinines on patients when they have been exposed to nephrotoxins. Consequently, the proportion of patients with creatinines measured will increase in the nephrotoxin-exposed patients. Unexposed patients may have AKI that is not detected because creatinines may not be ordered. Therefore, there is the potential for a relative increase in AKI detection among nephrotoxin-exposed patients as compared with unexposed patients, which would then affect the measured sensitivity and specificity of the alert. Second, the automated alerts require a baseline creatinine in order to trigger therefore are unable to identify AKI among patients who do not have a baseline serum creatinine measurement.

Our new nephrotoxin-related AKI detection model performed best when an alert was triggered for those medications or medication combinations with a predicted AKI risk of >8%. Forty-six medication combinations consisting of exactly two medications were determined to have a predicted AKI risk of >8% therefore would trigger an alert in our new model system. These medication combinations would not have triggered an alert using either of the previous system algorithms as both algorithms are based on the presence of three or more concomitant nephrotoxic medications.

From the list of suspected nephrotoxins, we identified 11 unique medications in 10 different combinations with a predicted AKI risk of >8% that were prescribed frequently (at least twice a month on average; Table 3). Notably, six out of 10 medication combinations involved vancomycin. Piperacillin-tazobactam was also represented in several combinations. These findings support the concern that others have reported regarding these two medications particularly when prescribed together.^22,23

Acknowledgments

Dr. Karyn E. Yonekawa conceptualized and designed the study, directed the data analysis, interpreted the data, drafted, revised and gave final approval of the manuscript. Dr. Chuan Zhou contributed to the study design, acquired data, conducted the data analysis, critically reviewed, and gave final approval of the manuscript. Ms. Wren L. Haaland contributed to the study design, acquired data, conducted the data analysis, critically reviewed, and gave final approval of the manuscript. Dr. Davene R. Wright contributed to the study design, data analysis, critically reviewed, revised, and gave final approval of the manuscript.

The authors would like to thank Holly Clifton and Suzanne Spencer for their assistance with data acquisition and Drs. Derya Caglar, Corrie McDaniel, and Thida Ong for their writing support.

All authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

Disclosures

The authors have no conflicts of interest to report.

Acute kidney injury (AKI) is increasingly common in the hospitalized patient^1,2 with recent adult and pediatric multinational studies reporting AKI rates of 57% and 27%, respectively.^3,4 The development of AKI is associated with significant adverse outcomes including an increased risk of mortality.^5-7 For those that survive, the history of AKI may contribute to a lifetime of impaired health with chronic kidney disease.^8,9 This is particularly concerning for pediatric patients as AKI may impact morbidity for many decades, influence available therapies for these morbidities, and ultimately contribute to a shortened lifespan.¹⁰

AKI in the hospitalized patient is no longer accepted as an unfortunate and unavoidable consequence of illness or the indicated therapy. Currently, there is strong interest in this hospital-acquired condition with global initiatives aimed at increased prevention and early detection and treatment of AKI.^11,12 To this objective, risk stratification tools or prediction models could assist clinicians in decision making. Numerous studies have tested AKI prediction models either in particular high-risk populations or based on associated comorbidities, biomarkers, and critical illness scores. These studies are predominantly in adult populations, and few have been externally validated.¹³ While associations between certain medications and AKI are well known, an AKI prediction model that is applicable to pediatric or adult populations and is based on medication exposure is difficult. However, there is a growing recognition of the potential to develop such a model using the electronic health record (EHR).¹⁴

In 2013, Seattle Children’s Hospital (SCH) implemented a nephrotoxin and AKI detection system to assist in clinical decision making within the EHR. This system instituted the automatic ordering of serum creatinines to screen for AKI when the provider ordered three or more medications that were suspected to be nephrotoxic. Other clinical factors such as the diagnoses or preexisting conditions were not considered in the decision-tool algorithm. This original algorithm (Algorithm 1) was later modified and the list of suspected nephrotoxins was expanded (Table 1) in order to align with a national pediatric AKI collaborative (Algorithm 2). However, it was unclear whether the algorithm modification would improve AKI detection.

Context

SCH is a tertiary care academic teaching hospital affiliated with the University of Washington School of Medicine, Harborview Medical Center, and the Seattle Cancer Care Alliance. The hospital has 371 licensed beds and approximately 18 medical subspecialty services.

Study Population

This was a retrospective cohort study examining all patients ages 0-21 years admitted to SCH between December 1, 2013 and November 30, 2015. The detection system was modified to align with the national pediatric AKI collaborative, Nephrotoxic Injury Negated by Just-in-Time Action (NINJA) in November 2014. Both acute care and intensive care patients were included (data not separated by location). Patients who had end-stage kidney disease and were receiving dialysis and patients who were evaluated in the emergency department without being admitted or admitted as observation status were excluded from analysis. Patients were also excluded if they did not have a baseline serum creatinine as defined below.

Study Measures

AKI is defined at SCH using the Kidney Disease: Improving Global Outcomes Stage 1 criteria as a guideline. The diagnosis of AKI is based on an increase in the baseline serum creatinine by 0.3 mg/dL or an increase in the serum creatinine by >1.5 times the baseline assuming the incoming creatinine is 0.5 mg/dL or higher. For our definition, the increase in serum creatinine needs to have occurred within a one-week timeframe and urine output is not a diagnostic criterion.¹⁵ Baseline serum creatinine is defined as the lowest serum creatinine in the previous six months. Forty medications were classified as nephrotoxins based on previous analysis¹⁶ and adapted for our institutional formulary.

Statistical Analysis

To evaluate the efficacy of our systems in detecting nephrotoxin-related AKI, the sensitivity and the specificity using both our original algorithm (Algorithm 1) and the modified algorithm (Algorithm 2) were generated on our complete data set. To test sensitivity, the proportion of AKI patients who would trigger alert using Algorithm 1 and then with Algorithm 2 was identified. Similarly, to test specificity, the proportion of non-AKI patients who did not trigger an alert by the surveillance systems was identified. The differences in sensitivity and specificity between the two algorithms were evaluated using two-sample tests of proportion.

The statistical method of Combinatorial Inference has been utilized in studies of cancer biology¹⁷ and in genomics.¹⁸ A variation of this approach was used in this study to identify the specific medication combinations most associated with AKI. First, all of the nephrotoxic medications and medication combinations that were prescribed during our study period were identified from a data set (ie, a training set) containing 75% of all encounters selected at random without replacement. Using this training set, the prevalence of each medication combination and the rate of AKI associated with each combination were identified. The predicted overall AKI risk of an individual medication is the average of all the AKI rates associated with each combination containing that specific medication. Also incorporated into the determination of the predicted AKI risk was the prevalence of that medication combination.

To test our model’s predictive capability, the algorithm was applied to the remaining 25% of the total patient data (ie, the test set). The predicted AKI risk was compared with the actual AKI rate in the test data set. Our model’s predictive capability was represented in a receiver operator characteristic (ROC) analysis. The goal was to achieve an area under the ROC curve (AUC) approaching one as this would reflect 100% sensitivity and 100% specificity, whereas an AUC of 0.5 would represent a random guess (50% chance of being correct).

Lastly, our final step was to use our model’s ROC curve to determine an optimal threshold of AKI risk for which to trigger an alert. This predicted risk threshold was based on our goal to increase our surveillance system’s sensitivity balanced with maintaining an acceptable specificity.

An a priori threshold of P = .05 was used to determine statistical significance of all results. Analyses were conducted in Stata 12.1 (StataCorp LP, College Station, Texas) and R 3.3.2 (R Foundation for Statistical Computing, Vienna, Austria). A sample data set containing replication code for our model can be found in an online repository (https://dataverse.harvard.edu/dataverse/chuan). This study was approved by the Seattle Children’s Institutional Review Board.

Sensitivity and Specificity

Of the patient encounters, 14,779 were eligible during the study period. The sensitivity of the system’s ability to identify nephrotoxin-related AKI decreased from 46.9% using Algorithm 1 to 43.3% using Algorithm 2, a change of 3.6% (P = .22). The specificity increased from 73.6% to 89.3%, a change of 15.7% (P < .001; Table 2).

Improvement of Our Nephrotoxin-Related AKI Detection System Using a Novel AKI Prediction Strategy

A total of 838 medication combinations were identified in our training set and the predicted AKI risk for every medication combination was determined. By comparing the predicted risk of AKI to the actual AKI occurrence, an ROC curve with an AUC of 0.756 (Figure) was generated. An increase in system sensitivity was prioritized when determining the optimal AKI risk at which the model would trigger an alert. Setting an alert threshold at a predicted AKI risk of >8%, our model performed with a sensitivity of 74% while decreasing the specificity to 70%.

Identification of High-Risk Nephrotoxic Medications and Medication Combinations

Approximately 200 medication combinations were associated with >8% AKI risk, our new AKI prediction model’s alert threshold. Medication combinations consisting of up to 11 concomitantly prescribed medications were present in our data set. However, many of these combinations were infrequently prescribed. Further analysis, conducted in order to increase the clinical relevance of our findings, identified 10 medications or medication combinations that were both associated with a predicted AKI risk of >8% and that were prescribed on average greater than twice a month (Table 3).

The nephrotoxin-related AKI detection system at SCH automatically places orders for serum creatinines on patients who have met criteria for concomitant nephrotoxin exposure. This has given us a robust database from which to develop our clinical decision-making tool. Both our original and updated systems were based on the absolute number of concomitant nephrotoxic medications prescribed.¹⁶ This is a reasonable approach given the complexity of building a surveillance system¹⁹ and resource limitations. However, a system based on observed rather than theoretical or in vitro data, adaptable to the institution and designed for ongoing refinement, would be more valuable.

The interest in AKI prediction tools continues to be high. Bedford et al. employed numerous variables and diagnostic codes to predict the development of AKI in adults during hospitalization. They were able to produce a prediction model with a reasonable fit (AUC 0.72) to identify patients at higher risk for AKI but were less successful in their attempts to predict progression to severe AKI.²⁰ Hodgson et al. recently developed an adult AKI prediction score (AUC 0.65-0.72) also based on numerous clinical factors that was able to positively impact inpatient mortality.²¹ To our knowledge, our model is unique in that it focuses on nephrotoxins using a predicted AKI risk algorithm based on observed AKI rates of previously ordered medications/medication combinations (two to 11 medications). Having a decision tool targeting medications gives the clinician guidance that can be used to make a specific intervention rather than identifying a patient at risk due to a diagnosis code or other difficult to modify factors.

There are abundant case studies and reports using logistic regression models identifying specific medications associated with AKI. Our choice of methodology was based on our assessment that logistic regression models would be inadequate for the development of a real-time clinical decision-making tool for several reasons. Using logistic regression to explore every medication combination based on our medication list would be challenging as there are approximately 5.5 × 10¹⁰ potential medication combinations. Additionally, logistic regression ignores any potential interactions between the medications. This is an important point as medication interactions can be synergistic, neutral, or antagonist. Consequently, the outcome generated from a set of combined variables may be different from one generated from the sum of each variable taken independently. Logistic regression also does not account for the potential prescribing trends among providers as it assumes that all medications or medication combinations are equally available at the same time. However, in practice, depending on numerous factors, such as hospital culture (eg, the presence of clinical standard work pathways), local bacterial resistance patterns, or medication shortages; certain medication combinations may occur more frequently while others not at all. Finally, logistic regression cannot account for the possibility of a medication combination occurring; therefore, logistic regression may identify a combination strongly associated with AKI that is rarely prescribed.

We theorized that AKI detection would improve with the Algorithm 2 modifications, including the expanded nephrotoxin list, which accompanied alignment with the national pediatric AKI collaborative, NINJA. The finding that our surveillance sensitivity did not improve with this system update supported our subsequent objective to develop a novel nephrotoxin-related AKI decision tool or detection system using our EHR data to identify which specific medications and/or medication combinations were associated with a higher rate of AKI. However, it should be noted that two factors related to measurement bias introduce limitations to our sensitivity and specificity analyses. First, regarding the presence of the alert system, our system will order serum creatinines on patients when they have been exposed to nephrotoxins. Consequently, the proportion of patients with creatinines measured will increase in the nephrotoxin-exposed patients. Unexposed patients may have AKI that is not detected because creatinines may not be ordered. Therefore, there is the potential for a relative increase in AKI detection among nephrotoxin-exposed patients as compared with unexposed patients, which would then affect the measured sensitivity and specificity of the alert. Second, the automated alerts require a baseline creatinine in order to trigger therefore are unable to identify AKI among patients who do not have a baseline serum creatinine measurement.

Our new nephrotoxin-related AKI detection model performed best when an alert was triggered for those medications or medication combinations with a predicted AKI risk of >8%. Forty-six medication combinations consisting of exactly two medications were determined to have a predicted AKI risk of >8% therefore would trigger an alert in our new model system. These medication combinations would not have triggered an alert using either of the previous system algorithms as both algorithms are based on the presence of three or more concomitant nephrotoxic medications.

From the list of suspected nephrotoxins, we identified 11 unique medications in 10 different combinations with a predicted AKI risk of >8% that were prescribed frequently (at least twice a month on average; Table 3). Notably, six out of 10 medication combinations involved vancomycin. Piperacillin-tazobactam was also represented in several combinations. These findings support the concern that others have reported regarding these two medications particularly when prescribed together.^22,23

Acknowledgments

Dr. Karyn E. Yonekawa conceptualized and designed the study, directed the data analysis, interpreted the data, drafted, revised and gave final approval of the manuscript. Dr. Chuan Zhou contributed to the study design, acquired data, conducted the data analysis, critically reviewed, and gave final approval of the manuscript. Ms. Wren L. Haaland contributed to the study design, acquired data, conducted the data analysis, critically reviewed, and gave final approval of the manuscript. Dr. Davene R. Wright contributed to the study design, data analysis, critically reviewed, revised, and gave final approval of the manuscript.

The authors would like to thank Holly Clifton and Suzanne Spencer for their assistance with data acquisition and Drs. Derya Caglar, Corrie McDaniel, and Thida Ong for their writing support.

All authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

Disclosures

The authors have no conflicts of interest to report.

Acute kidney injury (AKI) is increasingly common in the hospitalized patient^1,2 with recent adult and pediatric multinational studies reporting AKI rates of 57% and 27%, respectively.^3,4 The development of AKI is associated with significant adverse outcomes including an increased risk of mortality.^5-7 For those that survive, the history of AKI may contribute to a lifetime of impaired health with chronic kidney disease.^8,9 This is particularly concerning for pediatric patients as AKI may impact morbidity for many decades, influence available therapies for these morbidities, and ultimately contribute to a shortened lifespan.¹⁰

AKI in the hospitalized patient is no longer accepted as an unfortunate and unavoidable consequence of illness or the indicated therapy. Currently, there is strong interest in this hospital-acquired condition with global initiatives aimed at increased prevention and early detection and treatment of AKI.^11,12 To this objective, risk stratification tools or prediction models could assist clinicians in decision making. Numerous studies have tested AKI prediction models either in particular high-risk populations or based on associated comorbidities, biomarkers, and critical illness scores. These studies are predominantly in adult populations, and few have been externally validated.¹³ While associations between certain medications and AKI are well known, an AKI prediction model that is applicable to pediatric or adult populations and is based on medication exposure is difficult. However, there is a growing recognition of the potential to develop such a model using the electronic health record (EHR).¹⁴

In 2013, Seattle Children’s Hospital (SCH) implemented a nephrotoxin and AKI detection system to assist in clinical decision making within the EHR. This system instituted the automatic ordering of serum creatinines to screen for AKI when the provider ordered three or more medications that were suspected to be nephrotoxic. Other clinical factors such as the diagnoses or preexisting conditions were not considered in the decision-tool algorithm. This original algorithm (Algorithm 1) was later modified and the list of suspected nephrotoxins was expanded (Table 1) in order to align with a national pediatric AKI collaborative (Algorithm 2). However, it was unclear whether the algorithm modification would improve AKI detection.

Context

SCH is a tertiary care academic teaching hospital affiliated with the University of Washington School of Medicine, Harborview Medical Center, and the Seattle Cancer Care Alliance. The hospital has 371 licensed beds and approximately 18 medical subspecialty services.

Study Population

This was a retrospective cohort study examining all patients ages 0-21 years admitted to SCH between December 1, 2013 and November 30, 2015. The detection system was modified to align with the national pediatric AKI collaborative, Nephrotoxic Injury Negated by Just-in-Time Action (NINJA) in November 2014. Both acute care and intensive care patients were included (data not separated by location). Patients who had end-stage kidney disease and were receiving dialysis and patients who were evaluated in the emergency department without being admitted or admitted as observation status were excluded from analysis. Patients were also excluded if they did not have a baseline serum creatinine as defined below.

Study Measures

AKI is defined at SCH using the Kidney Disease: Improving Global Outcomes Stage 1 criteria as a guideline. The diagnosis of AKI is based on an increase in the baseline serum creatinine by 0.3 mg/dL or an increase in the serum creatinine by >1.5 times the baseline assuming the incoming creatinine is 0.5 mg/dL or higher. For our definition, the increase in serum creatinine needs to have occurred within a one-week timeframe and urine output is not a diagnostic criterion.¹⁵ Baseline serum creatinine is defined as the lowest serum creatinine in the previous six months. Forty medications were classified as nephrotoxins based on previous analysis¹⁶ and adapted for our institutional formulary.

Statistical Analysis

To evaluate the efficacy of our systems in detecting nephrotoxin-related AKI, the sensitivity and the specificity using both our original algorithm (Algorithm 1) and the modified algorithm (Algorithm 2) were generated on our complete data set. To test sensitivity, the proportion of AKI patients who would trigger alert using Algorithm 1 and then with Algorithm 2 was identified. Similarly, to test specificity, the proportion of non-AKI patients who did not trigger an alert by the surveillance systems was identified. The differences in sensitivity and specificity between the two algorithms were evaluated using two-sample tests of proportion.

The statistical method of Combinatorial Inference has been utilized in studies of cancer biology¹⁷ and in genomics.¹⁸ A variation of this approach was used in this study to identify the specific medication combinations most associated with AKI. First, all of the nephrotoxic medications and medication combinations that were prescribed during our study period were identified from a data set (ie, a training set) containing 75% of all encounters selected at random without replacement. Using this training set, the prevalence of each medication combination and the rate of AKI associated with each combination were identified. The predicted overall AKI risk of an individual medication is the average of all the AKI rates associated with each combination containing that specific medication. Also incorporated into the determination of the predicted AKI risk was the prevalence of that medication combination.

To test our model’s predictive capability, the algorithm was applied to the remaining 25% of the total patient data (ie, the test set). The predicted AKI risk was compared with the actual AKI rate in the test data set. Our model’s predictive capability was represented in a receiver operator characteristic (ROC) analysis. The goal was to achieve an area under the ROC curve (AUC) approaching one as this would reflect 100% sensitivity and 100% specificity, whereas an AUC of 0.5 would represent a random guess (50% chance of being correct).

Lastly, our final step was to use our model’s ROC curve to determine an optimal threshold of AKI risk for which to trigger an alert. This predicted risk threshold was based on our goal to increase our surveillance system’s sensitivity balanced with maintaining an acceptable specificity.

An a priori threshold of P = .05 was used to determine statistical significance of all results. Analyses were conducted in Stata 12.1 (StataCorp LP, College Station, Texas) and R 3.3.2 (R Foundation for Statistical Computing, Vienna, Austria). A sample data set containing replication code for our model can be found in an online repository (https://dataverse.harvard.edu/dataverse/chuan). This study was approved by the Seattle Children’s Institutional Review Board.

Sensitivity and Specificity

Of the patient encounters, 14,779 were eligible during the study period. The sensitivity of the system’s ability to identify nephrotoxin-related AKI decreased from 46.9% using Algorithm 1 to 43.3% using Algorithm 2, a change of 3.6% (P = .22). The specificity increased from 73.6% to 89.3%, a change of 15.7% (P < .001; Table 2).

Improvement of Our Nephrotoxin-Related AKI Detection System Using a Novel AKI Prediction Strategy

A total of 838 medication combinations were identified in our training set and the predicted AKI risk for every medication combination was determined. By comparing the predicted risk of AKI to the actual AKI occurrence, an ROC curve with an AUC of 0.756 (Figure) was generated. An increase in system sensitivity was prioritized when determining the optimal AKI risk at which the model would trigger an alert. Setting an alert threshold at a predicted AKI risk of >8%, our model performed with a sensitivity of 74% while decreasing the specificity to 70%.

Identification of High-Risk Nephrotoxic Medications and Medication Combinations

Approximately 200 medication combinations were associated with >8% AKI risk, our new AKI prediction model’s alert threshold. Medication combinations consisting of up to 11 concomitantly prescribed medications were present in our data set. However, many of these combinations were infrequently prescribed. Further analysis, conducted in order to increase the clinical relevance of our findings, identified 10 medications or medication combinations that were both associated with a predicted AKI risk of >8% and that were prescribed on average greater than twice a month (Table 3).

The nephrotoxin-related AKI detection system at SCH automatically places orders for serum creatinines on patients who have met criteria for concomitant nephrotoxin exposure. This has given us a robust database from which to develop our clinical decision-making tool. Both our original and updated systems were based on the absolute number of concomitant nephrotoxic medications prescribed.¹⁶ This is a reasonable approach given the complexity of building a surveillance system¹⁹ and resource limitations. However, a system based on observed rather than theoretical or in vitro data, adaptable to the institution and designed for ongoing refinement, would be more valuable.

The interest in AKI prediction tools continues to be high. Bedford et al. employed numerous variables and diagnostic codes to predict the development of AKI in adults during hospitalization. They were able to produce a prediction model with a reasonable fit (AUC 0.72) to identify patients at higher risk for AKI but were less successful in their attempts to predict progression to severe AKI.²⁰ Hodgson et al. recently developed an adult AKI prediction score (AUC 0.65-0.72) also based on numerous clinical factors that was able to positively impact inpatient mortality.²¹ To our knowledge, our model is unique in that it focuses on nephrotoxins using a predicted AKI risk algorithm based on observed AKI rates of previously ordered medications/medication combinations (two to 11 medications). Having a decision tool targeting medications gives the clinician guidance that can be used to make a specific intervention rather than identifying a patient at risk due to a diagnosis code or other difficult to modify factors.

There are abundant case studies and reports using logistic regression models identifying specific medications associated with AKI. Our choice of methodology was based on our assessment that logistic regression models would be inadequate for the development of a real-time clinical decision-making tool for several reasons. Using logistic regression to explore every medication combination based on our medication list would be challenging as there are approximately 5.5 × 10¹⁰ potential medication combinations. Additionally, logistic regression ignores any potential interactions between the medications. This is an important point as medication interactions can be synergistic, neutral, or antagonist. Consequently, the outcome generated from a set of combined variables may be different from one generated from the sum of each variable taken independently. Logistic regression also does not account for the potential prescribing trends among providers as it assumes that all medications or medication combinations are equally available at the same time. However, in practice, depending on numerous factors, such as hospital culture (eg, the presence of clinical standard work pathways), local bacterial resistance patterns, or medication shortages; certain medication combinations may occur more frequently while others not at all. Finally, logistic regression cannot account for the possibility of a medication combination occurring; therefore, logistic regression may identify a combination strongly associated with AKI that is rarely prescribed.

We theorized that AKI detection would improve with the Algorithm 2 modifications, including the expanded nephrotoxin list, which accompanied alignment with the national pediatric AKI collaborative, NINJA. The finding that our surveillance sensitivity did not improve with this system update supported our subsequent objective to develop a novel nephrotoxin-related AKI decision tool or detection system using our EHR data to identify which specific medications and/or medication combinations were associated with a higher rate of AKI. However, it should be noted that two factors related to measurement bias introduce limitations to our sensitivity and specificity analyses. First, regarding the presence of the alert system, our system will order serum creatinines on patients when they have been exposed to nephrotoxins. Consequently, the proportion of patients with creatinines measured will increase in the nephrotoxin-exposed patients. Unexposed patients may have AKI that is not detected because creatinines may not be ordered. Therefore, there is the potential for a relative increase in AKI detection among nephrotoxin-exposed patients as compared with unexposed patients, which would then affect the measured sensitivity and specificity of the alert. Second, the automated alerts require a baseline creatinine in order to trigger therefore are unable to identify AKI among patients who do not have a baseline serum creatinine measurement.

Our new nephrotoxin-related AKI detection model performed best when an alert was triggered for those medications or medication combinations with a predicted AKI risk of >8%. Forty-six medication combinations consisting of exactly two medications were determined to have a predicted AKI risk of >8% therefore would trigger an alert in our new model system. These medication combinations would not have triggered an alert using either of the previous system algorithms as both algorithms are based on the presence of three or more concomitant nephrotoxic medications.

From the list of suspected nephrotoxins, we identified 11 unique medications in 10 different combinations with a predicted AKI risk of >8% that were prescribed frequently (at least twice a month on average; Table 3). Notably, six out of 10 medication combinations involved vancomycin. Piperacillin-tazobactam was also represented in several combinations. These findings support the concern that others have reported regarding these two medications particularly when prescribed together.^22,23

Acknowledgments

Dr. Karyn E. Yonekawa conceptualized and designed the study, directed the data analysis, interpreted the data, drafted, revised and gave final approval of the manuscript. Dr. Chuan Zhou contributed to the study design, acquired data, conducted the data analysis, critically reviewed, and gave final approval of the manuscript. Ms. Wren L. Haaland contributed to the study design, acquired data, conducted the data analysis, critically reviewed, and gave final approval of the manuscript. Dr. Davene R. Wright contributed to the study design, data analysis, critically reviewed, revised, and gave final approval of the manuscript.

The authors would like to thank Holly Clifton and Suzanne Spencer for their assistance with data acquisition and Drs. Derya Caglar, Corrie McDaniel, and Thida Ong for their writing support.

All authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

Disclosures

The authors have no conflicts of interest to report.

In this issue of the Journal of Hospital Medicine^®, Drs. Arora and Moriates highlight an important deficiency in quality improvement efforts designed to reduce overuse of tests and treatments: the potential for trainees—and by extension, more seasoned clinicians—to rationalize minimizing under the guise of high-value care.¹ This insightful perspective from the Co-Directors of Costs of Care should serve as a catalyst for further robust and effective care redesign efforts to optimize the use of all medical resources, including tests, treatments, procedures, consultations, emergency department (ED) visits, and hospital admissions. The formula to root out minimizers is not straightforward and requires an evaluation of wasteful practices in a nuanced and holistic manner that considers not only the frequency that the overused test (or treatment) is ordered but also the collateral impact of not ordering it. This principle has implications for measuring, paying for, and studying high-value care.

Overuse of tests and treatments increases costs and carries a risk of harm, from unnecessary use of creatine kinase–myocardial band (CK-MB) in suspected acute coronary syndrome² to unwarranted administration of antibiotics for asymptomatic bacteriuria³ to over-administration of blood transfusions.⁴ However, decreasing the use of a commonly ordered test is not always clinically appropriate. To illustrate this point, we consider the evidence-based algorithm to deliver best practice in the work-up of pulmonary embolism (PE) by Raja et al; which integrates pretest probability, PERC assessment, and appropriate use of D-dimer and pulmonary CT angiography (CTA).⁵ Avoiding D-dimer testing is appropriate in patients with very low pretest probability who pass a pulmonary embolism rule-out criteria (PERC) clinical assessment and is also appropriate in patients who have sufficiently high clinical probability for PE to justify CTA regardless of a D-dimer result. On the other hand, avoiding D-dimer testing by attributing a patient’s symptoms to anxiety (as a minimizer might do) would increase patient risk, and could potentially increase cost if that patient ends up in intensive care after delayed diagnosis. Following diagnostic algorithms that include physician decision-making and evidence-based guidelines can prevent overuse and underuse, thereby maximizing efficiency and effectiveness. Engaging trainees in the development of such algorithms and decision support tools will serve to ingrain these principles into their practice.

Arora and Moriates highlight the importance of caring for a patient along a continuum rather than simply optimizing practice with respect to a single management decision or an isolated care episode. This approach is fundamental to the quality of care we provide, the public trust our profession still commands, and the total cost of care (TCOC). The two largest contributors to debilitating patient healthcare debt are not overuse of tests and treatments, but ED visits and hospitalizations.⁶ Thus, high-value quality improvement needs to anticipate future healthcare needs, including those that may result from delayed or missed diagnoses. Furthermore, excessive focus on the minutiae of high-value care (fewer daily basic metabolic panels) can lead to change fatigue and divert attention from higher impact utilization. We endorse a holistic approach in which the lens is shifted from the test—and even from the encounter or episode of care—to the entire continuum of care so that we can safeguard against inappropriate minimization. This approach has started to gain traction with policymakers. For example, the state of Maryland has implemented a TCOC hospital payment model predicted to save $1 billion by 2023.⁷ The TCOC model includes a Care Redesign Program whereby hospitals and nonhospital healthcare providers collaborate to improve the quality of care while reducing spending, and cost savings can be used for incentive payments to the nonhospital providers (gainsharing) while simultaneously monitoring quality measures to guard against rationing.⁷ In keeping with the authors’ call to prioritize overall health, this new reimbursement model and others similar to it aim to incentivize the delivery of high-value care across a continuum.

Research is needed to guide best practice from this global perspective; as such, value improvement projects aimed at optimizing use of tests and treatments should include rigorous methodology, measures of downstream outcomes and costs, and balancing safety measures.⁸ For example, the ROMICAT II randomized trial evaluated two diagnostic approaches in emergency department patients with suspected acute coronary syndrome: early coronary computed tomography angiogram (CCTA) and standard ED evaluation.⁹ In addition to outcomes related to the ED visit itself, downstream testing and outcomes for 28 days after the episode were studied. In the acute setting, CCTA decreased time to diagnosis, reduced mean hospital length of stay by 7.6 hours, and resulted in 47% of patients being discharged within 8.6 hours as opposed to only 12% of the standard evaluation cohort. No cases of ACS were missed, and the CCTA cohort has slightly fewer cardiovascular adverse events (P = .18). However, the CCTA patients received significantly more diagnostic and functional testing and higher radiation exposure than the standard evaluation cohort, and underwent modestly higher rates of coronary angiography and percutaneous coronary intervention. The TCOC over the 28-day period was similar at $4,289 for CCTA versus $4,060 for standard care (P = .65).⁹

Reducing the TCOC is imperative to protect patients from the burden of healthcare debt, but concerns have been raised about the ethics of high-value care if decision-making is driven by cost considerations.¹⁰ A recent viewpoint proposed a framework where high-value care recommendations are categorized as obligatory (protecting patients from harm), permissible (call for shared decision-making), or suspect (entirely cost-driven). By reframing care redesign as thoughtful, responsible care delivery, we can better incentivize physicians to exercise professionalism and maintain medical practice as a public trust.

High-value champions have a great deal of work ahead to redesign care to improve health, reduce TCOC, and investigate outcomes of care redesign. We applaud Drs. Arora and Moriates for once again leading the charge in preparing medical students and residents to deliver higher-value healthcare by emphasizing that effective patient care is not measured by a single episode or clinical decision, but is defined through a lifelong partnership between the patient and the healthcare system. As the country moves toward improved holistic models of care and financing, physician leadership in care redesign is essential to ensure that quality, safety, and patient well-being are not sacrificed at the altar of cost savings.

Disclosures

Dr. Johnson is a Consultant and Advisory Board Member at Oliver Wyman, receives salary support from an AHRQ grant, and has pending potential royalties from licensure of evidence-based appropriate use guidelines/criteria to AgilMD (Agile is a clinical decision support company). The other authors have no relevant disclosures. Dr. Johnson and Dr. Pahwa are Co-directors, High Value Practice Academic Alliance, www.hvpaa.org

In this issue of the Journal of Hospital Medicine^®, Drs. Arora and Moriates highlight an important deficiency in quality improvement efforts designed to reduce overuse of tests and treatments: the potential for trainees—and by extension, more seasoned clinicians—to rationalize minimizing under the guise of high-value care.¹ This insightful perspective from the Co-Directors of Costs of Care should serve as a catalyst for further robust and effective care redesign efforts to optimize the use of all medical resources, including tests, treatments, procedures, consultations, emergency department (ED) visits, and hospital admissions. The formula to root out minimizers is not straightforward and requires an evaluation of wasteful practices in a nuanced and holistic manner that considers not only the frequency that the overused test (or treatment) is ordered but also the collateral impact of not ordering it. This principle has implications for measuring, paying for, and studying high-value care.

Overuse of tests and treatments increases costs and carries a risk of harm, from unnecessary use of creatine kinase–myocardial band (CK-MB) in suspected acute coronary syndrome² to unwarranted administration of antibiotics for asymptomatic bacteriuria³ to over-administration of blood transfusions.⁴ However, decreasing the use of a commonly ordered test is not always clinically appropriate. To illustrate this point, we consider the evidence-based algorithm to deliver best practice in the work-up of pulmonary embolism (PE) by Raja et al; which integrates pretest probability, PERC assessment, and appropriate use of D-dimer and pulmonary CT angiography (CTA).⁵ Avoiding D-dimer testing is appropriate in patients with very low pretest probability who pass a pulmonary embolism rule-out criteria (PERC) clinical assessment and is also appropriate in patients who have sufficiently high clinical probability for PE to justify CTA regardless of a D-dimer result. On the other hand, avoiding D-dimer testing by attributing a patient’s symptoms to anxiety (as a minimizer might do) would increase patient risk, and could potentially increase cost if that patient ends up in intensive care after delayed diagnosis. Following diagnostic algorithms that include physician decision-making and evidence-based guidelines can prevent overuse and underuse, thereby maximizing efficiency and effectiveness. Engaging trainees in the development of such algorithms and decision support tools will serve to ingrain these principles into their practice.

Arora and Moriates highlight the importance of caring for a patient along a continuum rather than simply optimizing practice with respect to a single management decision or an isolated care episode. This approach is fundamental to the quality of care we provide, the public trust our profession still commands, and the total cost of care (TCOC). The two largest contributors to debilitating patient healthcare debt are not overuse of tests and treatments, but ED visits and hospitalizations.⁶ Thus, high-value quality improvement needs to anticipate future healthcare needs, including those that may result from delayed or missed diagnoses. Furthermore, excessive focus on the minutiae of high-value care (fewer daily basic metabolic panels) can lead to change fatigue and divert attention from higher impact utilization. We endorse a holistic approach in which the lens is shifted from the test—and even from the encounter or episode of care—to the entire continuum of care so that we can safeguard against inappropriate minimization. This approach has started to gain traction with policymakers. For example, the state of Maryland has implemented a TCOC hospital payment model predicted to save $1 billion by 2023.⁷ The TCOC model includes a Care Redesign Program whereby hospitals and nonhospital healthcare providers collaborate to improve the quality of care while reducing spending, and cost savings can be used for incentive payments to the nonhospital providers (gainsharing) while simultaneously monitoring quality measures to guard against rationing.⁷ In keeping with the authors’ call to prioritize overall health, this new reimbursement model and others similar to it aim to incentivize the delivery of high-value care across a continuum.

Research is needed to guide best practice from this global perspective; as such, value improvement projects aimed at optimizing use of tests and treatments should include rigorous methodology, measures of downstream outcomes and costs, and balancing safety measures.⁸ For example, the ROMICAT II randomized trial evaluated two diagnostic approaches in emergency department patients with suspected acute coronary syndrome: early coronary computed tomography angiogram (CCTA) and standard ED evaluation.⁹ In addition to outcomes related to the ED visit itself, downstream testing and outcomes for 28 days after the episode were studied. In the acute setting, CCTA decreased time to diagnosis, reduced mean hospital length of stay by 7.6 hours, and resulted in 47% of patients being discharged within 8.6 hours as opposed to only 12% of the standard evaluation cohort. No cases of ACS were missed, and the CCTA cohort has slightly fewer cardiovascular adverse events (P = .18). However, the CCTA patients received significantly more diagnostic and functional testing and higher radiation exposure than the standard evaluation cohort, and underwent modestly higher rates of coronary angiography and percutaneous coronary intervention. The TCOC over the 28-day period was similar at $4,289 for CCTA versus $4,060 for standard care (P = .65).⁹

Reducing the TCOC is imperative to protect patients from the burden of healthcare debt, but concerns have been raised about the ethics of high-value care if decision-making is driven by cost considerations.¹⁰ A recent viewpoint proposed a framework where high-value care recommendations are categorized as obligatory (protecting patients from harm), permissible (call for shared decision-making), or suspect (entirely cost-driven). By reframing care redesign as thoughtful, responsible care delivery, we can better incentivize physicians to exercise professionalism and maintain medical practice as a public trust.

High-value champions have a great deal of work ahead to redesign care to improve health, reduce TCOC, and investigate outcomes of care redesign. We applaud Drs. Arora and Moriates for once again leading the charge in preparing medical students and residents to deliver higher-value healthcare by emphasizing that effective patient care is not measured by a single episode or clinical decision, but is defined through a lifelong partnership between the patient and the healthcare system. As the country moves toward improved holistic models of care and financing, physician leadership in care redesign is essential to ensure that quality, safety, and patient well-being are not sacrificed at the altar of cost savings.

Disclosures

Dr. Johnson is a Consultant and Advisory Board Member at Oliver Wyman, receives salary support from an AHRQ grant, and has pending potential royalties from licensure of evidence-based appropriate use guidelines/criteria to AgilMD (Agile is a clinical decision support company). The other authors have no relevant disclosures. Dr. Johnson and Dr. Pahwa are Co-directors, High Value Practice Academic Alliance, www.hvpaa.org

In this issue of the Journal of Hospital Medicine^®, Drs. Arora and Moriates highlight an important deficiency in quality improvement efforts designed to reduce overuse of tests and treatments: the potential for trainees—and by extension, more seasoned clinicians—to rationalize minimizing under the guise of high-value care.¹ This insightful perspective from the Co-Directors of Costs of Care should serve as a catalyst for further robust and effective care redesign efforts to optimize the use of all medical resources, including tests, treatments, procedures, consultations, emergency department (ED) visits, and hospital admissions. The formula to root out minimizers is not straightforward and requires an evaluation of wasteful practices in a nuanced and holistic manner that considers not only the frequency that the overused test (or treatment) is ordered but also the collateral impact of not ordering it. This principle has implications for measuring, paying for, and studying high-value care.

Overuse of tests and treatments increases costs and carries a risk of harm, from unnecessary use of creatine kinase–myocardial band (CK-MB) in suspected acute coronary syndrome² to unwarranted administration of antibiotics for asymptomatic bacteriuria³ to over-administration of blood transfusions.⁴ However, decreasing the use of a commonly ordered test is not always clinically appropriate. To illustrate this point, we consider the evidence-based algorithm to deliver best practice in the work-up of pulmonary embolism (PE) by Raja et al; which integrates pretest probability, PERC assessment, and appropriate use of D-dimer and pulmonary CT angiography (CTA).⁵ Avoiding D-dimer testing is appropriate in patients with very low pretest probability who pass a pulmonary embolism rule-out criteria (PERC) clinical assessment and is also appropriate in patients who have sufficiently high clinical probability for PE to justify CTA regardless of a D-dimer result. On the other hand, avoiding D-dimer testing by attributing a patient’s symptoms to anxiety (as a minimizer might do) would increase patient risk, and could potentially increase cost if that patient ends up in intensive care after delayed diagnosis. Following diagnostic algorithms that include physician decision-making and evidence-based guidelines can prevent overuse and underuse, thereby maximizing efficiency and effectiveness. Engaging trainees in the development of such algorithms and decision support tools will serve to ingrain these principles into their practice.

Arora and Moriates highlight the importance of caring for a patient along a continuum rather than simply optimizing practice with respect to a single management decision or an isolated care episode. This approach is fundamental to the quality of care we provide, the public trust our profession still commands, and the total cost of care (TCOC). The two largest contributors to debilitating patient healthcare debt are not overuse of tests and treatments, but ED visits and hospitalizations.⁶ Thus, high-value quality improvement needs to anticipate future healthcare needs, including those that may result from delayed or missed diagnoses. Furthermore, excessive focus on the minutiae of high-value care (fewer daily basic metabolic panels) can lead to change fatigue and divert attention from higher impact utilization. We endorse a holistic approach in which the lens is shifted from the test—and even from the encounter or episode of care—to the entire continuum of care so that we can safeguard against inappropriate minimization. This approach has started to gain traction with policymakers. For example, the state of Maryland has implemented a TCOC hospital payment model predicted to save $1 billion by 2023.⁷ The TCOC model includes a Care Redesign Program whereby hospitals and nonhospital healthcare providers collaborate to improve the quality of care while reducing spending, and cost savings can be used for incentive payments to the nonhospital providers (gainsharing) while simultaneously monitoring quality measures to guard against rationing.⁷ In keeping with the authors’ call to prioritize overall health, this new reimbursement model and others similar to it aim to incentivize the delivery of high-value care across a continuum.

Research is needed to guide best practice from this global perspective; as such, value improvement projects aimed at optimizing use of tests and treatments should include rigorous methodology, measures of downstream outcomes and costs, and balancing safety measures.⁸ For example, the ROMICAT II randomized trial evaluated two diagnostic approaches in emergency department patients with suspected acute coronary syndrome: early coronary computed tomography angiogram (CCTA) and standard ED evaluation.⁹ In addition to outcomes related to the ED visit itself, downstream testing and outcomes for 28 days after the episode were studied. In the acute setting, CCTA decreased time to diagnosis, reduced mean hospital length of stay by 7.6 hours, and resulted in 47% of patients being discharged within 8.6 hours as opposed to only 12% of the standard evaluation cohort. No cases of ACS were missed, and the CCTA cohort has slightly fewer cardiovascular adverse events (P = .18). However, the CCTA patients received significantly more diagnostic and functional testing and higher radiation exposure than the standard evaluation cohort, and underwent modestly higher rates of coronary angiography and percutaneous coronary intervention. The TCOC over the 28-day period was similar at $4,289 for CCTA versus $4,060 for standard care (P = .65).⁹

Reducing the TCOC is imperative to protect patients from the burden of healthcare debt, but concerns have been raised about the ethics of high-value care if decision-making is driven by cost considerations.¹⁰ A recent viewpoint proposed a framework where high-value care recommendations are categorized as obligatory (protecting patients from harm), permissible (call for shared decision-making), or suspect (entirely cost-driven). By reframing care redesign as thoughtful, responsible care delivery, we can better incentivize physicians to exercise professionalism and maintain medical practice as a public trust.

High-value champions have a great deal of work ahead to redesign care to improve health, reduce TCOC, and investigate outcomes of care redesign. We applaud Drs. Arora and Moriates for once again leading the charge in preparing medical students and residents to deliver higher-value healthcare by emphasizing that effective patient care is not measured by a single episode or clinical decision, but is defined through a lifelong partnership between the patient and the healthcare system. As the country moves toward improved holistic models of care and financing, physician leadership in care redesign is essential to ensure that quality, safety, and patient well-being are not sacrificed at the altar of cost savings.

Disclosures

Dr. Johnson is a Consultant and Advisory Board Member at Oliver Wyman, receives salary support from an AHRQ grant, and has pending potential royalties from licensure of evidence-based appropriate use guidelines/criteria to AgilMD (Agile is a clinical decision support company). The other authors have no relevant disclosures. Dr. Johnson and Dr. Pahwa are Co-directors, High Value Practice Academic Alliance, www.hvpaa.org

With the escalating need for academic health centers to control costs, high-value care initiatives targeted at residents have exploded. Recent estimates suggest that more than two-thirds of internal medicine residency programs have high-value care curricula.¹ This growth has been catalyzed, in part, by compelling evidence suggesting that where the residents undergo training is strongly associated with their future utilization.² Although we encourage, support, and participate in high-value care education, as hospitalists, there are potential consequences of the high-value care movement in medical training.

Minimizers – physicians who underestimate the signs and symptoms of a patient, hastily concluding that they have the most benign condition possible – have always existed within residency training. The ethos of “doing nothing” has been around since at least the days of the widely read medical satire House of God.³ However, the increasing focus on high-value care creates a socially acceptable banner for minimizers to hide behind when defending inappropriately doing less. For an inpatient with unexplained localized abdominal pain not responding to conservative therapy, a minimizing resident may report to the attending, “They’re fine. I am trying to practice high-value care and avoid getting a CT scan.”

In their 2011 book, Your Medical Mind, Groopman and Hartzband described how people naturally fall on a scale between medical maximizing and minimizing and how this influences their approach toward healthcare.⁴ Researchers have expanded this construct to create a “Maximizer-Minimizer Scale,” which has been used for studying patients and how these traits affect the degree of medical care they receive.⁵ Similar approaches could be used for identifying physicians and trainees at risk of too much minimizer behavior. Although the vast majority of trainees are not minimizers, and overuse continues to be the bigger problem in the majority of academic settings, it is important to understand how the high-value care movement could facilitate minimalist behavior in some residents. Although this article focuses on the educational system, the potential for minimization exists at all levels of clinical practice, including faculty and practicing physicians. Tackling this problem requires understanding the factors that promote the creation of minimizers, how patients and trainees are affected, and the solutions for preventing the spread of minimizers.

Several factors may predispose a resident physician to become a minimizer. For example, resident burnout and overwhelming caseloads can contribute to the desire to decrease work by any means necessary. There are several ways a minimizer can accomplish this goal on inpatient rounds. First, a minimizer may present an important or acute problem as an “outpatient issue” that does not require inpatient workup. Second, minimizers may avoid requesting necessary consults, particularly those associated with intensive workups such as neurology, infectious disease, and rheumatology. Minimizers would claim that this is because of a concern of an unnecessary “costly workup,” when in reality they fear discovery of new problems, more tests to follow-up, and a potentially prolonged length of stay. Ironically, an institutional focus on hospital throughput can reinforce minimizers since the attending physicians or the hospital administrators may applaud them for avoiding “extra nights” in the hospital.

In addition to high workloads, inadequate clinical expertise favors the creation of minimizers. Although resident physicians may be aware that the probability of a rare disease is low, they may not recognize when ruling it out is appropriate. Thus, they could dismiss subtle cues or patterns that point to the need for further workup. Although attending physicians serve as a safety net, it could take time for them to recognize a resident minimizer who may be presenting biased information that influences their clinical decisions. Moreover, attending physicians may avoid further probing so that they are not perceived as promoting overuse and waste.

There are several dangers posed by minimizers, but the most concerning is the impact on patients. Missed diagnoses are a common source of patient maltreatment and contribute to avoidable deaths.⁶ Patients treated by minimizers may continue to experience their acute problem or have to be readmitted because of inadequate treatment. These patients may also lose faith or their trust in the medical system because of inattention to their problems. In fact, minimizing behaviors could have the greatest negative impact on the most vulnerable patients, who often cannot advocate for themselves or who may face conscious and unconscious biases, such as assumptions that they are “pain medication-seeking.”

In addition to harming patients, minimizers can jeopardize learning opportunities. A minimizer resident squanders the chance to recognize and contribute toward caring for a patient with a rare disease, diminishing their overall clinical development. Other trainees lose the opportunity to learn due to consultations or procedures never obtained. Lastly, as inappropriate attitudes and practices of minimizers spread through the hidden curriculum, particularly to medical students beginning their training, the overall clinical learning environment suffers.

There are specific techniques that academic hospitalists and teaching attending physicians can use to help curb the creation of minimizers and promote a clinical learning environment that counters these behaviors. First, instead of focusing on financial costs, it is important for educators to teach the true concept of healthcare value and the primary importance of improving patient outcomes. Embedding appropriateness criteria, such as those from the American College of Radiology, into daily workflows can enable residents to consider not just the cost of imaging but rather the appropriateness given a specific indication.⁷ Training programs can provide residents with a closed-loop feedback on patient outcomes so that they can recognize whether a diagnosis was missed or a necessary test was not ordered. Additionally, it is critical for residents to understand that improving healthcare value requires taking a big picture view of costs, particularly from the perspective of patients.⁸ A patient readmitted after receiving a minimalist workup is more costly to both the patient and the healthcare system.

Second, it is important for the hospitalist faculty to emphasize when a patient has failed a conservative approach and a more specialized, and sometimes intensive, workup or management strategy is appropriate. The classic example is a patient transferred from a community hospital to a tertiary center for further evaluation. Such patients are outside the scope of well-established guidelines. It is precisely these patients that Choosing Wisely or “Less is More” recommendations often do not apply. In contrast, transfer patients often do not end up receiving the specialty procedures that they were originally referred for⁹; it is important that all remain vigilant and committed to high-value care to avoid overuse in these situations.

Exposing residents to cognitive biases is equally important. For example, anchoring can lead to early closure, an easy path for a minimizer to follow. Given the recent focus on the harms related to diagnostic errors, more training in these biases can help promote better patient outcomes.¹⁰

Lastly, it is critical that hospitalists emphasize the importance of prioritizing a patient’s overall health to learners. Although it is tempting for trainees to focus only on acute episodes of a hospital stay, a holistic approach to patients and their quality of life can avoid the minimizer trap. The recent proposal to use home-to-home days in lieu of the routine length of hospital stay is a wonderful example of “measuring what matters to patients” and removing incentives for inappropriately shifting care to other clinicians or venues.¹¹ Likewise, alternative payment models for emphasizing patient outcomes over time can create systems that reinforce holistic views of patient health.

The increasing focus on delivering high-value care has created a socially acceptable excuse for minimizers, who could thrive relatively unchecked in the clinical learning environment. To counter this unintended consequence, hospitalists must learn to identify minimizing behavior and actively guard against these tendencies by highlighting the value of appropriate care, not just doing less, and always striving to provide the best care for patients.

Disclosures

Dr. Arora reports personal fees from the American Board of Internal Medicine and personal fees from McGraw Hill, outside the submitted work. Dr. Moriates reports personal fees from McGraw Hill, outside the submitted work.

With the escalating need for academic health centers to control costs, high-value care initiatives targeted at residents have exploded. Recent estimates suggest that more than two-thirds of internal medicine residency programs have high-value care curricula.¹ This growth has been catalyzed, in part, by compelling evidence suggesting that where the residents undergo training is strongly associated with their future utilization.² Although we encourage, support, and participate in high-value care education, as hospitalists, there are potential consequences of the high-value care movement in medical training.

Minimizers – physicians who underestimate the signs and symptoms of a patient, hastily concluding that they have the most benign condition possible – have always existed within residency training. The ethos of “doing nothing” has been around since at least the days of the widely read medical satire House of God.³ However, the increasing focus on high-value care creates a socially acceptable banner for minimizers to hide behind when defending inappropriately doing less. For an inpatient with unexplained localized abdominal pain not responding to conservative therapy, a minimizing resident may report to the attending, “They’re fine. I am trying to practice high-value care and avoid getting a CT scan.”

In their 2011 book, Your Medical Mind, Groopman and Hartzband described how people naturally fall on a scale between medical maximizing and minimizing and how this influences their approach toward healthcare.⁴ Researchers have expanded this construct to create a “Maximizer-Minimizer Scale,” which has been used for studying patients and how these traits affect the degree of medical care they receive.⁵ Similar approaches could be used for identifying physicians and trainees at risk of too much minimizer behavior. Although the vast majority of trainees are not minimizers, and overuse continues to be the bigger problem in the majority of academic settings, it is important to understand how the high-value care movement could facilitate minimalist behavior in some residents. Although this article focuses on the educational system, the potential for minimization exists at all levels of clinical practice, including faculty and practicing physicians. Tackling this problem requires understanding the factors that promote the creation of minimizers, how patients and trainees are affected, and the solutions for preventing the spread of minimizers.

Several factors may predispose a resident physician to become a minimizer. For example, resident burnout and overwhelming caseloads can contribute to the desire to decrease work by any means necessary. There are several ways a minimizer can accomplish this goal on inpatient rounds. First, a minimizer may present an important or acute problem as an “outpatient issue” that does not require inpatient workup. Second, minimizers may avoid requesting necessary consults, particularly those associated with intensive workups such as neurology, infectious disease, and rheumatology. Minimizers would claim that this is because of a concern of an unnecessary “costly workup,” when in reality they fear discovery of new problems, more tests to follow-up, and a potentially prolonged length of stay. Ironically, an institutional focus on hospital throughput can reinforce minimizers since the attending physicians or the hospital administrators may applaud them for avoiding “extra nights” in the hospital.

In addition to high workloads, inadequate clinical expertise favors the creation of minimizers. Although resident physicians may be aware that the probability of a rare disease is low, they may not recognize when ruling it out is appropriate. Thus, they could dismiss subtle cues or patterns that point to the need for further workup. Although attending physicians serve as a safety net, it could take time for them to recognize a resident minimizer who may be presenting biased information that influences their clinical decisions. Moreover, attending physicians may avoid further probing so that they are not perceived as promoting overuse and waste.

There are several dangers posed by minimizers, but the most concerning is the impact on patients. Missed diagnoses are a common source of patient maltreatment and contribute to avoidable deaths.⁶ Patients treated by minimizers may continue to experience their acute problem or have to be readmitted because of inadequate treatment. These patients may also lose faith or their trust in the medical system because of inattention to their problems. In fact, minimizing behaviors could have the greatest negative impact on the most vulnerable patients, who often cannot advocate for themselves or who may face conscious and unconscious biases, such as assumptions that they are “pain medication-seeking.”

In addition to harming patients, minimizers can jeopardize learning opportunities. A minimizer resident squanders the chance to recognize and contribute toward caring for a patient with a rare disease, diminishing their overall clinical development. Other trainees lose the opportunity to learn due to consultations or procedures never obtained. Lastly, as inappropriate attitudes and practices of minimizers spread through the hidden curriculum, particularly to medical students beginning their training, the overall clinical learning environment suffers.

There are specific techniques that academic hospitalists and teaching attending physicians can use to help curb the creation of minimizers and promote a clinical learning environment that counters these behaviors. First, instead of focusing on financial costs, it is important for educators to teach the true concept of healthcare value and the primary importance of improving patient outcomes. Embedding appropriateness criteria, such as those from the American College of Radiology, into daily workflows can enable residents to consider not just the cost of imaging but rather the appropriateness given a specific indication.⁷ Training programs can provide residents with a closed-loop feedback on patient outcomes so that they can recognize whether a diagnosis was missed or a necessary test was not ordered. Additionally, it is critical for residents to understand that improving healthcare value requires taking a big picture view of costs, particularly from the perspective of patients.⁸ A patient readmitted after receiving a minimalist workup is more costly to both the patient and the healthcare system.

Second, it is important for the hospitalist faculty to emphasize when a patient has failed a conservative approach and a more specialized, and sometimes intensive, workup or management strategy is appropriate. The classic example is a patient transferred from a community hospital to a tertiary center for further evaluation. Such patients are outside the scope of well-established guidelines. It is precisely these patients that Choosing Wisely or “Less is More” recommendations often do not apply. In contrast, transfer patients often do not end up receiving the specialty procedures that they were originally referred for⁹; it is important that all remain vigilant and committed to high-value care to avoid overuse in these situations.

Exposing residents to cognitive biases is equally important. For example, anchoring can lead to early closure, an easy path for a minimizer to follow. Given the recent focus on the harms related to diagnostic errors, more training in these biases can help promote better patient outcomes.¹⁰

Lastly, it is critical that hospitalists emphasize the importance of prioritizing a patient’s overall health to learners. Although it is tempting for trainees to focus only on acute episodes of a hospital stay, a holistic approach to patients and their quality of life can avoid the minimizer trap. The recent proposal to use home-to-home days in lieu of the routine length of hospital stay is a wonderful example of “measuring what matters to patients” and removing incentives for inappropriately shifting care to other clinicians or venues.¹¹ Likewise, alternative payment models for emphasizing patient outcomes over time can create systems that reinforce holistic views of patient health.

The increasing focus on delivering high-value care has created a socially acceptable excuse for minimizers, who could thrive relatively unchecked in the clinical learning environment. To counter this unintended consequence, hospitalists must learn to identify minimizing behavior and actively guard against these tendencies by highlighting the value of appropriate care, not just doing less, and always striving to provide the best care for patients.

Disclosures

Dr. Arora reports personal fees from the American Board of Internal Medicine and personal fees from McGraw Hill, outside the submitted work. Dr. Moriates reports personal fees from McGraw Hill, outside the submitted work.

With the escalating need for academic health centers to control costs, high-value care initiatives targeted at residents have exploded. Recent estimates suggest that more than two-thirds of internal medicine residency programs have high-value care curricula.¹ This growth has been catalyzed, in part, by compelling evidence suggesting that where the residents undergo training is strongly associated with their future utilization.² Although we encourage, support, and participate in high-value care education, as hospitalists, there are potential consequences of the high-value care movement in medical training.

Minimizers – physicians who underestimate the signs and symptoms of a patient, hastily concluding that they have the most benign condition possible – have always existed within residency training. The ethos of “doing nothing” has been around since at least the days of the widely read medical satire House of God.³ However, the increasing focus on high-value care creates a socially acceptable banner for minimizers to hide behind when defending inappropriately doing less. For an inpatient with unexplained localized abdominal pain not responding to conservative therapy, a minimizing resident may report to the attending, “They’re fine. I am trying to practice high-value care and avoid getting a CT scan.”

In their 2011 book, Your Medical Mind, Groopman and Hartzband described how people naturally fall on a scale between medical maximizing and minimizing and how this influences their approach toward healthcare.⁴ Researchers have expanded this construct to create a “Maximizer-Minimizer Scale,” which has been used for studying patients and how these traits affect the degree of medical care they receive.⁵ Similar approaches could be used for identifying physicians and trainees at risk of too much minimizer behavior. Although the vast majority of trainees are not minimizers, and overuse continues to be the bigger problem in the majority of academic settings, it is important to understand how the high-value care movement could facilitate minimalist behavior in some residents. Although this article focuses on the educational system, the potential for minimization exists at all levels of clinical practice, including faculty and practicing physicians. Tackling this problem requires understanding the factors that promote the creation of minimizers, how patients and trainees are affected, and the solutions for preventing the spread of minimizers.

Several factors may predispose a resident physician to become a minimizer. For example, resident burnout and overwhelming caseloads can contribute to the desire to decrease work by any means necessary. There are several ways a minimizer can accomplish this goal on inpatient rounds. First, a minimizer may present an important or acute problem as an “outpatient issue” that does not require inpatient workup. Second, minimizers may avoid requesting necessary consults, particularly those associated with intensive workups such as neurology, infectious disease, and rheumatology. Minimizers would claim that this is because of a concern of an unnecessary “costly workup,” when in reality they fear discovery of new problems, more tests to follow-up, and a potentially prolonged length of stay. Ironically, an institutional focus on hospital throughput can reinforce minimizers since the attending physicians or the hospital administrators may applaud them for avoiding “extra nights” in the hospital.

In addition to high workloads, inadequate clinical expertise favors the creation of minimizers. Although resident physicians may be aware that the probability of a rare disease is low, they may not recognize when ruling it out is appropriate. Thus, they could dismiss subtle cues or patterns that point to the need for further workup. Although attending physicians serve as a safety net, it could take time for them to recognize a resident minimizer who may be presenting biased information that influences their clinical decisions. Moreover, attending physicians may avoid further probing so that they are not perceived as promoting overuse and waste.

There are several dangers posed by minimizers, but the most concerning is the impact on patients. Missed diagnoses are a common source of patient maltreatment and contribute to avoidable deaths.⁶ Patients treated by minimizers may continue to experience their acute problem or have to be readmitted because of inadequate treatment. These patients may also lose faith or their trust in the medical system because of inattention to their problems. In fact, minimizing behaviors could have the greatest negative impact on the most vulnerable patients, who often cannot advocate for themselves or who may face conscious and unconscious biases, such as assumptions that they are “pain medication-seeking.”

In addition to harming patients, minimizers can jeopardize learning opportunities. A minimizer resident squanders the chance to recognize and contribute toward caring for a patient with a rare disease, diminishing their overall clinical development. Other trainees lose the opportunity to learn due to consultations or procedures never obtained. Lastly, as inappropriate attitudes and practices of minimizers spread through the hidden curriculum, particularly to medical students beginning their training, the overall clinical learning environment suffers.

There are specific techniques that academic hospitalists and teaching attending physicians can use to help curb the creation of minimizers and promote a clinical learning environment that counters these behaviors. First, instead of focusing on financial costs, it is important for educators to teach the true concept of healthcare value and the primary importance of improving patient outcomes. Embedding appropriateness criteria, such as those from the American College of Radiology, into daily workflows can enable residents to consider not just the cost of imaging but rather the appropriateness given a specific indication.⁷ Training programs can provide residents with a closed-loop feedback on patient outcomes so that they can recognize whether a diagnosis was missed or a necessary test was not ordered. Additionally, it is critical for residents to understand that improving healthcare value requires taking a big picture view of costs, particularly from the perspective of patients.⁸ A patient readmitted after receiving a minimalist workup is more costly to both the patient and the healthcare system.

Second, it is important for the hospitalist faculty to emphasize when a patient has failed a conservative approach and a more specialized, and sometimes intensive, workup or management strategy is appropriate. The classic example is a patient transferred from a community hospital to a tertiary center for further evaluation. Such patients are outside the scope of well-established guidelines. It is precisely these patients that Choosing Wisely or “Less is More” recommendations often do not apply. In contrast, transfer patients often do not end up receiving the specialty procedures that they were originally referred for⁹; it is important that all remain vigilant and committed to high-value care to avoid overuse in these situations.

Exposing residents to cognitive biases is equally important. For example, anchoring can lead to early closure, an easy path for a minimizer to follow. Given the recent focus on the harms related to diagnostic errors, more training in these biases can help promote better patient outcomes.¹⁰

Lastly, it is critical that hospitalists emphasize the importance of prioritizing a patient’s overall health to learners. Although it is tempting for trainees to focus only on acute episodes of a hospital stay, a holistic approach to patients and their quality of life can avoid the minimizer trap. The recent proposal to use home-to-home days in lieu of the routine length of hospital stay is a wonderful example of “measuring what matters to patients” and removing incentives for inappropriately shifting care to other clinicians or venues.¹¹ Likewise, alternative payment models for emphasizing patient outcomes over time can create systems that reinforce holistic views of patient health.

The increasing focus on delivering high-value care has created a socially acceptable excuse for minimizers, who could thrive relatively unchecked in the clinical learning environment. To counter this unintended consequence, hospitalists must learn to identify minimizing behavior and actively guard against these tendencies by highlighting the value of appropriate care, not just doing less, and always striving to provide the best care for patients.

Disclosures

Dr. Arora reports personal fees from the American Board of Internal Medicine and personal fees from McGraw Hill, outside the submitted work. Dr. Moriates reports personal fees from McGraw Hill, outside the submitted work.

Malignant bowel obstruction (MBO) is a catastrophic complication of cancer that often requires hospitalization and a multidisciplinary approach in its management. Hospitalists frequently collaborate with such specialties as Hematology/Oncology, Surgery, Palliative Medicine, and Interventional Radiology in arriving at a treatment plan.

Initial management is focused on hydration, bowel rest and decompression via nasogastric (NG) tube. Surgical resection or endoscopic stenting should be considered early.¹ However, patients who present in the terminal stages may be poor candidates for these options due to diminished functional status, multiple areas of obstruction, complicated anatomy limiting intervention, or an associated large volume of ascites.

Presence of inoperable MBO portends a poor prognosis, often measured in weeks.² Presentation often occurs in the context of a sentinel hospitalization, signifying a shift in disease course.^3,4 It is essential for hospitalists to be familiar with noninvasive therapies for inoperable MBO given the increasing role of hospitalists in providing inpatient palliative care. Palliative pharmacologic management of MBO can reduce symptom burden during these terminal stages and will be the focus of this paper.

Malignant bowel obstruction occurs in about 3%-15% of patients with cancer.² A consensus definition of MBO established the following specific criteria: (1) clinical evidence of bowel obstruction, (2) obstruction distal to the ligament of Treitz, and (3) the presence of primary intra-abdominal cancer with incurable disease or extra-abdominal cancer with peritoneal involvement.⁵ The most common malignancies are gastric, colorectal, and ovarian in origin.^1,2 The most common extra-abdominal malignancies associated with MBO are breast, melanoma, and lung. MBO is most frequently diagnosed during the advanced stages of cancer.² The obstruction can involve a partial or total blockage of the small or large intestine from either an intrinsic or extrinsic source. Peristalsis may also be impaired via direct tumor infiltration of the intestinal walls or within the enteric nervous system or celiac plexus. Other etiologies of MBO include peritoneal carcinomatosis and radiation-induced fibrosis.^1,6 The obstruction can occur at a single level or involve multiple areas, which usually precludes surgical intervention.²

Symptoms of MBO can be insidious in onset and take several weeks to manifest. The most prevalent symptoms are nausea, vomiting, constipation, abdominal pain, and distension.^2,6 The intermittent pattern of symptoms may evolve into continuous episodes with spontaneous remission in between. The etiology of symptoms can be attributed to distension proximal to the site of obstruction with concomitantly increased gastrointestinal and pancreaticobiliary secretions.

The distension creates a “hypertensive state” in the intestinal lumen causing enterochromaffin cells to release serotonin which activates the enteric nervous system and its effectors including substance P, nitric oxide, acetylcholine, somatostatin, and vasoactive intestinal peptide (VIP). These neurotransmitters stimulate the secretomotor actions that cause hypersecretion of mucus from cells of the intestinal crypts. Additional water and sodium secretions accumulate due to the expanded surface area of the bowel.^1,2 Overloaded with luminal contents, the bowel attempts to overcome the obstruction by contracting, which leads to colicky abdominal pain. Tumor burden can also damage the intestinal epithelium and cause continuous pain.

The buildup of secretions can lead to translocation of fluid into the peritoneum (“third spacing”), bowel ischemia, perforation, or sepsis. The combination of poor oral intake, gastrointestinal fluid loss, and sequestration can lead to profound dehydration on presentation.^2,7

Fluid resuscitation, electrolyte repletion, and a trial of NG tube decompression are part of the initial management of MBO (Figure ). While studies have shown that moderate intravenous hydration can minimize nausea and drowsiness, excessive fluids may worsen bowel edema and exacerbate vomiting.^1,8 NG tube decompression is most effective in patients with proximal obstructions but some studies suggest it can decrease vomiting in patients with colonic obstructions as well.⁹ Computed tomography imaging can identify the extent of the tumor, the transition point of the obstruction, and any distant metastases. Surgery, Gastroenterology, and/or Interventional Radiology consultation should be obtained early to evaluate options for direct decompression. Hematology/Oncology and Radiation/Oncology referral may help delineate prognosis and achievable outcomes. Emergent exploratory surgery may be required in cases of bowel perforation or ischemia. Otherwise, a planned surgical resection should be considered in those with an isolated resectable lesion and acceptable perioperative risk. Colorectal or duodenal stents may be an option for those who are not surgical candidates or as a bridge to surgery.

As bowel obstruction is often a late manifestation of advanced malignancy, many patients may not be appropriate candidates for operative/interventional treatment due to malnutrition, comorbid conditions, or anatomic considerations. For these individuals, pharmacologic management is the mainstay of treatment. Additionally, the pharmacologic approaches detailed below may provide benefit as adjunctive therapy for patients undergoing procedural intervention.⁷ Consultation for early palliative care can improve symptom control as well as clarify goals of care.

Given the pathophysiology of MBO, pharmacologic therapies are focused on controlling nausea and pain while reducing bowel edema and secretions.

Antiemetic Agents

Nausea and vomiting in MBO are due to activation of vagal nerve fibers in the gastric wall and stimulation of the chemoreceptor trigger zone (CTZ).¹⁰ Dopamine antagonists have started to gain favor for MBO compared to more commonly used antiemetics such as the serotonin antagonists. Haloperidol should be considered as a first-line antiemetic in patients with MBO. Its potent D₂-receptor antagonistic properties block receptors in the CTZ. The high affinity of the drug for only the D₂-receptor makes it preferable to alternative agents in the same class such as chlorpromazine. However, haloperidol may cause or worsen QT prolongation and should be avoided in patients with Parkinson’s disease. The medication has less sedative and unwanted anticholinergic side effects due to its limited interaction with histaminergic and acetylcholine receptors.¹¹ Haloperidol has been shown in the past to be efficacious for post-operative nausea but there are few randomized controlled trials in the terminally ill.¹² Nonetheless, recent consensus guidelines from the Multinational Association of Supportive Care recommended haloperidol as the initial treatment of nausea for individuals with MBO based on available systematic reviews.¹⁰

Other dopamine antagonists remain good options, though they may cause additional side effects due to actions on other receptor types. Metoclopramide, another D₂-receptor antagonist, has been shown to be effective in the treatment of nausea and vomiting due to advanced cancer.¹³ However as a prokinetic agent, this medication should be avoided in those with complete MBO and only considered in those with partial MBO.^10,14

Olanzapine, an atypical antipsychotic, may also have a role in controlling nausea in patients with MBO. It functions as a 5-HT_2A and D₂-receptor antagonist, with a slightly greater affinity for the 5-HT_2A receptor. Olanzapine thus can target two critical receptors playing a role in nausea and vomiting. A study of patients with incomplete bowel obstruction found the addition of olanzapine significantly decreased nausea and vomiting in patients who were refractory to other treatments including steroids and haloperidol.¹⁵ Olanzapine has the added advantage of single-day dosing as well as an oral disintegrating formulation.¹⁶

Intravenous and sublingual preparations of 5-HT₃ receptor antagonists such as ondansetron are commonly used in the inpatient setting. These medications are potent antiemetics that exhibit their effects via pathways where serotonin acts as a neurotransmitter.¹⁷ An alternative agent, tropisetron, has shown promise when used alone or in conjunction with metoclopramide but is not currently available in the US.¹⁸ Granisetron is available in a transdermal formulation, which can be very convenient for patients with bowel obstruction. Its mechanism of action differs from ondansetron as it is an allosteric inhibitor rather than a competitive inhibitor.¹⁹ Granisetron needs more specific study with regards to its role in MBO.

Although haloperidol remains the initial choice, combination therapy can help to decrease the risk of extrapyramidal symptoms seen with higher doses of dopaminergic monotherapy.

Analgesics

Pain control is an essential part of the palliative treatment of MBO as bowel distention, secretions, and edema can cause rapid onset of pain. Parenteral step three opioids remain the optimal initial choice since patients are unable to take medications orally and may have compromised absorption. Opioids address both the colicky and continuous aspects of MBO pain.

Short-acting intravenous opioids such as morphine or hydromorphone may be scheduled every four hours with breakthrough dosing every hour in between. Alternatively, analgesics can be administered via a patient-controlled analgesia (PCA) pump.¹ Although doses vary across patients, opioid-naïve patients can be initiated on a low dose therapy such as hydromorphone 0.2 mg IV/SC or morphine 1 mg IV/SC every four hours as needed for pain control.

Ongoing pain management for patients with MBO requires coordination of care. Many patients will elect to receive hospice care following discharge. Direct communication with palliative consultants and hospice providers can help facilitate a smooth transition. In patients for whom bowel obstruction resolves, transition to oral opioids based on morphine equivalent daily dose is indicated with further dose adjustment as patients may have reduced pain at this stage.

Options for patients with unresolved obstruction include transdermal and sublingual preparations as well as outpatient PCA with hospice support. Transdermal fentanyl patch can be useful but onset of peak levels occur within 8-12 hours.²⁰ The patch is usually exchanged every 72 hours and is most effective when applied to areas containing adipose tissue which may limit its use in cachectic patients. The liquid preparation of methadone can be useful even in patients with unresolved MBO. Its lipophilic properties allow for ease of absorption.²¹ A baseline electrocardiogram (EKG) is recommended prior to methadone initiation due to the potential for QT prolongation. Methadone should not be a first-line option for opioid-naïve individuals due to its longer onset of action which limits rapid dose titration. Close collaboration with palliative medicine is highly recommended when using longer acting opioids.

Antisecretory Agents

Antisecretory agents are a mainstay of the pharmacologic management of inoperable MBO. Medications that reduce secretions and bowel edema include: somatostatin analogs, H2-blockers, proton pump inhibitors (PPIs), steroids, and anticholinergic agents. Table 2 summarizes the major studies comparing various antisecretory medications.

Octreotide, a somatostatin analog, has been increasingly used for the palliative treatment of MBO. The mechanism of action involves splanchnic vasoconstriction, reduction of intestinal and pancreatic secretions (via inhibition of VIP), decrease in gastric emptying, and slowing of smooth muscle contractions.²² Octreotide comes in an immediate-release formulation with an initial subcutaneous dose of 100 µg three or four times per day. Most patients will require 300-800 µg/day, maximum dose being up to 1 mg/day.^22,23 A long-acting formulation, lanreotide, exists but can be difficult to obtain and may not provide the immediate relief needed in an acute care setting.

Initiation of octreotide should be considered in the presence of persistent symptoms. Studies have suggested that the benefit of octreotide is most apparent in the first three days of treatment (range 1-5 days).^6,22,24 The medication should be discontinued if there is no clinical improvement such as reduction of NG tube output. Octreotide has been shown to be more efficacious than anticholinergic agents in reducing secretions as well as frequency of nausea and vomiting.^8,25-28 Octreotide expedites NG tube removal, recovery of bowel function, and improvement in quality of life.^29-32 The medication should also be considered in cases of recurrent MBO that previously responded to the medication.

Octreotide is considered the first-line agent in the palliative treatment of MBO, however the medication is costly. Recent studies suggest combination therapy with steroids and H2-blockers or PPIs may be an equally effective and less expensive alternative. The primary rationale for the use of steroids in MBO is their ability to decrease peritumoral edema and promote salt and water absorption from the intestine.^1,2 PPIs and H2-blockers decrease distension, pain, and vomiting by reducing the volume of gastric secretions.³³ A recent meta-analysis of phase 3 trials found both PPIs and H2-blockers to be effective in lowering volumes of gastric aspirates with ranitidine being slightly superior.³⁴

Initial research into the utility of steroids in MBO garnered mixed results. One study showed marginal benefit for steroid plus octreotide combination therapy compared to octreotide, in a cohort of 27 patients.³⁵ A subsequent review of practice patterns in the management of terminal MBO in Japan found that patients given steroids in combination with octreotide compared to octreotide alone were more likely to undergo early NG tube removal.³⁶ A 1999 systematic review of corticosteroid treatment of MBO concluded low morbidity associated with the medications with a trend toward benefit that was not statistically significant.³⁷ A 2015 study by Currow showed the addition of octreotide in patients already on a regime of dexamethasone and ranitidine did not improve the number of days free from vomiting but did reduce vomiting episodes in those with the most refractory symptoms.³⁸

Collectively, the studies suggest that combination therapy with steroid and PPI or H2 blocker could be a less expensive option in the initial management of MBO. Alternatively, steroids may provide additional relief in patients with continued symptoms on octreotide and H2-blockers. Dexamethasone is preferable given its longer half-life and decreased propensity for sodium retention. Dosing of dexamethasone should be 8 mg IV once a day.³⁸

Anticholinergic agents also reduce secretions. However, they are considered second-line therapy given their lower efficacy compared to other treatment options as well as their propensity to worsen cognitive function.^1,2 Anticholinergics may benefit patients with continued symptoms who cannot tolerate the side effects of other treatments. Scopolamine, also known as hyoscine hydrobromide in the US, should be avoided as it crosses the blood-brain barrier. The quaternary formulation, scopolamine butylbromide (hyoscine butylbromide), does not pass this barrier but is currently not available in the US. Glycopyrrolate may be considered as it is also a quaternary ammonium compound that does not cross the blood-brain barrier. Several case reports have described its effectiveness in the resolution of refractory nausea and vomiting in combination with haloperidol and hydromorphone for symptom control.³⁹ Effective oral care is imperative if anticholinergics are used in order to prevent the unpleasant feeling of dry mouth.

While initial management of MBO often requires placement of an NG tube, prolonged placement can increase the risk for erosions, aspiration, and sinus infections. Removal of the NG tube is most successful when secretions are minimal, but this may not happen unless the obstruction resolves. Some patients may elect to keep an NG tube if symptoms cannot be otherwise controlled by medications.

A venting gastrostomy tube can be considered as an alternative to prolonged NG tube placement. The tube may help alleviate distressing symptoms and can enhance the quality of life of patients by allowing the sensation of oral intake, though it will not allow for absorption of nutrients.⁴⁰ Although a low risk procedure, patients may be too frail to undergo the procedure and may have postprocedure pain and complications. Anatomic abnormalities such as overlying bowel may also prevent the noninvasive percutaneous approach.

In patients with unresolved obstruction, oral intake should be reinitiated with caution with the patient’s wishes taken into account at all times. Some patients may prioritize the comfort derived from eating small amounts over any associated risks of increased nausea and vomiting.

Parenteral nutrition should be avoided in those with inoperable MBO in the advanced stages. The risks of infection, refeeding syndrome, and the discomfort of an intravenous line and intermittent testing may outweigh any benefits given the overall prognosis.^41,42

Hospitalists are often involved in the initial care of patients with advanced malignancy who present with MBO. When interventions or surgeries to directly alleviate the obstruction are not possible, pharmacologic options are essential in managing burdensome symptoms and improving quality of life. Early Palliative Care referral can also assist with symptom management, emotional support, clarification of goals of care, and transition to the outpatient setting. While patients with inoperable MBO have a poor prognosis, hospitalists can play a vital role in alleviation of suffering in this devastating complication of advanced cancer.

Disclosures

The authors have nothing to disclose.

Malignant bowel obstruction (MBO) is a catastrophic complication of cancer that often requires hospitalization and a multidisciplinary approach in its management. Hospitalists frequently collaborate with such specialties as Hematology/Oncology, Surgery, Palliative Medicine, and Interventional Radiology in arriving at a treatment plan.

Initial management is focused on hydration, bowel rest and decompression via nasogastric (NG) tube. Surgical resection or endoscopic stenting should be considered early.¹ However, patients who present in the terminal stages may be poor candidates for these options due to diminished functional status, multiple areas of obstruction, complicated anatomy limiting intervention, or an associated large volume of ascites.

Presence of inoperable MBO portends a poor prognosis, often measured in weeks.² Presentation often occurs in the context of a sentinel hospitalization, signifying a shift in disease course.^3,4 It is essential for hospitalists to be familiar with noninvasive therapies for inoperable MBO given the increasing role of hospitalists in providing inpatient palliative care. Palliative pharmacologic management of MBO can reduce symptom burden during these terminal stages and will be the focus of this paper.

Malignant bowel obstruction occurs in about 3%-15% of patients with cancer.² A consensus definition of MBO established the following specific criteria: (1) clinical evidence of bowel obstruction, (2) obstruction distal to the ligament of Treitz, and (3) the presence of primary intra-abdominal cancer with incurable disease or extra-abdominal cancer with peritoneal involvement.⁵ The most common malignancies are gastric, colorectal, and ovarian in origin.^1,2 The most common extra-abdominal malignancies associated with MBO are breast, melanoma, and lung. MBO is most frequently diagnosed during the advanced stages of cancer.² The obstruction can involve a partial or total blockage of the small or large intestine from either an intrinsic or extrinsic source. Peristalsis may also be impaired via direct tumor infiltration of the intestinal walls or within the enteric nervous system or celiac plexus. Other etiologies of MBO include peritoneal carcinomatosis and radiation-induced fibrosis.^1,6 The obstruction can occur at a single level or involve multiple areas, which usually precludes surgical intervention.²

Symptoms of MBO can be insidious in onset and take several weeks to manifest. The most prevalent symptoms are nausea, vomiting, constipation, abdominal pain, and distension.^2,6 The intermittent pattern of symptoms may evolve into continuous episodes with spontaneous remission in between. The etiology of symptoms can be attributed to distension proximal to the site of obstruction with concomitantly increased gastrointestinal and pancreaticobiliary secretions.

The distension creates a “hypertensive state” in the intestinal lumen causing enterochromaffin cells to release serotonin which activates the enteric nervous system and its effectors including substance P, nitric oxide, acetylcholine, somatostatin, and vasoactive intestinal peptide (VIP). These neurotransmitters stimulate the secretomotor actions that cause hypersecretion of mucus from cells of the intestinal crypts. Additional water and sodium secretions accumulate due to the expanded surface area of the bowel.^1,2 Overloaded with luminal contents, the bowel attempts to overcome the obstruction by contracting, which leads to colicky abdominal pain. Tumor burden can also damage the intestinal epithelium and cause continuous pain.

The buildup of secretions can lead to translocation of fluid into the peritoneum (“third spacing”), bowel ischemia, perforation, or sepsis. The combination of poor oral intake, gastrointestinal fluid loss, and sequestration can lead to profound dehydration on presentation.^2,7

Fluid resuscitation, electrolyte repletion, and a trial of NG tube decompression are part of the initial management of MBO (Figure ). While studies have shown that moderate intravenous hydration can minimize nausea and drowsiness, excessive fluids may worsen bowel edema and exacerbate vomiting.^1,8 NG tube decompression is most effective in patients with proximal obstructions but some studies suggest it can decrease vomiting in patients with colonic obstructions as well.⁹ Computed tomography imaging can identify the extent of the tumor, the transition point of the obstruction, and any distant metastases. Surgery, Gastroenterology, and/or Interventional Radiology consultation should be obtained early to evaluate options for direct decompression. Hematology/Oncology and Radiation/Oncology referral may help delineate prognosis and achievable outcomes. Emergent exploratory surgery may be required in cases of bowel perforation or ischemia. Otherwise, a planned surgical resection should be considered in those with an isolated resectable lesion and acceptable perioperative risk. Colorectal or duodenal stents may be an option for those who are not surgical candidates or as a bridge to surgery.

As bowel obstruction is often a late manifestation of advanced malignancy, many patients may not be appropriate candidates for operative/interventional treatment due to malnutrition, comorbid conditions, or anatomic considerations. For these individuals, pharmacologic management is the mainstay of treatment. Additionally, the pharmacologic approaches detailed below may provide benefit as adjunctive therapy for patients undergoing procedural intervention.⁷ Consultation for early palliative care can improve symptom control as well as clarify goals of care.

Given the pathophysiology of MBO, pharmacologic therapies are focused on controlling nausea and pain while reducing bowel edema and secretions.

Antiemetic Agents

Nausea and vomiting in MBO are due to activation of vagal nerve fibers in the gastric wall and stimulation of the chemoreceptor trigger zone (CTZ).¹⁰ Dopamine antagonists have started to gain favor for MBO compared to more commonly used antiemetics such as the serotonin antagonists. Haloperidol should be considered as a first-line antiemetic in patients with MBO. Its potent D₂-receptor antagonistic properties block receptors in the CTZ. The high affinity of the drug for only the D₂-receptor makes it preferable to alternative agents in the same class such as chlorpromazine. However, haloperidol may cause or worsen QT prolongation and should be avoided in patients with Parkinson’s disease. The medication has less sedative and unwanted anticholinergic side effects due to its limited interaction with histaminergic and acetylcholine receptors.¹¹ Haloperidol has been shown in the past to be efficacious for post-operative nausea but there are few randomized controlled trials in the terminally ill.¹² Nonetheless, recent consensus guidelines from the Multinational Association of Supportive Care recommended haloperidol as the initial treatment of nausea for individuals with MBO based on available systematic reviews.¹⁰

Other dopamine antagonists remain good options, though they may cause additional side effects due to actions on other receptor types. Metoclopramide, another D₂-receptor antagonist, has been shown to be effective in the treatment of nausea and vomiting due to advanced cancer.¹³ However as a prokinetic agent, this medication should be avoided in those with complete MBO and only considered in those with partial MBO.^10,14

Olanzapine, an atypical antipsychotic, may also have a role in controlling nausea in patients with MBO. It functions as a 5-HT_2A and D₂-receptor antagonist, with a slightly greater affinity for the 5-HT_2A receptor. Olanzapine thus can target two critical receptors playing a role in nausea and vomiting. A study of patients with incomplete bowel obstruction found the addition of olanzapine significantly decreased nausea and vomiting in patients who were refractory to other treatments including steroids and haloperidol.¹⁵ Olanzapine has the added advantage of single-day dosing as well as an oral disintegrating formulation.¹⁶

Intravenous and sublingual preparations of 5-HT₃ receptor antagonists such as ondansetron are commonly used in the inpatient setting. These medications are potent antiemetics that exhibit their effects via pathways where serotonin acts as a neurotransmitter.¹⁷ An alternative agent, tropisetron, has shown promise when used alone or in conjunction with metoclopramide but is not currently available in the US.¹⁸ Granisetron is available in a transdermal formulation, which can be very convenient for patients with bowel obstruction. Its mechanism of action differs from ondansetron as it is an allosteric inhibitor rather than a competitive inhibitor.¹⁹ Granisetron needs more specific study with regards to its role in MBO.

Although haloperidol remains the initial choice, combination therapy can help to decrease the risk of extrapyramidal symptoms seen with higher doses of dopaminergic monotherapy.

Analgesics

Pain control is an essential part of the palliative treatment of MBO as bowel distention, secretions, and edema can cause rapid onset of pain. Parenteral step three opioids remain the optimal initial choice since patients are unable to take medications orally and may have compromised absorption. Opioids address both the colicky and continuous aspects of MBO pain.

Short-acting intravenous opioids such as morphine or hydromorphone may be scheduled every four hours with breakthrough dosing every hour in between. Alternatively, analgesics can be administered via a patient-controlled analgesia (PCA) pump.¹ Although doses vary across patients, opioid-naïve patients can be initiated on a low dose therapy such as hydromorphone 0.2 mg IV/SC or morphine 1 mg IV/SC every four hours as needed for pain control.

Ongoing pain management for patients with MBO requires coordination of care. Many patients will elect to receive hospice care following discharge. Direct communication with palliative consultants and hospice providers can help facilitate a smooth transition. In patients for whom bowel obstruction resolves, transition to oral opioids based on morphine equivalent daily dose is indicated with further dose adjustment as patients may have reduced pain at this stage.

Options for patients with unresolved obstruction include transdermal and sublingual preparations as well as outpatient PCA with hospice support. Transdermal fentanyl patch can be useful but onset of peak levels occur within 8-12 hours.²⁰ The patch is usually exchanged every 72 hours and is most effective when applied to areas containing adipose tissue which may limit its use in cachectic patients. The liquid preparation of methadone can be useful even in patients with unresolved MBO. Its lipophilic properties allow for ease of absorption.²¹ A baseline electrocardiogram (EKG) is recommended prior to methadone initiation due to the potential for QT prolongation. Methadone should not be a first-line option for opioid-naïve individuals due to its longer onset of action which limits rapid dose titration. Close collaboration with palliative medicine is highly recommended when using longer acting opioids.

Antisecretory Agents

Antisecretory agents are a mainstay of the pharmacologic management of inoperable MBO. Medications that reduce secretions and bowel edema include: somatostatin analogs, H2-blockers, proton pump inhibitors (PPIs), steroids, and anticholinergic agents. Table 2 summarizes the major studies comparing various antisecretory medications.

Octreotide, a somatostatin analog, has been increasingly used for the palliative treatment of MBO. The mechanism of action involves splanchnic vasoconstriction, reduction of intestinal and pancreatic secretions (via inhibition of VIP), decrease in gastric emptying, and slowing of smooth muscle contractions.²² Octreotide comes in an immediate-release formulation with an initial subcutaneous dose of 100 µg three or four times per day. Most patients will require 300-800 µg/day, maximum dose being up to 1 mg/day.^22,23 A long-acting formulation, lanreotide, exists but can be difficult to obtain and may not provide the immediate relief needed in an acute care setting.

Initiation of octreotide should be considered in the presence of persistent symptoms. Studies have suggested that the benefit of octreotide is most apparent in the first three days of treatment (range 1-5 days).^6,22,24 The medication should be discontinued if there is no clinical improvement such as reduction of NG tube output. Octreotide has been shown to be more efficacious than anticholinergic agents in reducing secretions as well as frequency of nausea and vomiting.^8,25-28 Octreotide expedites NG tube removal, recovery of bowel function, and improvement in quality of life.^29-32 The medication should also be considered in cases of recurrent MBO that previously responded to the medication.

Octreotide is considered the first-line agent in the palliative treatment of MBO, however the medication is costly. Recent studies suggest combination therapy with steroids and H2-blockers or PPIs may be an equally effective and less expensive alternative. The primary rationale for the use of steroids in MBO is their ability to decrease peritumoral edema and promote salt and water absorption from the intestine.^1,2 PPIs and H2-blockers decrease distension, pain, and vomiting by reducing the volume of gastric secretions.³³ A recent meta-analysis of phase 3 trials found both PPIs and H2-blockers to be effective in lowering volumes of gastric aspirates with ranitidine being slightly superior.³⁴

Initial research into the utility of steroids in MBO garnered mixed results. One study showed marginal benefit for steroid plus octreotide combination therapy compared to octreotide, in a cohort of 27 patients.³⁵ A subsequent review of practice patterns in the management of terminal MBO in Japan found that patients given steroids in combination with octreotide compared to octreotide alone were more likely to undergo early NG tube removal.³⁶ A 1999 systematic review of corticosteroid treatment of MBO concluded low morbidity associated with the medications with a trend toward benefit that was not statistically significant.³⁷ A 2015 study by Currow showed the addition of octreotide in patients already on a regime of dexamethasone and ranitidine did not improve the number of days free from vomiting but did reduce vomiting episodes in those with the most refractory symptoms.³⁸

Collectively, the studies suggest that combination therapy with steroid and PPI or H2 blocker could be a less expensive option in the initial management of MBO. Alternatively, steroids may provide additional relief in patients with continued symptoms on octreotide and H2-blockers. Dexamethasone is preferable given its longer half-life and decreased propensity for sodium retention. Dosing of dexamethasone should be 8 mg IV once a day.³⁸

Anticholinergic agents also reduce secretions. However, they are considered second-line therapy given their lower efficacy compared to other treatment options as well as their propensity to worsen cognitive function.^1,2 Anticholinergics may benefit patients with continued symptoms who cannot tolerate the side effects of other treatments. Scopolamine, also known as hyoscine hydrobromide in the US, should be avoided as it crosses the blood-brain barrier. The quaternary formulation, scopolamine butylbromide (hyoscine butylbromide), does not pass this barrier but is currently not available in the US. Glycopyrrolate may be considered as it is also a quaternary ammonium compound that does not cross the blood-brain barrier. Several case reports have described its effectiveness in the resolution of refractory nausea and vomiting in combination with haloperidol and hydromorphone for symptom control.³⁹ Effective oral care is imperative if anticholinergics are used in order to prevent the unpleasant feeling of dry mouth.

While initial management of MBO often requires placement of an NG tube, prolonged placement can increase the risk for erosions, aspiration, and sinus infections. Removal of the NG tube is most successful when secretions are minimal, but this may not happen unless the obstruction resolves. Some patients may elect to keep an NG tube if symptoms cannot be otherwise controlled by medications.

A venting gastrostomy tube can be considered as an alternative to prolonged NG tube placement. The tube may help alleviate distressing symptoms and can enhance the quality of life of patients by allowing the sensation of oral intake, though it will not allow for absorption of nutrients.⁴⁰ Although a low risk procedure, patients may be too frail to undergo the procedure and may have postprocedure pain and complications. Anatomic abnormalities such as overlying bowel may also prevent the noninvasive percutaneous approach.

In patients with unresolved obstruction, oral intake should be reinitiated with caution with the patient’s wishes taken into account at all times. Some patients may prioritize the comfort derived from eating small amounts over any associated risks of increased nausea and vomiting.

Parenteral nutrition should be avoided in those with inoperable MBO in the advanced stages. The risks of infection, refeeding syndrome, and the discomfort of an intravenous line and intermittent testing may outweigh any benefits given the overall prognosis.^41,42

Hospitalists are often involved in the initial care of patients with advanced malignancy who present with MBO. When interventions or surgeries to directly alleviate the obstruction are not possible, pharmacologic options are essential in managing burdensome symptoms and improving quality of life. Early Palliative Care referral can also assist with symptom management, emotional support, clarification of goals of care, and transition to the outpatient setting. While patients with inoperable MBO have a poor prognosis, hospitalists can play a vital role in alleviation of suffering in this devastating complication of advanced cancer.

Disclosures

The authors have nothing to disclose.

Malignant bowel obstruction (MBO) is a catastrophic complication of cancer that often requires hospitalization and a multidisciplinary approach in its management. Hospitalists frequently collaborate with such specialties as Hematology/Oncology, Surgery, Palliative Medicine, and Interventional Radiology in arriving at a treatment plan.

Initial management is focused on hydration, bowel rest and decompression via nasogastric (NG) tube. Surgical resection or endoscopic stenting should be considered early.¹ However, patients who present in the terminal stages may be poor candidates for these options due to diminished functional status, multiple areas of obstruction, complicated anatomy limiting intervention, or an associated large volume of ascites.

Presence of inoperable MBO portends a poor prognosis, often measured in weeks.² Presentation often occurs in the context of a sentinel hospitalization, signifying a shift in disease course.^3,4 It is essential for hospitalists to be familiar with noninvasive therapies for inoperable MBO given the increasing role of hospitalists in providing inpatient palliative care. Palliative pharmacologic management of MBO can reduce symptom burden during these terminal stages and will be the focus of this paper.

Malignant bowel obstruction occurs in about 3%-15% of patients with cancer.² A consensus definition of MBO established the following specific criteria: (1) clinical evidence of bowel obstruction, (2) obstruction distal to the ligament of Treitz, and (3) the presence of primary intra-abdominal cancer with incurable disease or extra-abdominal cancer with peritoneal involvement.⁵ The most common malignancies are gastric, colorectal, and ovarian in origin.^1,2 The most common extra-abdominal malignancies associated with MBO are breast, melanoma, and lung. MBO is most frequently diagnosed during the advanced stages of cancer.² The obstruction can involve a partial or total blockage of the small or large intestine from either an intrinsic or extrinsic source. Peristalsis may also be impaired via direct tumor infiltration of the intestinal walls or within the enteric nervous system or celiac plexus. Other etiologies of MBO include peritoneal carcinomatosis and radiation-induced fibrosis.^1,6 The obstruction can occur at a single level or involve multiple areas, which usually precludes surgical intervention.²

Symptoms of MBO can be insidious in onset and take several weeks to manifest. The most prevalent symptoms are nausea, vomiting, constipation, abdominal pain, and distension.^2,6 The intermittent pattern of symptoms may evolve into continuous episodes with spontaneous remission in between. The etiology of symptoms can be attributed to distension proximal to the site of obstruction with concomitantly increased gastrointestinal and pancreaticobiliary secretions.

The distension creates a “hypertensive state” in the intestinal lumen causing enterochromaffin cells to release serotonin which activates the enteric nervous system and its effectors including substance P, nitric oxide, acetylcholine, somatostatin, and vasoactive intestinal peptide (VIP). These neurotransmitters stimulate the secretomotor actions that cause hypersecretion of mucus from cells of the intestinal crypts. Additional water and sodium secretions accumulate due to the expanded surface area of the bowel.^1,2 Overloaded with luminal contents, the bowel attempts to overcome the obstruction by contracting, which leads to colicky abdominal pain. Tumor burden can also damage the intestinal epithelium and cause continuous pain.

The buildup of secretions can lead to translocation of fluid into the peritoneum (“third spacing”), bowel ischemia, perforation, or sepsis. The combination of poor oral intake, gastrointestinal fluid loss, and sequestration can lead to profound dehydration on presentation.^2,7

Fluid resuscitation, electrolyte repletion, and a trial of NG tube decompression are part of the initial management of MBO (Figure ). While studies have shown that moderate intravenous hydration can minimize nausea and drowsiness, excessive fluids may worsen bowel edema and exacerbate vomiting.^1,8 NG tube decompression is most effective in patients with proximal obstructions but some studies suggest it can decrease vomiting in patients with colonic obstructions as well.⁹ Computed tomography imaging can identify the extent of the tumor, the transition point of the obstruction, and any distant metastases. Surgery, Gastroenterology, and/or Interventional Radiology consultation should be obtained early to evaluate options for direct decompression. Hematology/Oncology and Radiation/Oncology referral may help delineate prognosis and achievable outcomes. Emergent exploratory surgery may be required in cases of bowel perforation or ischemia. Otherwise, a planned surgical resection should be considered in those with an isolated resectable lesion and acceptable perioperative risk. Colorectal or duodenal stents may be an option for those who are not surgical candidates or as a bridge to surgery.

As bowel obstruction is often a late manifestation of advanced malignancy, many patients may not be appropriate candidates for operative/interventional treatment due to malnutrition, comorbid conditions, or anatomic considerations. For these individuals, pharmacologic management is the mainstay of treatment. Additionally, the pharmacologic approaches detailed below may provide benefit as adjunctive therapy for patients undergoing procedural intervention.⁷ Consultation for early palliative care can improve symptom control as well as clarify goals of care.

Given the pathophysiology of MBO, pharmacologic therapies are focused on controlling nausea and pain while reducing bowel edema and secretions.

Antiemetic Agents

Nausea and vomiting in MBO are due to activation of vagal nerve fibers in the gastric wall and stimulation of the chemoreceptor trigger zone (CTZ).¹⁰ Dopamine antagonists have started to gain favor for MBO compared to more commonly used antiemetics such as the serotonin antagonists. Haloperidol should be considered as a first-line antiemetic in patients with MBO. Its potent D₂-receptor antagonistic properties block receptors in the CTZ. The high affinity of the drug for only the D₂-receptor makes it preferable to alternative agents in the same class such as chlorpromazine. However, haloperidol may cause or worsen QT prolongation and should be avoided in patients with Parkinson’s disease. The medication has less sedative and unwanted anticholinergic side effects due to its limited interaction with histaminergic and acetylcholine receptors.¹¹ Haloperidol has been shown in the past to be efficacious for post-operative nausea but there are few randomized controlled trials in the terminally ill.¹² Nonetheless, recent consensus guidelines from the Multinational Association of Supportive Care recommended haloperidol as the initial treatment of nausea for individuals with MBO based on available systematic reviews.¹⁰

Other dopamine antagonists remain good options, though they may cause additional side effects due to actions on other receptor types. Metoclopramide, another D₂-receptor antagonist, has been shown to be effective in the treatment of nausea and vomiting due to advanced cancer.¹³ However as a prokinetic agent, this medication should be avoided in those with complete MBO and only considered in those with partial MBO.^10,14

Olanzapine, an atypical antipsychotic, may also have a role in controlling nausea in patients with MBO. It functions as a 5-HT_2A and D₂-receptor antagonist, with a slightly greater affinity for the 5-HT_2A receptor. Olanzapine thus can target two critical receptors playing a role in nausea and vomiting. A study of patients with incomplete bowel obstruction found the addition of olanzapine significantly decreased nausea and vomiting in patients who were refractory to other treatments including steroids and haloperidol.¹⁵ Olanzapine has the added advantage of single-day dosing as well as an oral disintegrating formulation.¹⁶

Intravenous and sublingual preparations of 5-HT₃ receptor antagonists such as ondansetron are commonly used in the inpatient setting. These medications are potent antiemetics that exhibit their effects via pathways where serotonin acts as a neurotransmitter.¹⁷ An alternative agent, tropisetron, has shown promise when used alone or in conjunction with metoclopramide but is not currently available in the US.¹⁸ Granisetron is available in a transdermal formulation, which can be very convenient for patients with bowel obstruction. Its mechanism of action differs from ondansetron as it is an allosteric inhibitor rather than a competitive inhibitor.¹⁹ Granisetron needs more specific study with regards to its role in MBO.

Although haloperidol remains the initial choice, combination therapy can help to decrease the risk of extrapyramidal symptoms seen with higher doses of dopaminergic monotherapy.

Analgesics

Pain control is an essential part of the palliative treatment of MBO as bowel distention, secretions, and edema can cause rapid onset of pain. Parenteral step three opioids remain the optimal initial choice since patients are unable to take medications orally and may have compromised absorption. Opioids address both the colicky and continuous aspects of MBO pain.

Short-acting intravenous opioids such as morphine or hydromorphone may be scheduled every four hours with breakthrough dosing every hour in between. Alternatively, analgesics can be administered via a patient-controlled analgesia (PCA) pump.¹ Although doses vary across patients, opioid-naïve patients can be initiated on a low dose therapy such as hydromorphone 0.2 mg IV/SC or morphine 1 mg IV/SC every four hours as needed for pain control.

Ongoing pain management for patients with MBO requires coordination of care. Many patients will elect to receive hospice care following discharge. Direct communication with palliative consultants and hospice providers can help facilitate a smooth transition. In patients for whom bowel obstruction resolves, transition to oral opioids based on morphine equivalent daily dose is indicated with further dose adjustment as patients may have reduced pain at this stage.

Options for patients with unresolved obstruction include transdermal and sublingual preparations as well as outpatient PCA with hospice support. Transdermal fentanyl patch can be useful but onset of peak levels occur within 8-12 hours.²⁰ The patch is usually exchanged every 72 hours and is most effective when applied to areas containing adipose tissue which may limit its use in cachectic patients. The liquid preparation of methadone can be useful even in patients with unresolved MBO. Its lipophilic properties allow for ease of absorption.²¹ A baseline electrocardiogram (EKG) is recommended prior to methadone initiation due to the potential for QT prolongation. Methadone should not be a first-line option for opioid-naïve individuals due to its longer onset of action which limits rapid dose titration. Close collaboration with palliative medicine is highly recommended when using longer acting opioids.

Antisecretory Agents

Antisecretory agents are a mainstay of the pharmacologic management of inoperable MBO. Medications that reduce secretions and bowel edema include: somatostatin analogs, H2-blockers, proton pump inhibitors (PPIs), steroids, and anticholinergic agents. Table 2 summarizes the major studies comparing various antisecretory medications.

Octreotide, a somatostatin analog, has been increasingly used for the palliative treatment of MBO. The mechanism of action involves splanchnic vasoconstriction, reduction of intestinal and pancreatic secretions (via inhibition of VIP), decrease in gastric emptying, and slowing of smooth muscle contractions.²² Octreotide comes in an immediate-release formulation with an initial subcutaneous dose of 100 µg three or four times per day. Most patients will require 300-800 µg/day, maximum dose being up to 1 mg/day.^22,23 A long-acting formulation, lanreotide, exists but can be difficult to obtain and may not provide the immediate relief needed in an acute care setting.

Initiation of octreotide should be considered in the presence of persistent symptoms. Studies have suggested that the benefit of octreotide is most apparent in the first three days of treatment (range 1-5 days).^6,22,24 The medication should be discontinued if there is no clinical improvement such as reduction of NG tube output. Octreotide has been shown to be more efficacious than anticholinergic agents in reducing secretions as well as frequency of nausea and vomiting.^8,25-28 Octreotide expedites NG tube removal, recovery of bowel function, and improvement in quality of life.^29-32 The medication should also be considered in cases of recurrent MBO that previously responded to the medication.

Octreotide is considered the first-line agent in the palliative treatment of MBO, however the medication is costly. Recent studies suggest combination therapy with steroids and H2-blockers or PPIs may be an equally effective and less expensive alternative. The primary rationale for the use of steroids in MBO is their ability to decrease peritumoral edema and promote salt and water absorption from the intestine.^1,2 PPIs and H2-blockers decrease distension, pain, and vomiting by reducing the volume of gastric secretions.³³ A recent meta-analysis of phase 3 trials found both PPIs and H2-blockers to be effective in lowering volumes of gastric aspirates with ranitidine being slightly superior.³⁴

Initial research into the utility of steroids in MBO garnered mixed results. One study showed marginal benefit for steroid plus octreotide combination therapy compared to octreotide, in a cohort of 27 patients.³⁵ A subsequent review of practice patterns in the management of terminal MBO in Japan found that patients given steroids in combination with octreotide compared to octreotide alone were more likely to undergo early NG tube removal.³⁶ A 1999 systematic review of corticosteroid treatment of MBO concluded low morbidity associated with the medications with a trend toward benefit that was not statistically significant.³⁷ A 2015 study by Currow showed the addition of octreotide in patients already on a regime of dexamethasone and ranitidine did not improve the number of days free from vomiting but did reduce vomiting episodes in those with the most refractory symptoms.³⁸

Collectively, the studies suggest that combination therapy with steroid and PPI or H2 blocker could be a less expensive option in the initial management of MBO. Alternatively, steroids may provide additional relief in patients with continued symptoms on octreotide and H2-blockers. Dexamethasone is preferable given its longer half-life and decreased propensity for sodium retention. Dosing of dexamethasone should be 8 mg IV once a day.³⁸

Anticholinergic agents also reduce secretions. However, they are considered second-line therapy given their lower efficacy compared to other treatment options as well as their propensity to worsen cognitive function.^1,2 Anticholinergics may benefit patients with continued symptoms who cannot tolerate the side effects of other treatments. Scopolamine, also known as hyoscine hydrobromide in the US, should be avoided as it crosses the blood-brain barrier. The quaternary formulation, scopolamine butylbromide (hyoscine butylbromide), does not pass this barrier but is currently not available in the US. Glycopyrrolate may be considered as it is also a quaternary ammonium compound that does not cross the blood-brain barrier. Several case reports have described its effectiveness in the resolution of refractory nausea and vomiting in combination with haloperidol and hydromorphone for symptom control.³⁹ Effective oral care is imperative if anticholinergics are used in order to prevent the unpleasant feeling of dry mouth.

While initial management of MBO often requires placement of an NG tube, prolonged placement can increase the risk for erosions, aspiration, and sinus infections. Removal of the NG tube is most successful when secretions are minimal, but this may not happen unless the obstruction resolves. Some patients may elect to keep an NG tube if symptoms cannot be otherwise controlled by medications.

A venting gastrostomy tube can be considered as an alternative to prolonged NG tube placement. The tube may help alleviate distressing symptoms and can enhance the quality of life of patients by allowing the sensation of oral intake, though it will not allow for absorption of nutrients.⁴⁰ Although a low risk procedure, patients may be too frail to undergo the procedure and may have postprocedure pain and complications. Anatomic abnormalities such as overlying bowel may also prevent the noninvasive percutaneous approach.

In patients with unresolved obstruction, oral intake should be reinitiated with caution with the patient’s wishes taken into account at all times. Some patients may prioritize the comfort derived from eating small amounts over any associated risks of increased nausea and vomiting.

Parenteral nutrition should be avoided in those with inoperable MBO in the advanced stages. The risks of infection, refeeding syndrome, and the discomfort of an intravenous line and intermittent testing may outweigh any benefits given the overall prognosis.^41,42

Hospitalists are often involved in the initial care of patients with advanced malignancy who present with MBO. When interventions or surgeries to directly alleviate the obstruction are not possible, pharmacologic options are essential in managing burdensome symptoms and improving quality of life. Early Palliative Care referral can also assist with symptom management, emotional support, clarification of goals of care, and transition to the outpatient setting. While patients with inoperable MBO have a poor prognosis, hospitalists can play a vital role in alleviation of suffering in this devastating complication of advanced cancer.

Disclosures

The authors have nothing to disclose.

Hospital medicine continues to expand with respect to the number of practitioners as well as the scope of the practice of those practitioners. In addition, the commitment to, and rigor of, scientific inquiry in the field continues to grow. The authors of this article conducted a review of the medical literature, including articles published between March 2017 and March 2018. The key articles reported studies with high methodological quality, clear findings, and a high potential for impact on clinical practice. The literature was independently reviewed by each author, and candidate works were chosen on the basis of relevance to hospital medicine and expected clinical impact. The articles were organized by subject matter, ranked by applicability to the audience, and selected to meet the time constraints of each talk. Twenty-nine articles were presented at the Update in Hospital Medicine at the 2018 Society of Hospital Medicine and Society of General Internal Medicine annual meetings (B Sharpe, A Burger at SGIM and B Slawski, C Cooper at SHM). Nine articles were included in this review through an iterative voting process. Each author ranked their top five articles from one to five. Points were tallied for each article, and the five articles with the highest points were included. A second round of voting identified the remaining four articles for inclusion. Ties were adjudicated by group discussion. Each article is summarized below, and their key points are highlighted in the table.

Aspirin in Patients with Previous Percutaneous Coronary Intervention Undergoing Noncardiac Surgery. Graham MM et al. Ann Intern Med. 2018;168(4):237-244.¹

Background

The Perioperative Ischemic Evaluation 2 (POISE-2) trial found that perioperative aspirin use had no significant effect on the risk of perioperative death and nonfatal myocardial infarction (MI) in patients who are at risk for vascular complications; however, the risk of major bleeding increased with aspirin use in these patients.² Nevertheless, the POISE-2 trial did not specifically address the role of aspirin in patients who had undergone previous percutaneous coronary intervention (PCI).

Methods

A post hoc subgroup analysis of POISE-2 evaluated 470 PCI patients (234 aspirin-treated and 236 placebo-treated patients) aged >45 years, 90% of whom had stents. The administration of the study drug was initiated within four hours preoperatively and continued postoperatively. Patients who had bare metal stents placed within the six weeks prior to the study or drug-eluting stents placed within the preceding 12 months were excluded.

Findings

The composite endpoint of risk of death and nonfatal MI was 11.5% in the placebo group and 6% in aspirin-treated patients (HR 0.50; CI, 0.26-0.95). Most of the difference in primary outcome was attributed to an increase in nonfatal MI in the placebo group. Major and life-threatening bleeding were not substantially increased in PCI patients but increased in the overall POISE-2 trial (absolute risk increase 0.8% for major bleeding [95% CI, 0.1%-1.6%]; HR 1.22 [95% CI, 1.01-1.48]). Stent type had no effect on death and nonfatal MI.

Cautions

This was a non-prespecified subgroup analysis with a small sample size.

Implications

Perioperative aspirin use in patients with previous PCI appears to provide more benefit than harm, unless a substantial bleeding risk exists.

Association Between Wait Time and 30-Day Mortality in Adults Undergoing Hip Fracture Surgery. Pincus D et al. JAMA. 2017;318(20):1994-2003.³

Background

Wait times to hip fracture surgery have been associated with mortality in previous studies; however, the wait time associated with complications remains controversial.^4,5

Methods

This retrospective cohort study of 42,230 adults modeled the probability of complications in accordance with wait time from hospital arrival to hip fracture surgery. It aimed to identify the optimal time window in which to conduct surgery before complications increased. This window to increased complications was used to define early and delayed surgery. The matched cohorts of early and delayed patients were then used to compare outcomes.

Findings

Overall 30-day mortality was 7%. Complication rates increased when wait times reached 24 hours. Comparing the propensity-matched early (<24 hours) and late (>24 hours) surgery patients revealed that late surgery patients had significantly higher 30-day mortality (6.5% vs 5.8%; % absolute RD 0.79; 95% CI, 0.23-1.35) than early surgery patients and the composite outcome of mortality or other medical complications (MI, DVT, PE, and pneumonia; 12.2% vs 10.1%; % absolute RD 2.16; 95% CI, 1.43-2.89).

Cautions

Only 34% of patients in this study had surgery within 24 hours. The observational cohort study design may result in unmeasured confounders, eg, less sick patients go to surgery more quickly than sicker patients.

Implications

A preoperative wait time of 24 hours appears to represent a threshold of increased risk for 30-day perioperative complications and mortality in hip fracture surgery.

When are Oral Antibiotics a Safe and Effective Choice for Bacterial Bloodstream Infections? An Evidence-Based Narrative Review. Hale AJ et al. J Hosp Med. 2018;13(5):328-335.⁶

Background

Bloodstream infections (BSIs) are significant causes of morbidity and mortality in the United States. Traditionally, clinicians have relied on intravenous antibiotics for treatment. A recent “Choosing Wisely^®” initiative recommends that clinicians should use “oral formulations of highly bioavailable antimicrobials wherever possible.”⁷ Thus, the authors searched for evidence for scenarios wherein BSIs could be safely treated with oral antibiotics.

Methods

A narrative review was conducted given that robust clinical data for an extensive systematic review were insufficient.

Findings

Key decision points on the use of an oral antibiotic for a diagnosed BSI are as follows: (1) Source control must be attained prior to the consideration of oral antibiotics. (2) A highly bioavailable oral option to which the pathogen is sensitive must be available. (3) Patients must be able to comply with the therapy for the full course and not be on interfering medications. Good evidence for use of oral antibiotics against sensitive gram-negative bacilli other than Pseudomonas exists. Evidence for treating Streptococcus pneumoniae with early transition (within three days) to oral antibiotics is robust when treating bacteremia and pneumonia but not for other primary sites of infection. Evidence for the use of oral antibiotics for B-hemolytic streptococcus, including necrotizing fasciitis and Enterococcus, is insufficient. The evidence supports at least two weeks of IV antibiotics for the treatment of Staphylococcus aureus.

Cautions

This is a narrative review due to limited evidence.

Implications

The early use of oral antibiotics in the setting of bacteremia may be appropriate in select clinical situations.

Prevalence of Pulmonary Embolism in Patients with Syncope. Costantino et al. JAMA Intern Med. 2018;178(3):356-362.⁸

Background

Data on the prevalence of pulmonary embolism in patients presenting with syncope are conflicting.

Methods

This was a retrospective observational study involving five databases in four countries of >1.6 million adults identified through syncope ICD codes. The rates of pulmonary embolism at first evaluation and pulmonary embolism or venous thromboembolism within 90 days were calculated for emergency room patients and a hospitalized subgroup.

Findings

Pulmonary embolism was rare in patients with syncope, eg, less than 3% for hospitalized patients in this database study.

Cautions

The results of this study are based on the use of administrative databases to confirm the diagnosis of syncope. Additionally, the results include hospitalized and nonhospitalized patients. The design of this study differs significantly from those of the PESIT study, which showed a prevalence of 17% in hospitalized patients.⁹ The PESIT study specifically sought the diagnosis of pulmonary embolism even when other etiologies for syncope existed.

Implications

Ultimately, the clinical impetus to search for pulmonary embolism in hospitalized patients admitted with syncope will depend on individual presentations. The authors argued that pulmonary embolism is rare in syncope and much lower than 17% but should be considered in appropriate patients.

Balanced Crystalloids versus Saline in Noncritically Ill Patients. Self WH et al. N Engl J Med. 2018;378(9):819-828.¹⁰

Background

Data on the optimal composition of intravenous fluids (IVF) are limited. Limited experimental evidence suggests that IVF-induced hyperchloremia results in renal vasoconstriction and acute kidney injury.

Methods

This was a single-center, open-label, multiple crossover trial of >13,000 non-ICU hospitalized patients admitted from the Emergency Department. Patients were randomized to receive either only normal saline or a “balanced crystalloid,” eg, either Lactated Ringer’s or Plasmalyte. The primary outcome was hospital-free days. Secondary outcomes were major adverse kidney events (MAKE) at 30 days.

Findings

The study found no difference in the primary outcome of hospital-free days. However, balanced IVF resulted in a lower incidence of hyperchloremia and a slightly reduced incidence of MAKE 30 (4.7% vs 5.6%; adjusted OR 0.82).

Cautions

The incidence of acute kidney injury was low in this single-center ED population. This study, however, did not include hospitalized patients. The long-term effects on renal function could not be ascertained.

Implications

Hospital-free days after inpatient randomization to either normal saline or “balanced IVF” were not significantly different. “Balanced IVF” may be beneficial in select high renal-risk populations.

Speaker Introductions at Internal Medicine Grand Rounds: Forms of Address Reveal Speaker Bias. Files et al. J Womens Health. 2017;26(5):413-419.¹¹

Background

Gender bias is known to contribute to leadership disparities between men and women in several academic medical centers.

Methods

This was a retrospective observational study reviewing video-archived introductions at Internal Medicine Grand Rounds at two connected institutions. All speakers had doctoral degrees. The outcome measured was the use of a speaker’s professional title during his/her introduction as a function of the introducer’s gender.

Findings

Women were more likely than men to introduce speakers of any gender by their professional title in the 321 forms of address analyzed (96% vs 66%, P < .001). When the introducer and speaker were of different genders, women were more likely to introduce male speakers with formal titles than men introducing female speakers (95% vs 49%, P < .001).

Cautions

This study was done at two associated academic institutions and may not reflect the practice or customs of physicians in other departments or institutions.

Implications

Despite the study’s limitations, it supports a theme of prevalent gender bias within academic medical institutions that may affect the outcomes of leadership, promotion, and scholarship.

Edoxaban for the Treatment of Cancer-Associated Venous Thromboembolism. Raskob GE et al. N Engl J Med. 2018;378(7):615-624.¹²

Background

Low-molecular-weight heparin (LMWH) is the standard of care for the treatment of venous thromboembolism (VTE) in patients with cancer. Direct oral anticoagulants have not been studied for this indication.

Methods

This open-label, noninferiority trial randomized patients with cancer and acute VTE to either LMWH for a minimum of five days followed by oral edoxaban vs subcutaneous dalteparin.

Findings

A total of 1,046 patients were included in the modified intention-to-treat analysis. Patients received treatment for six to twelve months total. A composite outcome of recurrent VTE or major bleed within 12 months occurred in 67 of 522 (12.8%) of patients in the edoxaban group vs 71 of 524 (13.5%) of patients in the dalteparin group (HR 0.91, 95% CI 0.70-1.36, P = .006 for noninferiority). Recurrent VTE occurred more commonly with dalteparin than with edoxaban (11.3% vs 7.9%), whereas major bleeding was less common with dalteparin than with edoxaban (4% vs 6.9%). The increased bleeding rate with edoxaban was predominantly in patients with an upper gastrointestinal (GI) malignancy.

Cautions

This was an open-label study. Patients in the edoxaban still received five days of LMWH prior to oral edoxaban. More patients in the edoxaban group continued treatment for the entire 12-month period, which contributes to the observed decreased bleeding and increased VTE rates in the dalteparin group.

Implications

Oral edoxaban is noninferior to subcutaneous dalteparin for the primary composite endpoint of VTE and bleeding. Notably, the patients in the edoxaban group experienced a lower rate of recurrent VTE and a higher rate of major bleeding than the patients in the dalteparin group. Additional caution about bleeding risk in those with a GI malignancy is recommended.

Can High-flow Nasal Cannula Reduce the Rate of Endotracheal Intubation in Adult Patients with Acute Respiratory Failure Compared with Conventional Oxygen Therapy and Noninvasive Positive Pressure Ventilation? Ni Y-N et al. Chest. 2017;151(4):764-775.¹³

Background

High-flow nasal cannula (HFNC) can deliver heated and humidified oxygen at rates of up to 60 L/min. Evidence on the benefits of HFNC over usual oxygen therapy or noninvasive positive pressure ventilation (NIPPV) is conflicting.

Methods

This systematic review and meta-analysis included 18 studies (12 RCTs, four retrospective, and two prospective cohort studies) with 3,881 patients with respiratory failure (medical and surgical causes). The included studies compared HFNC with usual oxygen therapy or NIPPV.

Findings

HFNC was associated with lower rates of endotracheal intubation (OR 0.47, 95% CI 0.27-0.84, P = .01) relative to oxygen therapy. Intubation rates did not differ between HFNC and NIPPV (OR 0.73, 95% CI 0.47-1.13, P = .16). No differences in ICU mortality or ICU length of stay (LOS) were found when HFNC was compared with either usual oxygen therapy or NIPPV.

Cautions

The significant heterogeneity in study design across studies is mainly attributable to varying causes of respiratory failure and differences in flow rate, oxygen concentration, and treatment duration across studies.

Implications

In patients with respiratory failure, HFNC may reduce intubation when compared with usual oxygen therapy and has similar ICU mortality when compared with usual oxygen and NIPPV.

Errors in the Diagnosis of Spinal Epidural Abscesses in the Era of Electronic Health Records. Bhise V et al. Am J Med. 2017;130(8):975-981.¹⁴

Background

Diagnostic errors are common in patients with spinal epidural abscess, but the main contributing factors are unclear.¹⁵

Methods

All patients who were newly diagnosed with spinal epidural abscess in 2013 were identified from the Veterans Affairs (VA) national database. Charts were reviewed for diagnostic delay and contributing factors, including the presence of “red flag” symptoms (eg, fever and neurological deficits).

Findings

Of the 119 patients with a new diagnosis of spinal epidural abscess, 66 (56%) had a diagnostic error. The median time to diagnosis in those with a diagnostic error was 12 days vs four days in those without error (P < .01). Common missed red flags in error cases included fever (n = 57, 86.4%), focal neurologic deficit (n = 54, 81.8%), and active infection (n = 54, 81.8%). Most errors occurred during the provider–patient encounter (eg, information not gathered during the history or physical). The magnitude of harm was serious for most patients (n = 40, 60.6%) and contributed to death in eight patients (12.1%).

Cautions

The study may not be generalizable because it was limited to the VA health system.

Implications

Diagnostic errors are common in patients with spinal epidural abscesses and can lead to serious harm. Health systems should build mechanisms to support providers in the evaluation of patients with back pain.

Hospital medicine continues to expand with respect to the number of practitioners as well as the scope of the practice of those practitioners. In addition, the commitment to, and rigor of, scientific inquiry in the field continues to grow. The authors of this article conducted a review of the medical literature, including articles published between March 2017 and March 2018. The key articles reported studies with high methodological quality, clear findings, and a high potential for impact on clinical practice. The literature was independently reviewed by each author, and candidate works were chosen on the basis of relevance to hospital medicine and expected clinical impact. The articles were organized by subject matter, ranked by applicability to the audience, and selected to meet the time constraints of each talk. Twenty-nine articles were presented at the Update in Hospital Medicine at the 2018 Society of Hospital Medicine and Society of General Internal Medicine annual meetings (B Sharpe, A Burger at SGIM and B Slawski, C Cooper at SHM). Nine articles were included in this review through an iterative voting process. Each author ranked their top five articles from one to five. Points were tallied for each article, and the five articles with the highest points were included. A second round of voting identified the remaining four articles for inclusion. Ties were adjudicated by group discussion. Each article is summarized below, and their key points are highlighted in the table.

Aspirin in Patients with Previous Percutaneous Coronary Intervention Undergoing Noncardiac Surgery. Graham MM et al. Ann Intern Med. 2018;168(4):237-244.¹

Background

The Perioperative Ischemic Evaluation 2 (POISE-2) trial found that perioperative aspirin use had no significant effect on the risk of perioperative death and nonfatal myocardial infarction (MI) in patients who are at risk for vascular complications; however, the risk of major bleeding increased with aspirin use in these patients.² Nevertheless, the POISE-2 trial did not specifically address the role of aspirin in patients who had undergone previous percutaneous coronary intervention (PCI).

Methods

A post hoc subgroup analysis of POISE-2 evaluated 470 PCI patients (234 aspirin-treated and 236 placebo-treated patients) aged >45 years, 90% of whom had stents. The administration of the study drug was initiated within four hours preoperatively and continued postoperatively. Patients who had bare metal stents placed within the six weeks prior to the study or drug-eluting stents placed within the preceding 12 months were excluded.

Findings

The composite endpoint of risk of death and nonfatal MI was 11.5% in the placebo group and 6% in aspirin-treated patients (HR 0.50; CI, 0.26-0.95). Most of the difference in primary outcome was attributed to an increase in nonfatal MI in the placebo group. Major and life-threatening bleeding were not substantially increased in PCI patients but increased in the overall POISE-2 trial (absolute risk increase 0.8% for major bleeding [95% CI, 0.1%-1.6%]; HR 1.22 [95% CI, 1.01-1.48]). Stent type had no effect on death and nonfatal MI.

Cautions

This was a non-prespecified subgroup analysis with a small sample size.

Implications

Perioperative aspirin use in patients with previous PCI appears to provide more benefit than harm, unless a substantial bleeding risk exists.

Association Between Wait Time and 30-Day Mortality in Adults Undergoing Hip Fracture Surgery. Pincus D et al. JAMA. 2017;318(20):1994-2003.³

Background

Wait times to hip fracture surgery have been associated with mortality in previous studies; however, the wait time associated with complications remains controversial.^4,5

Methods

This retrospective cohort study of 42,230 adults modeled the probability of complications in accordance with wait time from hospital arrival to hip fracture surgery. It aimed to identify the optimal time window in which to conduct surgery before complications increased. This window to increased complications was used to define early and delayed surgery. The matched cohorts of early and delayed patients were then used to compare outcomes.

Findings

Overall 30-day mortality was 7%. Complication rates increased when wait times reached 24 hours. Comparing the propensity-matched early (<24 hours) and late (>24 hours) surgery patients revealed that late surgery patients had significantly higher 30-day mortality (6.5% vs 5.8%; % absolute RD 0.79; 95% CI, 0.23-1.35) than early surgery patients and the composite outcome of mortality or other medical complications (MI, DVT, PE, and pneumonia; 12.2% vs 10.1%; % absolute RD 2.16; 95% CI, 1.43-2.89).

Cautions

Only 34% of patients in this study had surgery within 24 hours. The observational cohort study design may result in unmeasured confounders, eg, less sick patients go to surgery more quickly than sicker patients.

Implications

A preoperative wait time of 24 hours appears to represent a threshold of increased risk for 30-day perioperative complications and mortality in hip fracture surgery.

When are Oral Antibiotics a Safe and Effective Choice for Bacterial Bloodstream Infections? An Evidence-Based Narrative Review. Hale AJ et al. J Hosp Med. 2018;13(5):328-335.⁶

Background

Bloodstream infections (BSIs) are significant causes of morbidity and mortality in the United States. Traditionally, clinicians have relied on intravenous antibiotics for treatment. A recent “Choosing Wisely^®” initiative recommends that clinicians should use “oral formulations of highly bioavailable antimicrobials wherever possible.”⁷ Thus, the authors searched for evidence for scenarios wherein BSIs could be safely treated with oral antibiotics.

Methods

A narrative review was conducted given that robust clinical data for an extensive systematic review were insufficient.

Findings

Key decision points on the use of an oral antibiotic for a diagnosed BSI are as follows: (1) Source control must be attained prior to the consideration of oral antibiotics. (2) A highly bioavailable oral option to which the pathogen is sensitive must be available. (3) Patients must be able to comply with the therapy for the full course and not be on interfering medications. Good evidence for use of oral antibiotics against sensitive gram-negative bacilli other than Pseudomonas exists. Evidence for treating Streptococcus pneumoniae with early transition (within three days) to oral antibiotics is robust when treating bacteremia and pneumonia but not for other primary sites of infection. Evidence for the use of oral antibiotics for B-hemolytic streptococcus, including necrotizing fasciitis and Enterococcus, is insufficient. The evidence supports at least two weeks of IV antibiotics for the treatment of Staphylococcus aureus.

Cautions

This is a narrative review due to limited evidence.

Implications

The early use of oral antibiotics in the setting of bacteremia may be appropriate in select clinical situations.

Prevalence of Pulmonary Embolism in Patients with Syncope. Costantino et al. JAMA Intern Med. 2018;178(3):356-362.⁸

Background

Data on the prevalence of pulmonary embolism in patients presenting with syncope are conflicting.

Methods

This was a retrospective observational study involving five databases in four countries of >1.6 million adults identified through syncope ICD codes. The rates of pulmonary embolism at first evaluation and pulmonary embolism or venous thromboembolism within 90 days were calculated for emergency room patients and a hospitalized subgroup.

Findings

Pulmonary embolism was rare in patients with syncope, eg, less than 3% for hospitalized patients in this database study.

Cautions

The results of this study are based on the use of administrative databases to confirm the diagnosis of syncope. Additionally, the results include hospitalized and nonhospitalized patients. The design of this study differs significantly from those of the PESIT study, which showed a prevalence of 17% in hospitalized patients.⁹ The PESIT study specifically sought the diagnosis of pulmonary embolism even when other etiologies for syncope existed.

Implications

Ultimately, the clinical impetus to search for pulmonary embolism in hospitalized patients admitted with syncope will depend on individual presentations. The authors argued that pulmonary embolism is rare in syncope and much lower than 17% but should be considered in appropriate patients.

Balanced Crystalloids versus Saline in Noncritically Ill Patients. Self WH et al. N Engl J Med. 2018;378(9):819-828.¹⁰

Background

Data on the optimal composition of intravenous fluids (IVF) are limited. Limited experimental evidence suggests that IVF-induced hyperchloremia results in renal vasoconstriction and acute kidney injury.

Methods

This was a single-center, open-label, multiple crossover trial of >13,000 non-ICU hospitalized patients admitted from the Emergency Department. Patients were randomized to receive either only normal saline or a “balanced crystalloid,” eg, either Lactated Ringer’s or Plasmalyte. The primary outcome was hospital-free days. Secondary outcomes were major adverse kidney events (MAKE) at 30 days.

Findings

The study found no difference in the primary outcome of hospital-free days. However, balanced IVF resulted in a lower incidence of hyperchloremia and a slightly reduced incidence of MAKE 30 (4.7% vs 5.6%; adjusted OR 0.82).

Cautions

The incidence of acute kidney injury was low in this single-center ED population. This study, however, did not include hospitalized patients. The long-term effects on renal function could not be ascertained.

Implications

Hospital-free days after inpatient randomization to either normal saline or “balanced IVF” were not significantly different. “Balanced IVF” may be beneficial in select high renal-risk populations.

Speaker Introductions at Internal Medicine Grand Rounds: Forms of Address Reveal Speaker Bias. Files et al. J Womens Health. 2017;26(5):413-419.¹¹

Background

Gender bias is known to contribute to leadership disparities between men and women in several academic medical centers.

Methods

This was a retrospective observational study reviewing video-archived introductions at Internal Medicine Grand Rounds at two connected institutions. All speakers had doctoral degrees. The outcome measured was the use of a speaker’s professional title during his/her introduction as a function of the introducer’s gender.

Findings

Women were more likely than men to introduce speakers of any gender by their professional title in the 321 forms of address analyzed (96% vs 66%, P < .001). When the introducer and speaker were of different genders, women were more likely to introduce male speakers with formal titles than men introducing female speakers (95% vs 49%, P < .001).

Cautions

This study was done at two associated academic institutions and may not reflect the practice or customs of physicians in other departments or institutions.

Implications

Despite the study’s limitations, it supports a theme of prevalent gender bias within academic medical institutions that may affect the outcomes of leadership, promotion, and scholarship.

Edoxaban for the Treatment of Cancer-Associated Venous Thromboembolism. Raskob GE et al. N Engl J Med. 2018;378(7):615-624.¹²

Background

Low-molecular-weight heparin (LMWH) is the standard of care for the treatment of venous thromboembolism (VTE) in patients with cancer. Direct oral anticoagulants have not been studied for this indication.

Methods

This open-label, noninferiority trial randomized patients with cancer and acute VTE to either LMWH for a minimum of five days followed by oral edoxaban vs subcutaneous dalteparin.

Findings

A total of 1,046 patients were included in the modified intention-to-treat analysis. Patients received treatment for six to twelve months total. A composite outcome of recurrent VTE or major bleed within 12 months occurred in 67 of 522 (12.8%) of patients in the edoxaban group vs 71 of 524 (13.5%) of patients in the dalteparin group (HR 0.91, 95% CI 0.70-1.36, P = .006 for noninferiority). Recurrent VTE occurred more commonly with dalteparin than with edoxaban (11.3% vs 7.9%), whereas major bleeding was less common with dalteparin than with edoxaban (4% vs 6.9%). The increased bleeding rate with edoxaban was predominantly in patients with an upper gastrointestinal (GI) malignancy.

Cautions

This was an open-label study. Patients in the edoxaban still received five days of LMWH prior to oral edoxaban. More patients in the edoxaban group continued treatment for the entire 12-month period, which contributes to the observed decreased bleeding and increased VTE rates in the dalteparin group.

Implications

Oral edoxaban is noninferior to subcutaneous dalteparin for the primary composite endpoint of VTE and bleeding. Notably, the patients in the edoxaban group experienced a lower rate of recurrent VTE and a higher rate of major bleeding than the patients in the dalteparin group. Additional caution about bleeding risk in those with a GI malignancy is recommended.

Can High-flow Nasal Cannula Reduce the Rate of Endotracheal Intubation in Adult Patients with Acute Respiratory Failure Compared with Conventional Oxygen Therapy and Noninvasive Positive Pressure Ventilation? Ni Y-N et al. Chest. 2017;151(4):764-775.¹³

Background

High-flow nasal cannula (HFNC) can deliver heated and humidified oxygen at rates of up to 60 L/min. Evidence on the benefits of HFNC over usual oxygen therapy or noninvasive positive pressure ventilation (NIPPV) is conflicting.

Methods

This systematic review and meta-analysis included 18 studies (12 RCTs, four retrospective, and two prospective cohort studies) with 3,881 patients with respiratory failure (medical and surgical causes). The included studies compared HFNC with usual oxygen therapy or NIPPV.

Findings

HFNC was associated with lower rates of endotracheal intubation (OR 0.47, 95% CI 0.27-0.84, P = .01) relative to oxygen therapy. Intubation rates did not differ between HFNC and NIPPV (OR 0.73, 95% CI 0.47-1.13, P = .16). No differences in ICU mortality or ICU length of stay (LOS) were found when HFNC was compared with either usual oxygen therapy or NIPPV.

Cautions

The significant heterogeneity in study design across studies is mainly attributable to varying causes of respiratory failure and differences in flow rate, oxygen concentration, and treatment duration across studies.

Implications

In patients with respiratory failure, HFNC may reduce intubation when compared with usual oxygen therapy and has similar ICU mortality when compared with usual oxygen and NIPPV.

Errors in the Diagnosis of Spinal Epidural Abscesses in the Era of Electronic Health Records. Bhise V et al. Am J Med. 2017;130(8):975-981.¹⁴

Background

Diagnostic errors are common in patients with spinal epidural abscess, but the main contributing factors are unclear.¹⁵

Methods

All patients who were newly diagnosed with spinal epidural abscess in 2013 were identified from the Veterans Affairs (VA) national database. Charts were reviewed for diagnostic delay and contributing factors, including the presence of “red flag” symptoms (eg, fever and neurological deficits).

Findings

Of the 119 patients with a new diagnosis of spinal epidural abscess, 66 (56%) had a diagnostic error. The median time to diagnosis in those with a diagnostic error was 12 days vs four days in those without error (P < .01). Common missed red flags in error cases included fever (n = 57, 86.4%), focal neurologic deficit (n = 54, 81.8%), and active infection (n = 54, 81.8%). Most errors occurred during the provider–patient encounter (eg, information not gathered during the history or physical). The magnitude of harm was serious for most patients (n = 40, 60.6%) and contributed to death in eight patients (12.1%).

Cautions

The study may not be generalizable because it was limited to the VA health system.

Implications

Diagnostic errors are common in patients with spinal epidural abscesses and can lead to serious harm. Health systems should build mechanisms to support providers in the evaluation of patients with back pain.

Hospital medicine continues to expand with respect to the number of practitioners as well as the scope of the practice of those practitioners. In addition, the commitment to, and rigor of, scientific inquiry in the field continues to grow. The authors of this article conducted a review of the medical literature, including articles published between March 2017 and March 2018. The key articles reported studies with high methodological quality, clear findings, and a high potential for impact on clinical practice. The literature was independently reviewed by each author, and candidate works were chosen on the basis of relevance to hospital medicine and expected clinical impact. The articles were organized by subject matter, ranked by applicability to the audience, and selected to meet the time constraints of each talk. Twenty-nine articles were presented at the Update in Hospital Medicine at the 2018 Society of Hospital Medicine and Society of General Internal Medicine annual meetings (B Sharpe, A Burger at SGIM and B Slawski, C Cooper at SHM). Nine articles were included in this review through an iterative voting process. Each author ranked their top five articles from one to five. Points were tallied for each article, and the five articles with the highest points were included. A second round of voting identified the remaining four articles for inclusion. Ties were adjudicated by group discussion. Each article is summarized below, and their key points are highlighted in the table.

Aspirin in Patients with Previous Percutaneous Coronary Intervention Undergoing Noncardiac Surgery. Graham MM et al. Ann Intern Med. 2018;168(4):237-244.¹

Background

The Perioperative Ischemic Evaluation 2 (POISE-2) trial found that perioperative aspirin use had no significant effect on the risk of perioperative death and nonfatal myocardial infarction (MI) in patients who are at risk for vascular complications; however, the risk of major bleeding increased with aspirin use in these patients.² Nevertheless, the POISE-2 trial did not specifically address the role of aspirin in patients who had undergone previous percutaneous coronary intervention (PCI).

Methods

A post hoc subgroup analysis of POISE-2 evaluated 470 PCI patients (234 aspirin-treated and 236 placebo-treated patients) aged >45 years, 90% of whom had stents. The administration of the study drug was initiated within four hours preoperatively and continued postoperatively. Patients who had bare metal stents placed within the six weeks prior to the study or drug-eluting stents placed within the preceding 12 months were excluded.

Findings

The composite endpoint of risk of death and nonfatal MI was 11.5% in the placebo group and 6% in aspirin-treated patients (HR 0.50; CI, 0.26-0.95). Most of the difference in primary outcome was attributed to an increase in nonfatal MI in the placebo group. Major and life-threatening bleeding were not substantially increased in PCI patients but increased in the overall POISE-2 trial (absolute risk increase 0.8% for major bleeding [95% CI, 0.1%-1.6%]; HR 1.22 [95% CI, 1.01-1.48]). Stent type had no effect on death and nonfatal MI.

Cautions

This was a non-prespecified subgroup analysis with a small sample size.

Implications

Perioperative aspirin use in patients with previous PCI appears to provide more benefit than harm, unless a substantial bleeding risk exists.

Association Between Wait Time and 30-Day Mortality in Adults Undergoing Hip Fracture Surgery. Pincus D et al. JAMA. 2017;318(20):1994-2003.³

Background

Wait times to hip fracture surgery have been associated with mortality in previous studies; however, the wait time associated with complications remains controversial.^4,5

Methods

This retrospective cohort study of 42,230 adults modeled the probability of complications in accordance with wait time from hospital arrival to hip fracture surgery. It aimed to identify the optimal time window in which to conduct surgery before complications increased. This window to increased complications was used to define early and delayed surgery. The matched cohorts of early and delayed patients were then used to compare outcomes.

Findings

Overall 30-day mortality was 7%. Complication rates increased when wait times reached 24 hours. Comparing the propensity-matched early (<24 hours) and late (>24 hours) surgery patients revealed that late surgery patients had significantly higher 30-day mortality (6.5% vs 5.8%; % absolute RD 0.79; 95% CI, 0.23-1.35) than early surgery patients and the composite outcome of mortality or other medical complications (MI, DVT, PE, and pneumonia; 12.2% vs 10.1%; % absolute RD 2.16; 95% CI, 1.43-2.89).

Cautions

Only 34% of patients in this study had surgery within 24 hours. The observational cohort study design may result in unmeasured confounders, eg, less sick patients go to surgery more quickly than sicker patients.

Implications

A preoperative wait time of 24 hours appears to represent a threshold of increased risk for 30-day perioperative complications and mortality in hip fracture surgery.

When are Oral Antibiotics a Safe and Effective Choice for Bacterial Bloodstream Infections? An Evidence-Based Narrative Review. Hale AJ et al. J Hosp Med. 2018;13(5):328-335.⁶

Background

Bloodstream infections (BSIs) are significant causes of morbidity and mortality in the United States. Traditionally, clinicians have relied on intravenous antibiotics for treatment. A recent “Choosing Wisely^®” initiative recommends that clinicians should use “oral formulations of highly bioavailable antimicrobials wherever possible.”⁷ Thus, the authors searched for evidence for scenarios wherein BSIs could be safely treated with oral antibiotics.

Methods

A narrative review was conducted given that robust clinical data for an extensive systematic review were insufficient.

Findings

Key decision points on the use of an oral antibiotic for a diagnosed BSI are as follows: (1) Source control must be attained prior to the consideration of oral antibiotics. (2) A highly bioavailable oral option to which the pathogen is sensitive must be available. (3) Patients must be able to comply with the therapy for the full course and not be on interfering medications. Good evidence for use of oral antibiotics against sensitive gram-negative bacilli other than Pseudomonas exists. Evidence for treating Streptococcus pneumoniae with early transition (within three days) to oral antibiotics is robust when treating bacteremia and pneumonia but not for other primary sites of infection. Evidence for the use of oral antibiotics for B-hemolytic streptococcus, including necrotizing fasciitis and Enterococcus, is insufficient. The evidence supports at least two weeks of IV antibiotics for the treatment of Staphylococcus aureus.

Cautions

This is a narrative review due to limited evidence.

Implications

The early use of oral antibiotics in the setting of bacteremia may be appropriate in select clinical situations.

Prevalence of Pulmonary Embolism in Patients with Syncope. Costantino et al. JAMA Intern Med. 2018;178(3):356-362.⁸

Background

Data on the prevalence of pulmonary embolism in patients presenting with syncope are conflicting.

Methods

This was a retrospective observational study involving five databases in four countries of >1.6 million adults identified through syncope ICD codes. The rates of pulmonary embolism at first evaluation and pulmonary embolism or venous thromboembolism within 90 days were calculated for emergency room patients and a hospitalized subgroup.

Findings

Pulmonary embolism was rare in patients with syncope, eg, less than 3% for hospitalized patients in this database study.

Cautions

The results of this study are based on the use of administrative databases to confirm the diagnosis of syncope. Additionally, the results include hospitalized and nonhospitalized patients. The design of this study differs significantly from those of the PESIT study, which showed a prevalence of 17% in hospitalized patients.⁹ The PESIT study specifically sought the diagnosis of pulmonary embolism even when other etiologies for syncope existed.

Implications

Ultimately, the clinical impetus to search for pulmonary embolism in hospitalized patients admitted with syncope will depend on individual presentations. The authors argued that pulmonary embolism is rare in syncope and much lower than 17% but should be considered in appropriate patients.

Balanced Crystalloids versus Saline in Noncritically Ill Patients. Self WH et al. N Engl J Med. 2018;378(9):819-828.¹⁰

Background

Data on the optimal composition of intravenous fluids (IVF) are limited. Limited experimental evidence suggests that IVF-induced hyperchloremia results in renal vasoconstriction and acute kidney injury.

Methods

This was a single-center, open-label, multiple crossover trial of >13,000 non-ICU hospitalized patients admitted from the Emergency Department. Patients were randomized to receive either only normal saline or a “balanced crystalloid,” eg, either Lactated Ringer’s or Plasmalyte. The primary outcome was hospital-free days. Secondary outcomes were major adverse kidney events (MAKE) at 30 days.

Findings

The study found no difference in the primary outcome of hospital-free days. However, balanced IVF resulted in a lower incidence of hyperchloremia and a slightly reduced incidence of MAKE 30 (4.7% vs 5.6%; adjusted OR 0.82).

Cautions

The incidence of acute kidney injury was low in this single-center ED population. This study, however, did not include hospitalized patients. The long-term effects on renal function could not be ascertained.

Implications

Hospital-free days after inpatient randomization to either normal saline or “balanced IVF” were not significantly different. “Balanced IVF” may be beneficial in select high renal-risk populations.

Speaker Introductions at Internal Medicine Grand Rounds: Forms of Address Reveal Speaker Bias. Files et al. J Womens Health. 2017;26(5):413-419.¹¹

Background

Gender bias is known to contribute to leadership disparities between men and women in several academic medical centers.

Methods

This was a retrospective observational study reviewing video-archived introductions at Internal Medicine Grand Rounds at two connected institutions. All speakers had doctoral degrees. The outcome measured was the use of a speaker’s professional title during his/her introduction as a function of the introducer’s gender.

Findings

Women were more likely than men to introduce speakers of any gender by their professional title in the 321 forms of address analyzed (96% vs 66%, P < .001). When the introducer and speaker were of different genders, women were more likely to introduce male speakers with formal titles than men introducing female speakers (95% vs 49%, P < .001).

Cautions

This study was done at two associated academic institutions and may not reflect the practice or customs of physicians in other departments or institutions.

Implications

Despite the study’s limitations, it supports a theme of prevalent gender bias within academic medical institutions that may affect the outcomes of leadership, promotion, and scholarship.

Edoxaban for the Treatment of Cancer-Associated Venous Thromboembolism. Raskob GE et al. N Engl J Med. 2018;378(7):615-624.¹²

Background

Low-molecular-weight heparin (LMWH) is the standard of care for the treatment of venous thromboembolism (VTE) in patients with cancer. Direct oral anticoagulants have not been studied for this indication.

Methods

This open-label, noninferiority trial randomized patients with cancer and acute VTE to either LMWH for a minimum of five days followed by oral edoxaban vs subcutaneous dalteparin.

Findings

A total of 1,046 patients were included in the modified intention-to-treat analysis. Patients received treatment for six to twelve months total. A composite outcome of recurrent VTE or major bleed within 12 months occurred in 67 of 522 (12.8%) of patients in the edoxaban group vs 71 of 524 (13.5%) of patients in the dalteparin group (HR 0.91, 95% CI 0.70-1.36, P = .006 for noninferiority). Recurrent VTE occurred more commonly with dalteparin than with edoxaban (11.3% vs 7.9%), whereas major bleeding was less common with dalteparin than with edoxaban (4% vs 6.9%). The increased bleeding rate with edoxaban was predominantly in patients with an upper gastrointestinal (GI) malignancy.

Cautions

This was an open-label study. Patients in the edoxaban still received five days of LMWH prior to oral edoxaban. More patients in the edoxaban group continued treatment for the entire 12-month period, which contributes to the observed decreased bleeding and increased VTE rates in the dalteparin group.

Implications

Oral edoxaban is noninferior to subcutaneous dalteparin for the primary composite endpoint of VTE and bleeding. Notably, the patients in the edoxaban group experienced a lower rate of recurrent VTE and a higher rate of major bleeding than the patients in the dalteparin group. Additional caution about bleeding risk in those with a GI malignancy is recommended.

Can High-flow Nasal Cannula Reduce the Rate of Endotracheal Intubation in Adult Patients with Acute Respiratory Failure Compared with Conventional Oxygen Therapy and Noninvasive Positive Pressure Ventilation? Ni Y-N et al. Chest. 2017;151(4):764-775.¹³

Background

High-flow nasal cannula (HFNC) can deliver heated and humidified oxygen at rates of up to 60 L/min. Evidence on the benefits of HFNC over usual oxygen therapy or noninvasive positive pressure ventilation (NIPPV) is conflicting.

Methods

This systematic review and meta-analysis included 18 studies (12 RCTs, four retrospective, and two prospective cohort studies) with 3,881 patients with respiratory failure (medical and surgical causes). The included studies compared HFNC with usual oxygen therapy or NIPPV.

Findings

HFNC was associated with lower rates of endotracheal intubation (OR 0.47, 95% CI 0.27-0.84, P = .01) relative to oxygen therapy. Intubation rates did not differ between HFNC and NIPPV (OR 0.73, 95% CI 0.47-1.13, P = .16). No differences in ICU mortality or ICU length of stay (LOS) were found when HFNC was compared with either usual oxygen therapy or NIPPV.

Cautions

The significant heterogeneity in study design across studies is mainly attributable to varying causes of respiratory failure and differences in flow rate, oxygen concentration, and treatment duration across studies.

Implications

In patients with respiratory failure, HFNC may reduce intubation when compared with usual oxygen therapy and has similar ICU mortality when compared with usual oxygen and NIPPV.

Errors in the Diagnosis of Spinal Epidural Abscesses in the Era of Electronic Health Records. Bhise V et al. Am J Med. 2017;130(8):975-981.¹⁴

Background

Diagnostic errors are common in patients with spinal epidural abscess, but the main contributing factors are unclear.¹⁵

Methods

All patients who were newly diagnosed with spinal epidural abscess in 2013 were identified from the Veterans Affairs (VA) national database. Charts were reviewed for diagnostic delay and contributing factors, including the presence of “red flag” symptoms (eg, fever and neurological deficits).

Findings

Of the 119 patients with a new diagnosis of spinal epidural abscess, 66 (56%) had a diagnostic error. The median time to diagnosis in those with a diagnostic error was 12 days vs four days in those without error (P < .01). Common missed red flags in error cases included fever (n = 57, 86.4%), focal neurologic deficit (n = 54, 81.8%), and active infection (n = 54, 81.8%). Most errors occurred during the provider–patient encounter (eg, information not gathered during the history or physical). The magnitude of harm was serious for most patients (n = 40, 60.6%) and contributed to death in eight patients (12.1%).

Cautions

The study may not be generalizable because it was limited to the VA health system.

Implications

Diagnostic errors are common in patients with spinal epidural abscesses and can lead to serious harm. Health systems should build mechanisms to support providers in the evaluation of patients with back pain.

Total hip or knee replacement (also called total joint arthroplasty) is highly successful at relieving pain and restoring function, but at the risk of acute kidney injury, which is a sudden loss of renal function. Various factors have been associated with this risk, some of which are potentially modifiable, notably, the use of nephrotoxic antibiotics and other drugs.

This review examines the incidence of acute kidney injury using current criteria in total joint arthroplasty of the hip or knee in general, and in the setting of revision surgery for prosthetic joint infection in particular, in which the risk is higher. We identify risk factors for acute kidney injury and propose ways to lower the risk.

MILLIONS OF PROCEDURES ANNUALLY

Total replacement of the hip^1,2 or knee³ is being done more and more. Kurtz et al⁴ estimate that by the year 2030, we will see approximately 3.5 million primary total knee and 500,000 primary total hip replacements every year. In addition, revision total knee procedures are expected to exceed 250,000 per year, and revision total hip procedures are expected to exceed 90,000 per year.⁴

Chronic infection may complicate up to 2% of these procedures and is associated with significant morbidity, death, and financial costs. Currently, it may be the reason for 25% of total joint arthroplasty revisions,⁵ but by the year 2030, it is projected to account for 66% of revision total knee arthroplasties and 48% of revision total hip arthroplasties.⁶

PRIMARY TOTAL JOINT ARTHROPLASTY AND ACUTE KIDNEY INJURY

Study designs, findings varied widely

The incidence of acute kidney injury varied markedly among the studies of primary total joint arthroplasty or revision for aseptic reasons. Numerous factors explain this heterogeneity.

Designs ranged from single-center studies with relatively small numbers of patients to large regional and national samples based on administrative data.

Almost all of the studies were retrospective. We are not aware of any randomized controlled trials.

Discharge diagnosis may miss many cases

Several studies based the diagnosis of acute kidney injury on International Classification of Diseases, Ninth Revision (ICD-9) coding from hospital discharge summaries.

Nadkarni et al,²⁹ in the largest study published to date, used the nationwide inpatient sample database of more than 7 million total joint arthroplasties and found an incidence of acute kidney injury based on ICD-9 coding of 1.3% over the years 2002 to 2012, although this increased to 1.8% to 1.9% from 2010 to 2012.

Lopez-de-Andres et al,³⁰ in a similar study using the Spanish national hospital discharge database, evaluated 20,188 patients who underwent revision total hip or knee arthroplasty and found an overall incidence of acute kidney injury of 0.94%, also using ICD-9 coding.

Gharaibeh et al³¹ used similar methods to diagnose acute kidney injury in a single-center study of 8,949 patients and found an incidence of 1.1%.

Although these 3 studies suggest that the incidence of acute kidney injury is relatively low, Grams et al³⁵ found the sensitivity of ICD-9 coding from hospital records for the diagnosis of acute kidney injury to be only 11.7% compared with KDIGO serum creatinine and urine output criteria. This suggests that the true incidence in these studies may be many times higher, possibly near 10%.

Do all stages of kidney injury count?

Jafari et al,⁷ in a large series from a single medical center, used only the “I” (injury) and “F” (failure) levels of the RIFLE criteria (corresponding to stages 2 and 3 of the KDIGO criteria) and found an incidence of 0.55% in more than 17,000 total joint arthroplasties.

Jamsa et al⁸ used the same criteria for acute kidney injury (only “I” and “F”) and found 58 cases in 5,609 patients in whom postoperative serum creatinine was measured, for an incidence of 1%; the remaining 14,966 patients in their cohort did not have serum creatinine measured, and it was assumed they did not have acute kidney injury. Neither of these studies included the most common “R” (risk) stage of acute kidney injury.

Parr et al³⁶ recently studied a nationwide sample of 657,840 hospitalized veterans and found that of 90,614 who developed acute kidney injury based on KDIGO creatinine criteria, 84% reached only stage R. This suggests that if all stages were considered, the true incidence of acute kidney injury would have been higher—possibly 4% in the Jafari series and possibly 7% in the Jamsa series.

Smaller studies had higher rates

Smaller, single-center series reported much higher incidences of acute kidney injury.

Kimmel et al¹¹ found an incidence of 14.8% in 425 total joint arthroplasties using RIFLE creatinine criteria.

Johansson et al²⁵ found an incidence of 19.9% in 136 total joint arthroplasties using KDIGO creatinine criteria.

Sehgal et al⁹ found an incidence of 21.9% in 659 total joint arthroplasties using AKIN creatinine criteria.

Challagundla et al²⁴ found an incidence of 23.7% in 198 procedures using RIFLE creatinine criteria.

Weingarten et al,¹⁰ in a single-center series of 7,463 total joint arthroplasties, found an incidence of acute kidney injury of only 2.2% using AKIN criteria, although 12% of the patients with acute kidney injury did not return to their baseline serum creatinine levels by 3 months.

Our estimate: Nearly 10%

In total, in the 20 studies in Table 1 that included all stages of acute kidney injury, there were 1,909 cases of acute kidney injury in 34,337 patients, for an incidence of 5.6%. Considering that all studies but one were retrospective and none considered urine output criteria for acute kidney injury, we believe that using current KDIGO criteria, the true incidence of acute kidney injury complicating primary lower-extremity total joint arthroplasties is really closer to 10%.

Various factors have been associated with development of acute kidney injury by multivariate analysis in these studies. Some are modifiable, while others are not, at least in the short term.

Nonmodifiable risk factors

Older age is often significant in studies assessing primary total joint arthroplasty or revision total joint arthroplasty not specifically for infection.^{11,12,16,17,26,28}

Obesity is also a major factor in the development of acute kidney injury,^{7,10–12,17,18} and, along with age, is a major factor contributing to the need for joint replacement in the first place.

Male sex may increase risk.²⁹

Diabetes mellitus was identified as a risk factor in several studies,^10,12,17,20 and hypertension in a few.^7,10,24

Other comorbidities and factors such as cardiovascular disease,^7,10 liver disease,⁷ pulmonary disease,⁷ high American Society of Anesthesiology score,^8,19 and benign heart murmurs preoperatively by routine physical examination have also been linked to acute kidney injury after joint arthroplasty.²⁸

Chronic kidney disease as a risk factor

Chronic kidney disease at baseline was associated with acute kidney injury in several of these series.^{7,11–13,15,19,29}

Warth et al¹² studied 1,038 patients and found an incidence of acute kidney injury of 11% in the 135 with chronic kidney disease (defined as serum creatinine > 1.2 mg/dL) and who received acetaminophen or narcotics for pain control, compared with 4.8% in the remaining 903 patients without chronic kidney disease, who received ketorolac or celecoxib.

Perregaard et al¹³ studied 3,410 patients who underwent total hip arthroplasty and found an incidence of acute kidney injury (per KDIGO creatinine criteria) of 2.2% overall, but 7% in the 134 patients with chronic kidney disease based on KDIGO creatinine criteria.

Nowicka et al¹⁵ found an incidence of acute kidney injury of 16.7% in the 48 patients with chronic kidney disease (defined as a glomerular filtration rate estimated by the Cockroft-Gault formula of less than 60 mL/min/1.73 m²), compared with 4.5% in the remaining 289.

Modifiable risk factors

Modifiable risk factors that should be considered in high-risk cases include anemia, perioperative blood transfusion, perioperative use of renin-angiotensin-aldosterone system inhibitors such as angiotensin-converting enzyme (ACE) inhibitors and angiotensin II receptor blockers (ARBs), particular antibiotics used for prophylaxis, and nonsteroidal anti-inflammatory drugs used postoperatively.

Anemia and blood transfusion

Preoperative anemia has been associated with postoperative acute kidney injury in various surgical settings such as cardiac surgery.^37,38 Perioperative red blood cell transfusions have also been associated with acute kidney injury in cardiac surgery; similar results may apply to total joint arthroplasty.

Choi et al,¹⁷ in 2,467 patients undergoing hip replacement, found a significant risk for acute kidney injury if postoperative hemoglobin was consistently below 10 g/dL compared with consistently above this level, with an inverse probability-of-treatment weighted odds ratio of 1.817 (P = .011).

Others have found a significant association of perioperative blood transfusion with acute kidney injury in total joint arthroplasty.^10,29

Nadkarni et al,²⁹ for example, used the nationwide inpatient sample database and found by multivariate analysis that perioperative blood transfusion was strongly associated with acute kidney injury, with an adjusted odds ratio of 2.28 (95% confidence interval [CI] 2.15–2.42, P < .0001).

Comment. A higher incidence of acute kidney injury may represent confounding by indication bias, as sicker patients or complicated surgeries may require transfusion, and this risk may not be completely accounted for by multivariate analysis. It is also possible, however, that transfusions per se may contribute to acute kidney injury. Possible direct or indirect mechanisms mediating acute kidney injury include hemolytic reactions, circulatory overload, acute lung injury, and immunomodulatory effects.³⁹

Preoperative transfusion in anemic patients undergoing cardiac surgery may also reduce the incidence of postoperative acute kidney injury both by correcting the anemia and by limiting the need for perioperative transfusions.⁴⁰ It remains to be determined whether elective preoperative transfusion to correct anemia would reduce postoperative development of acute kidney injury in total joint arthroplasty. As an aside, perioperative transfusion has also been linked to development of periprosthetic joint infection.⁴¹

Renin-angiotensin-aldosterone system inhibitors

Several studies found perioperative use of renin-angiotensin-aldosterone system inhibitors to be a risk factor for acute kidney injury.

Kimmel et al¹¹ reported adjusted odds ratios of 2.70 (95% CI 1.12–6.48) for ACE inhibitor use and 2.64 (95% CI 1.18–5.93) for ARB use in a study of 425 primary total joint arthroplasties.

Challagundla et al²⁴ found an odds ratio of 3.07 (95% CI 1.40–6.74) with ACE inhibitor or ARB use by multivariate analysis in 198 total joint arthroplasties.

Nielson et al¹⁸ studied 798 patients who underwent total joint arthroplasty and found that preoperative use of renin-angiotensin system inhibitors was associated with a significantly higher rate of postoperative acute kidney injury (8.3% vs 1.7% without inhibition), which was statistically significant by multivariate analysis (odds ratio 2.6, 95% CI 1.04–6.51).

We recommend holding renin-angiotensin-aldosterone system inhibitors 7 days before surgery through the postoperative period in high-risk cases.

Aminoglycoside use as a risk factor

Prophylactic administration of systemic antibiotics is the standard of care. In a systematic review of 26 studies and meta-analysis of 7 studies (3,065 patients), prophylactic antibiotics reduced the relative risk of wound infection by 81% with an absolute risk reduction of 8%.⁴²

A modifiable risk factor for acute kidney injury is the specific antibiotic used for prophylaxis. Multiple studies assessed the risk of acute kidney injury comparing regimens containing an aminoglycoside (typically gentamicin) with regimens lacking these agents.^20–26 In general, these studies found a significantly higher risk of acute kidney injury when gentamicin was used.

Challagundla et al²⁴ found an incidence of acute kidney injury of 52% using RIFLE creatinine criteria in 52 patients receiving 8 g total of flucloxacillin plus 160 mg of gentamicin (120 mg if they weighed less than 60 kg) compared with 8% in 48 patients given cefuroxime (3 g total) and 14% in an additional 52 patients also given cefuroxime.

Johansson et al²⁵ found an incidence of KDIGO creatinine-based acute kidney injury of 13% in 70 patients given dicloxacillin alone prophylactically compared with 27% given dicloxacillin and gentamicin, with a relative risk of 3.

Bell et al,²¹ in a large registry-based analysis from Scotland involving 7,666 elective orthopedic procedures, found that use of flucloxacillin 2 g plus a single dose of gentamicin 4 mg/kg was significantly associated with a 94% higher risk of acute kidney injury (KDIGO creatinine criteria) compared with a cefuroxime-based regimen, with absolute rates increasing from 6.2% to 10.8%.

Dubrovskaya et al²⁰ and Ferguson et al,²⁶ in contrast, found no increased risk with addition of gentamicin.

We recommend avoiding aminoglycosides for prophylaxis in primary lower-extremity total joint arthroplasty in patients at higher risk unless required for specific microbiologic reasons.

Vancomycin may also increase risk

Courtney et al¹⁹ assessed the risk of adding vancomycin to cefazolin for routine prophylaxis in a retrospective series of 1,828 total hip or knee arthroplasties and found a significantly higher rate of acute kidney injury, using AKIN criteria (13% vs 8%, odds ratio by multivariate analysis 1.82, P = .002).¹⁹

Other agents shown to be effective in treating periprosthetic joint infections or complicated skin and soft-tissue infections with resistant organisms include daptomycin⁴³ and linezolid.⁴⁴ These nonnephrotoxic alternatives to vancomycin may be a consideration if prophylaxis for methicillin-resistant Staphylococcus aureus is deemed necessary in patients at risk for acute kidney injury.

Deep infection may complicate nearly 1% of total hip⁴⁵ and 2% of total knee arthroplasties.⁴⁶ Kurtz et al^4,6 have projected that by 2030, infection will be the cause of two-thirds of the estimated 268,000 revision total knee arthroplasties and about half of the estimated 96,700 revision total hip arthroplasties.

The most common method of treating a chronically infected replacement joint is a 2-stage procedure.⁵ First, the prosthesis is removed, all infected bone and soft tissue is debrided, and an antibiotic-loaded cement spacer is implanted. Systemic antibiotics are given concurrently, typically for about 6 weeks. After the infection is brought under control, perhaps 2 to 3 months later, the spacer is removed and a new joint is implanted with antibiotic-loaded cement. A 1-stage procedure may be an option in selected cases and would obviate the need for an antibiotic-loaded cement spacer.^47,48

Of obvious relevance to development of acute kidney injury is the choice and amount of antibiotics embedded in the cement used for spacers and in implantation. Very high antibiotic levels are achieved within the joint space, usually with little systemic absorption, although significant systemic exposure has been documented in some cases.

The polymethylmethacrylate cement used for these purposes comes in 40-g bags. Multiple bags are typically required per joint, perhaps 2 to 4.⁴⁹

The rate of elution of antibiotics is determined by several factors, including surface area, porosity, and the number of antibiotics. In general, elution is greatest early on, with exponential decline lasting perhaps 1 week,  followed by slow, sustained release over weeks to months.⁵⁰ However, several in vitro studies have indicated that only about 5%^50,51 of the total antibiotic actually elutes over time.

Initially, multiple antibiotic-laden cement beads were used to fill the joint space, but this significantly limited function and mobility.⁵² Now, cement spacers are used, and they can be nonarticulating or articulating for maximal joint mobility.⁵³ Although much greater antibiotic elution occurs from beads due to their high surface area-to-volume ratio, spacers still provide an adequate dose.

ANTIBIOTIC-LOADED CEMENT: DOSAGE AND ELUTION CHARACTERISTICS

Antibiotic-loaded cement can be either low-dose or high-dose.

Low-dose cement

Low-dose cement typically consists of 0.5 to 1.0 g of antibiotic per 40-g bag of cement, usually an aminoglycoside (gentamicin or tobramycin) or vancomycin, and can be purchased premixed by the manufacturer. Such cement is only used prophylactically with primary total joint arthroplasty or revision for aseptic reasons, a practice common in Europe but less so in the United States. Some American authors propose antibiotic-loaded cement prophylaxis for patients at high risk, eg, those with immunosuppression, inflammatory cause of arthritis, or diabetes.⁵⁴

Vrabec et al,⁵⁵ in a study of low-dose tobramycin-loaded cement used for primary total knee arthroplasty, found a peak median intra-articular tobramycin concentration of 32 mg/L at 6 hours, declining to 6 mg/L at 48 hours with all serum levels 0.3 mg/L or less (unmeasureable) at similar time points.

Sterling et al,⁵⁶ studying primary total hip arthroplasties with low-dose tobramycin-loaded cement, found mean levels in drainage fluid of 103 mg/L at 6 hours, declining to 15 mg/L at 48 hours. Serum levels peaked at 0.94 mg/L at 3 hours, declining to 0.2 mg/L by 48 hours.

Although most of the antibiotic elution occurs early (within the first week), antibiotic can be found in joint aspirates up to 20 years later.⁵⁷ We are unaware of any well-documented cases of acute kidney injury ascribable to low-dose antibiotic-loaded cement used prophylactically. One case report making this assertion did not determine serum levels of aminoglycoside.⁵⁸

High-dose cement

High-dose antibiotic-loaded cement typically contains about 4 to 8 g of antibiotic per 40-g bag of cement and is used in the treatment of prosthetic joint infection to form the spacers. The antibiotic must be mixed into the cement powder by the surgeon in the operating room.

There is no standard combination or dosage. The choice of antibiotic can be tailored to the infecting organism if known. Otherwise, gram-positive organisms are most common, and vancomycin and aminoglycosides are often used together. This particular combination will enhance the elution of both antibiotics when studied in vitro, a process termed “passive opportunism.”⁵⁹ Other antibiotics in use include aztreonam, piperacillin, teicoplanin, fluoroquinolones, cephalosporins, and daptomycin, among others.

About 8 g of antibiotic total per 40-g bag is the maximum to allow easy molding.⁵² As an example, this may include 4 g of vancomycin and 3.6 g of tobramycin per 40 g. Given that 3 to 4 such bags are often used per joint, there is significant risk of systemic exposure.

Kalil et al⁶⁰ studied 8 patients who received high-dose tobramycin-loaded cement to treat periprosthetic joint infections of the hip or knee and found that 7 had detectable serum levels (mean 0.84 mg/L, highest 2.0 mg/L), including 1 with a level of 0.9 mg/L on day 38; 4 of these 8 developed acute kidney injury by AKIN criteria, although other risk factors for acute kidney injury existed. Nearly all had concomitant vancomycin (3 to 8 g) added to the cement as well.

Hsieh et al⁶¹ studied 46 patients with infected total hip arthroplasties treated with high-dose antibiotic-loaded cement spacers (vancomycin 4 g and aztreonam 4 g per 40-g bag) and found vancomycin levels in joint drainage higher than 1,500 mg/L on day 1, decreasing to 571 mg/L on day 7; serum levels were low (range 0.1–1.6 mg/L at 24 hours), falling to undetectable by 72 hours.

Case reports have associated high-dose antibiotic-loaded cement spacers with acute kidney injury.

Curtis et al⁶² described an 85-year-old patient with stage 3 chronic kidney disease who was treated for an infected total knee arthroplasty with an antibiotic-loaded cement spacer (containing 3.6 g of tobramycin and 3 g of cefazolin per 40-g bag, 3 bags total) and developed stage 3 acute kidney injury. After 16 days and 3 hemodialysis sessions, the patient’s serum tobramycin level was still 2 mg/L despite receiving no systemic tobramycin.

Wu et al⁶³ reported a case of acute kidney injury that required dialysis after implantation of a tobramycin- and vancomycin-loaded spacer, with persistent serum tobramycin levels despite repeated hemodialysis sessions until the spacer was removed.

Chalmers et al⁶⁴ described 2 patients with acute kidney injury and persistently elevated serum tobramycin levels (3.9 mg/L on day 39 in 1 patient and 2.0 mg/L on day 24 in the other patient) despite no systemic administration.

In these and other case reports,^65–67 dialysis and spacer explantation were usually required. 


Comment. It is intuitive that acute kidney injury would more likely complicate revision total joint arthroplasties for infection than for primary total joint arthroplasties or revisions for aseptic reasons, given the systemic effects of infection and exposure to nephrotoxic or allergenic antibiotics. And the available data suggest that the risk of acute kidney injury is higher with revision for prosthetic joint infection than with revision for aseptic reasons. However, many of the studies were retrospective, relatively small, single-center series and used different definitions of acute kidney injury.

Luu et al⁸³ performed a systematic review of studies published between January 1989 and June 2012 reporting systemic complications (including acute kidney injury) of 2-stage revision arthroplasties including placement of an antibiotic-loaded cement spacer for treatment of periprosthetic joint infection. Overall, 10 studies were identified with 544 total patients. Five of these studies, with 409 patients, reported at least 1 case of acute kidney injury for a total of 27 patients, giving an incidence of 6.6% in these studies.^68–71 The remaining 5 studies, totaling 135 patients, did not report any cases of acute kidney injury,^{50,61,76–78} although that was not the primary focus of any of those trials.

Most notable from this systematic review, the study of Menge et al⁶⁹ retrospectively determined the incidence of acute kidney injury (defined as a 50% rise in serum creatinine to > 1.4 mg/dL within 90 days of surgery) to be 17% in 84 patients with infected total knee arthroplasties treated with antibiotic-loaded cement spacers. A mean of 3.5 bags of cement per spacer were used in the 35 articulating spacers, compared with 2.9 per nonarticulating spacer. These spacers contained vancomycin in 82% (median 4.0 g, range 1–16 g) and tobramycin in 94% (median 4.8 g, range 1–12 g), among others in small percentages. The dose of tobramycin in the spacer considered either as a dichotomous variable (> 4.8 g, OR 5.87) or linearly (OR 1.24 per 1-g increase) was significantly associated with acute kidney injury, although systemic administration of aminoglycosides or vancomycin was not.

Additional single-center series that were published subsequent to this review have generally used more current diagnostic criteria.

Noto et al⁷² found that 10 of 46 patients treated with antibiotic-loaded cement spacers had a greater than 50% rise in serum creatinine (average increase 260%). All spacers contained tobramycin (mean dose 8.2 g), and 9 of 10 also contained vancomycin (mean 7.6 g). All of the 9 patients with acute kidney injury with follow-up data recovered renal function.

Reed et al⁷⁵ found 26 cases of acute kidney injury (based on RIFLE creatinine criteria) in 306 patients with antibiotic-loaded cement spacers treating various periprosthetic joint infections (including hips, knees, shoulders, and digits) and compared them with 74 controls who did not develop acute kidney injury. By multivariable analysis, receipt of an ACE inhibitor within 7 days of surgery and receipt of piperacillin-tazobactam within 7 days after surgery were both significantly more common in cases with acute kidney injury than in controls without acute kidney injury.

Aeng et al⁷³ prospectively studied 50 consecutive patients receiving antibiotic-loaded spacers containing tobramycin (with or without vancomycin) for treatment of infected hip or knee replacements. Using RIFLE creatinine criteria, they found an incidence of acute kidney injury of 20% (10 of 50). Factors significantly associated with acute kidney injury included cement premixed by the manufacturer with gentamicin (0.5 g per 40-g bag) in addition to the tobramycin they added, intraoperative blood transfusions, and postoperative use of nonsteroidal anti-inflammatory drugs.

Geller et al,⁷⁴ in a multicenter retrospective study of 247 patients with prosthetic joint infections (156 knees and 91 hips) undergoing antibiotic-loaded cement spacer placement, found an incidence of acute kidney injury of 26% based on KDIGO creatinine criteria. Significant risk factors included higher body mass index, lower preoperative hemoglobin level, drop in hemoglobin after surgery, and comorbidity (hypertension, diabetes, chronic kidney disease, or cardiovascular disease). Most of the spacers contained a combination of vancomycin and either tobramycin (81%) or gentamicin (13%). The spacers contained an average of 5.3 g (range 0.6–18 g) of vancomycin (average 2.65 g per 40-g bag) and an average of 5.2 g (range 0.5–16.4 g) of tobramycin (average 2.6 g per bag).

As in Menge et al,⁶⁹ this study illustrates the wide range of antibiotic dosages in use and the lack of standardization. In contrast to the study by Menge et al, however, development of acute kidney injury was not related to the amount of vancomycin or tobramycin contained in the spacers. Eventual clearance of infection (at 1 and 2 years) was significantly related to increasing amounts of vancomycin. Multiple different systemic antibiotics were used, most commonly vancomycin (44%), and systemic vancomycin was not associated with acute kidney injury.

Yadav et al,⁸¹ in a study of 3,129 consecutive revision procedures of the knee or hip, found an incidence of acute kidney injury by RIFLE creatinine criteria of 29% in the 197 patients who received antibiotic-loaded cement spacers for periprosthetic joint infection compared with 3.4% in the 2,848 who underwent revision for aseptic reasons. In 84 patients with prosthetic joint infection having various surgeries not including placement of a spacer, the acute kidney injury rate at some point in their course was an alarmingly high 82%. In the group that received spacers, only age and comorbidity as assessed by Charlson comorbidity index were independently associated with acute kidney injury by multivariate analysis. Surprisingly, modest renal impairment was protective, possibly because physicians of patients with chronic kidney disease were more vigilant and took appropriate measures to prevent acute kidney injury.

Overall, the risk of acute kidney injury appears to be much higher during treatment of prosthetic joint infection with a 2-stage procedure using an antibiotic-loaded cement spacer than after primary total joint arthroplasty or revision for aseptic reasons, and may complicate up to one-third of cases.

As in primary total joint arthroplasty in general, higher-risk cases should be identified based on age, body mass index, chronic kidney disease, comorbidities (hypertension, diabetes, established cardiovascular disease), and anemia.

Preoperative transfusion can be considered case by case depending on degree of anemia and associated risk factors.

All renin-angiotensin-aldosterone system inhibitors should be withheld starting 1 week before surgery.

Both nonselective and cyclooxygenase-2 selective nonsteroidal anti-inflammatory drugs should be avoided, if possible.

Strict attention should be paid to adequate intraoperative and postoperative fluid resuscitation.

Kidney function should be monitored closely in the early postoperative period, including urine output and daily creatinine for at least 72 hours.

Systemic administration of potentially nephrotoxic antibiotics should be minimized, especially the combination of vancomycin with piperacillin-tazobactam.⁸⁴ Daptomycin is a consideration.⁴³

If acute kidney injury should develop, serum levels of vancomycin or aminoglycosides should be measured if the spacer contains these antibiotics. The spacer may need to be removed if toxic serum levels persist.

TAKE-HOME POINTS

Acute kidney injury may complicate up to 10% of primary lower-extremity total joint arthroplasties and up to 25% of periprosthetic joint infections treated with a 2-stage procedure including placement of an antibiotic-loaded cement spacer in the first stage.

Risk factors for acute kidney injury include older age, obesity, chronic kidney disease, and overall comorbidity. Potentially modifiable risk factors include anemia, perioperative transfusions, aminoglycoside prophylaxis, perioperative renin-angiotensin system blockade, and postoperative nonsteroidal anti-inflammatory drugs. These should be mitigated when possible.

In patients with periprosthetic joint infection who receive antibiotic-loaded cement spacers, especially patients  with additional risk factors for acute kidney injury, strict attention should be paid to the dose of antibiotic in the spacer, with levels checked postoperatively if necessary. Nonnephrotoxic antibiotics should be chosen for systemic administration when possible.

Prospective randomized controlled trials are needed to guide therapy after total joint arthroplasty, and to verify the adverse long-term outcomes of acute kidney injury in this setting.

Total hip or knee replacement (also called total joint arthroplasty) is highly successful at relieving pain and restoring function, but at the risk of acute kidney injury, which is a sudden loss of renal function. Various factors have been associated with this risk, some of which are potentially modifiable, notably, the use of nephrotoxic antibiotics and other drugs.

This review examines the incidence of acute kidney injury using current criteria in total joint arthroplasty of the hip or knee in general, and in the setting of revision surgery for prosthetic joint infection in particular, in which the risk is higher. We identify risk factors for acute kidney injury and propose ways to lower the risk.

MILLIONS OF PROCEDURES ANNUALLY

Total replacement of the hip^1,2 or knee³ is being done more and more. Kurtz et al⁴ estimate that by the year 2030, we will see approximately 3.5 million primary total knee and 500,000 primary total hip replacements every year. In addition, revision total knee procedures are expected to exceed 250,000 per year, and revision total hip procedures are expected to exceed 90,000 per year.⁴

Chronic infection may complicate up to 2% of these procedures and is associated with significant morbidity, death, and financial costs. Currently, it may be the reason for 25% of total joint arthroplasty revisions,⁵ but by the year 2030, it is projected to account for 66% of revision total knee arthroplasties and 48% of revision total hip arthroplasties.⁶

PRIMARY TOTAL JOINT ARTHROPLASTY AND ACUTE KIDNEY INJURY

Study designs, findings varied widely

The incidence of acute kidney injury varied markedly among the studies of primary total joint arthroplasty or revision for aseptic reasons. Numerous factors explain this heterogeneity.

Designs ranged from single-center studies with relatively small numbers of patients to large regional and national samples based on administrative data.

Almost all of the studies were retrospective. We are not aware of any randomized controlled trials.

Discharge diagnosis may miss many cases

Several studies based the diagnosis of acute kidney injury on International Classification of Diseases, Ninth Revision (ICD-9) coding from hospital discharge summaries.

Nadkarni et al,²⁹ in the largest study published to date, used the nationwide inpatient sample database of more than 7 million total joint arthroplasties and found an incidence of acute kidney injury based on ICD-9 coding of 1.3% over the years 2002 to 2012, although this increased to 1.8% to 1.9% from 2010 to 2012.

Lopez-de-Andres et al,³⁰ in a similar study using the Spanish national hospital discharge database, evaluated 20,188 patients who underwent revision total hip or knee arthroplasty and found an overall incidence of acute kidney injury of 0.94%, also using ICD-9 coding.

Gharaibeh et al³¹ used similar methods to diagnose acute kidney injury in a single-center study of 8,949 patients and found an incidence of 1.1%.

Although these 3 studies suggest that the incidence of acute kidney injury is relatively low, Grams et al³⁵ found the sensitivity of ICD-9 coding from hospital records for the diagnosis of acute kidney injury to be only 11.7% compared with KDIGO serum creatinine and urine output criteria. This suggests that the true incidence in these studies may be many times higher, possibly near 10%.

Do all stages of kidney injury count?

Jafari et al,⁷ in a large series from a single medical center, used only the “I” (injury) and “F” (failure) levels of the RIFLE criteria (corresponding to stages 2 and 3 of the KDIGO criteria) and found an incidence of 0.55% in more than 17,000 total joint arthroplasties.

Jamsa et al⁸ used the same criteria for acute kidney injury (only “I” and “F”) and found 58 cases in 5,609 patients in whom postoperative serum creatinine was measured, for an incidence of 1%; the remaining 14,966 patients in their cohort did not have serum creatinine measured, and it was assumed they did not have acute kidney injury. Neither of these studies included the most common “R” (risk) stage of acute kidney injury.

Parr et al³⁶ recently studied a nationwide sample of 657,840 hospitalized veterans and found that of 90,614 who developed acute kidney injury based on KDIGO creatinine criteria, 84% reached only stage R. This suggests that if all stages were considered, the true incidence of acute kidney injury would have been higher—possibly 4% in the Jafari series and possibly 7% in the Jamsa series.

Smaller studies had higher rates

Smaller, single-center series reported much higher incidences of acute kidney injury.

Kimmel et al¹¹ found an incidence of 14.8% in 425 total joint arthroplasties using RIFLE creatinine criteria.

Johansson et al²⁵ found an incidence of 19.9% in 136 total joint arthroplasties using KDIGO creatinine criteria.

Sehgal et al⁹ found an incidence of 21.9% in 659 total joint arthroplasties using AKIN creatinine criteria.

Challagundla et al²⁴ found an incidence of 23.7% in 198 procedures using RIFLE creatinine criteria.

Weingarten et al,¹⁰ in a single-center series of 7,463 total joint arthroplasties, found an incidence of acute kidney injury of only 2.2% using AKIN criteria, although 12% of the patients with acute kidney injury did not return to their baseline serum creatinine levels by 3 months.

Our estimate: Nearly 10%

In total, in the 20 studies in Table 1 that included all stages of acute kidney injury, there were 1,909 cases of acute kidney injury in 34,337 patients, for an incidence of 5.6%. Considering that all studies but one were retrospective and none considered urine output criteria for acute kidney injury, we believe that using current KDIGO criteria, the true incidence of acute kidney injury complicating primary lower-extremity total joint arthroplasties is really closer to 10%.

Various factors have been associated with development of acute kidney injury by multivariate analysis in these studies. Some are modifiable, while others are not, at least in the short term.

Nonmodifiable risk factors

Older age is often significant in studies assessing primary total joint arthroplasty or revision total joint arthroplasty not specifically for infection.^{11,12,16,17,26,28}

Obesity is also a major factor in the development of acute kidney injury,^{7,10–12,17,18} and, along with age, is a major factor contributing to the need for joint replacement in the first place.

Male sex may increase risk.²⁹

Diabetes mellitus was identified as a risk factor in several studies,^10,12,17,20 and hypertension in a few.^7,10,24

Other comorbidities and factors such as cardiovascular disease,^7,10 liver disease,⁷ pulmonary disease,⁷ high American Society of Anesthesiology score,^8,19 and benign heart murmurs preoperatively by routine physical examination have also been linked to acute kidney injury after joint arthroplasty.²⁸

Chronic kidney disease as a risk factor

Chronic kidney disease at baseline was associated with acute kidney injury in several of these series.^{7,11–13,15,19,29}

Warth et al¹² studied 1,038 patients and found an incidence of acute kidney injury of 11% in the 135 with chronic kidney disease (defined as serum creatinine > 1.2 mg/dL) and who received acetaminophen or narcotics for pain control, compared with 4.8% in the remaining 903 patients without chronic kidney disease, who received ketorolac or celecoxib.

Perregaard et al¹³ studied 3,410 patients who underwent total hip arthroplasty and found an incidence of acute kidney injury (per KDIGO creatinine criteria) of 2.2% overall, but 7% in the 134 patients with chronic kidney disease based on KDIGO creatinine criteria.

Nowicka et al¹⁵ found an incidence of acute kidney injury of 16.7% in the 48 patients with chronic kidney disease (defined as a glomerular filtration rate estimated by the Cockroft-Gault formula of less than 60 mL/min/1.73 m²), compared with 4.5% in the remaining 289.

Modifiable risk factors

Modifiable risk factors that should be considered in high-risk cases include anemia, perioperative blood transfusion, perioperative use of renin-angiotensin-aldosterone system inhibitors such as angiotensin-converting enzyme (ACE) inhibitors and angiotensin II receptor blockers (ARBs), particular antibiotics used for prophylaxis, and nonsteroidal anti-inflammatory drugs used postoperatively.

Anemia and blood transfusion

Preoperative anemia has been associated with postoperative acute kidney injury in various surgical settings such as cardiac surgery.^37,38 Perioperative red blood cell transfusions have also been associated with acute kidney injury in cardiac surgery; similar results may apply to total joint arthroplasty.

Choi et al,¹⁷ in 2,467 patients undergoing hip replacement, found a significant risk for acute kidney injury if postoperative hemoglobin was consistently below 10 g/dL compared with consistently above this level, with an inverse probability-of-treatment weighted odds ratio of 1.817 (P = .011).

Others have found a significant association of perioperative blood transfusion with acute kidney injury in total joint arthroplasty.^10,29

Nadkarni et al,²⁹ for example, used the nationwide inpatient sample database and found by multivariate analysis that perioperative blood transfusion was strongly associated with acute kidney injury, with an adjusted odds ratio of 2.28 (95% confidence interval [CI] 2.15–2.42, P < .0001).

Comment. A higher incidence of acute kidney injury may represent confounding by indication bias, as sicker patients or complicated surgeries may require transfusion, and this risk may not be completely accounted for by multivariate analysis. It is also possible, however, that transfusions per se may contribute to acute kidney injury. Possible direct or indirect mechanisms mediating acute kidney injury include hemolytic reactions, circulatory overload, acute lung injury, and immunomodulatory effects.³⁹

Preoperative transfusion in anemic patients undergoing cardiac surgery may also reduce the incidence of postoperative acute kidney injury both by correcting the anemia and by limiting the need for perioperative transfusions.⁴⁰ It remains to be determined whether elective preoperative transfusion to correct anemia would reduce postoperative development of acute kidney injury in total joint arthroplasty. As an aside, perioperative transfusion has also been linked to development of periprosthetic joint infection.⁴¹

Renin-angiotensin-aldosterone system inhibitors

Several studies found perioperative use of renin-angiotensin-aldosterone system inhibitors to be a risk factor for acute kidney injury.

Kimmel et al¹¹ reported adjusted odds ratios of 2.70 (95% CI 1.12–6.48) for ACE inhibitor use and 2.64 (95% CI 1.18–5.93) for ARB use in a study of 425 primary total joint arthroplasties.

Challagundla et al²⁴ found an odds ratio of 3.07 (95% CI 1.40–6.74) with ACE inhibitor or ARB use by multivariate analysis in 198 total joint arthroplasties.

Nielson et al¹⁸ studied 798 patients who underwent total joint arthroplasty and found that preoperative use of renin-angiotensin system inhibitors was associated with a significantly higher rate of postoperative acute kidney injury (8.3% vs 1.7% without inhibition), which was statistically significant by multivariate analysis (odds ratio 2.6, 95% CI 1.04–6.51).

We recommend holding renin-angiotensin-aldosterone system inhibitors 7 days before surgery through the postoperative period in high-risk cases.

Aminoglycoside use as a risk factor

Prophylactic administration of systemic antibiotics is the standard of care. In a systematic review of 26 studies and meta-analysis of 7 studies (3,065 patients), prophylactic antibiotics reduced the relative risk of wound infection by 81% with an absolute risk reduction of 8%.⁴²

A modifiable risk factor for acute kidney injury is the specific antibiotic used for prophylaxis. Multiple studies assessed the risk of acute kidney injury comparing regimens containing an aminoglycoside (typically gentamicin) with regimens lacking these agents.^20–26 In general, these studies found a significantly higher risk of acute kidney injury when gentamicin was used.

Challagundla et al²⁴ found an incidence of acute kidney injury of 52% using RIFLE creatinine criteria in 52 patients receiving 8 g total of flucloxacillin plus 160 mg of gentamicin (120 mg if they weighed less than 60 kg) compared with 8% in 48 patients given cefuroxime (3 g total) and 14% in an additional 52 patients also given cefuroxime.

Johansson et al²⁵ found an incidence of KDIGO creatinine-based acute kidney injury of 13% in 70 patients given dicloxacillin alone prophylactically compared with 27% given dicloxacillin and gentamicin, with a relative risk of 3.

Bell et al,²¹ in a large registry-based analysis from Scotland involving 7,666 elective orthopedic procedures, found that use of flucloxacillin 2 g plus a single dose of gentamicin 4 mg/kg was significantly associated with a 94% higher risk of acute kidney injury (KDIGO creatinine criteria) compared with a cefuroxime-based regimen, with absolute rates increasing from 6.2% to 10.8%.

Dubrovskaya et al²⁰ and Ferguson et al,²⁶ in contrast, found no increased risk with addition of gentamicin.

We recommend avoiding aminoglycosides for prophylaxis in primary lower-extremity total joint arthroplasty in patients at higher risk unless required for specific microbiologic reasons.

Vancomycin may also increase risk

Courtney et al¹⁹ assessed the risk of adding vancomycin to cefazolin for routine prophylaxis in a retrospective series of 1,828 total hip or knee arthroplasties and found a significantly higher rate of acute kidney injury, using AKIN criteria (13% vs 8%, odds ratio by multivariate analysis 1.82, P = .002).¹⁹

Other agents shown to be effective in treating periprosthetic joint infections or complicated skin and soft-tissue infections with resistant organisms include daptomycin⁴³ and linezolid.⁴⁴ These nonnephrotoxic alternatives to vancomycin may be a consideration if prophylaxis for methicillin-resistant Staphylococcus aureus is deemed necessary in patients at risk for acute kidney injury.

Deep infection may complicate nearly 1% of total hip⁴⁵ and 2% of total knee arthroplasties.⁴⁶ Kurtz et al^4,6 have projected that by 2030, infection will be the cause of two-thirds of the estimated 268,000 revision total knee arthroplasties and about half of the estimated 96,700 revision total hip arthroplasties.

The most common method of treating a chronically infected replacement joint is a 2-stage procedure.⁵ First, the prosthesis is removed, all infected bone and soft tissue is debrided, and an antibiotic-loaded cement spacer is implanted. Systemic antibiotics are given concurrently, typically for about 6 weeks. After the infection is brought under control, perhaps 2 to 3 months later, the spacer is removed and a new joint is implanted with antibiotic-loaded cement. A 1-stage procedure may be an option in selected cases and would obviate the need for an antibiotic-loaded cement spacer.^47,48

Of obvious relevance to development of acute kidney injury is the choice and amount of antibiotics embedded in the cement used for spacers and in implantation. Very high antibiotic levels are achieved within the joint space, usually with little systemic absorption, although significant systemic exposure has been documented in some cases.

The polymethylmethacrylate cement used for these purposes comes in 40-g bags. Multiple bags are typically required per joint, perhaps 2 to 4.⁴⁹

The rate of elution of antibiotics is determined by several factors, including surface area, porosity, and the number of antibiotics. In general, elution is greatest early on, with exponential decline lasting perhaps 1 week,  followed by slow, sustained release over weeks to months.⁵⁰ However, several in vitro studies have indicated that only about 5%^50,51 of the total antibiotic actually elutes over time.

Initially, multiple antibiotic-laden cement beads were used to fill the joint space, but this significantly limited function and mobility.⁵² Now, cement spacers are used, and they can be nonarticulating or articulating for maximal joint mobility.⁵³ Although much greater antibiotic elution occurs from beads due to their high surface area-to-volume ratio, spacers still provide an adequate dose.

ANTIBIOTIC-LOADED CEMENT: DOSAGE AND ELUTION CHARACTERISTICS

Antibiotic-loaded cement can be either low-dose or high-dose.

Low-dose cement

Low-dose cement typically consists of 0.5 to 1.0 g of antibiotic per 40-g bag of cement, usually an aminoglycoside (gentamicin or tobramycin) or vancomycin, and can be purchased premixed by the manufacturer. Such cement is only used prophylactically with primary total joint arthroplasty or revision for aseptic reasons, a practice common in Europe but less so in the United States. Some American authors propose antibiotic-loaded cement prophylaxis for patients at high risk, eg, those with immunosuppression, inflammatory cause of arthritis, or diabetes.⁵⁴

Vrabec et al,⁵⁵ in a study of low-dose tobramycin-loaded cement used for primary total knee arthroplasty, found a peak median intra-articular tobramycin concentration of 32 mg/L at 6 hours, declining to 6 mg/L at 48 hours with all serum levels 0.3 mg/L or less (unmeasureable) at similar time points.

Sterling et al,⁵⁶ studying primary total hip arthroplasties with low-dose tobramycin-loaded cement, found mean levels in drainage fluid of 103 mg/L at 6 hours, declining to 15 mg/L at 48 hours. Serum levels peaked at 0.94 mg/L at 3 hours, declining to 0.2 mg/L by 48 hours.

Although most of the antibiotic elution occurs early (within the first week), antibiotic can be found in joint aspirates up to 20 years later.⁵⁷ We are unaware of any well-documented cases of acute kidney injury ascribable to low-dose antibiotic-loaded cement used prophylactically. One case report making this assertion did not determine serum levels of aminoglycoside.⁵⁸

High-dose cement

High-dose antibiotic-loaded cement typically contains about 4 to 8 g of antibiotic per 40-g bag of cement and is used in the treatment of prosthetic joint infection to form the spacers. The antibiotic must be mixed into the cement powder by the surgeon in the operating room.

There is no standard combination or dosage. The choice of antibiotic can be tailored to the infecting organism if known. Otherwise, gram-positive organisms are most common, and vancomycin and aminoglycosides are often used together. This particular combination will enhance the elution of both antibiotics when studied in vitro, a process termed “passive opportunism.”⁵⁹ Other antibiotics in use include aztreonam, piperacillin, teicoplanin, fluoroquinolones, cephalosporins, and daptomycin, among others.

About 8 g of antibiotic total per 40-g bag is the maximum to allow easy molding.⁵² As an example, this may include 4 g of vancomycin and 3.6 g of tobramycin per 40 g. Given that 3 to 4 such bags are often used per joint, there is significant risk of systemic exposure.

Kalil et al⁶⁰ studied 8 patients who received high-dose tobramycin-loaded cement to treat periprosthetic joint infections of the hip or knee and found that 7 had detectable serum levels (mean 0.84 mg/L, highest 2.0 mg/L), including 1 with a level of 0.9 mg/L on day 38; 4 of these 8 developed acute kidney injury by AKIN criteria, although other risk factors for acute kidney injury existed. Nearly all had concomitant vancomycin (3 to 8 g) added to the cement as well.

Hsieh et al⁶¹ studied 46 patients with infected total hip arthroplasties treated with high-dose antibiotic-loaded cement spacers (vancomycin 4 g and aztreonam 4 g per 40-g bag) and found vancomycin levels in joint drainage higher than 1,500 mg/L on day 1, decreasing to 571 mg/L on day 7; serum levels were low (range 0.1–1.6 mg/L at 24 hours), falling to undetectable by 72 hours.

Case reports have associated high-dose antibiotic-loaded cement spacers with acute kidney injury.

Curtis et al⁶² described an 85-year-old patient with stage 3 chronic kidney disease who was treated for an infected total knee arthroplasty with an antibiotic-loaded cement spacer (containing 3.6 g of tobramycin and 3 g of cefazolin per 40-g bag, 3 bags total) and developed stage 3 acute kidney injury. After 16 days and 3 hemodialysis sessions, the patient’s serum tobramycin level was still 2 mg/L despite receiving no systemic tobramycin.

Wu et al⁶³ reported a case of acute kidney injury that required dialysis after implantation of a tobramycin- and vancomycin-loaded spacer, with persistent serum tobramycin levels despite repeated hemodialysis sessions until the spacer was removed.

Chalmers et al⁶⁴ described 2 patients with acute kidney injury and persistently elevated serum tobramycin levels (3.9 mg/L on day 39 in 1 patient and 2.0 mg/L on day 24 in the other patient) despite no systemic administration.

In these and other case reports,^65–67 dialysis and spacer explantation were usually required. 


Comment. It is intuitive that acute kidney injury would more likely complicate revision total joint arthroplasties for infection than for primary total joint arthroplasties or revisions for aseptic reasons, given the systemic effects of infection and exposure to nephrotoxic or allergenic antibiotics. And the available data suggest that the risk of acute kidney injury is higher with revision for prosthetic joint infection than with revision for aseptic reasons. However, many of the studies were retrospective, relatively small, single-center series and used different definitions of acute kidney injury.

Luu et al⁸³ performed a systematic review of studies published between January 1989 and June 2012 reporting systemic complications (including acute kidney injury) of 2-stage revision arthroplasties including placement of an antibiotic-loaded cement spacer for treatment of periprosthetic joint infection. Overall, 10 studies were identified with 544 total patients. Five of these studies, with 409 patients, reported at least 1 case of acute kidney injury for a total of 27 patients, giving an incidence of 6.6% in these studies.^68–71 The remaining 5 studies, totaling 135 patients, did not report any cases of acute kidney injury,^{50,61,76–78} although that was not the primary focus of any of those trials.

Most notable from this systematic review, the study of Menge et al⁶⁹ retrospectively determined the incidence of acute kidney injury (defined as a 50% rise in serum creatinine to > 1.4 mg/dL within 90 days of surgery) to be 17% in 84 patients with infected total knee arthroplasties treated with antibiotic-loaded cement spacers. A mean of 3.5 bags of cement per spacer were used in the 35 articulating spacers, compared with 2.9 per nonarticulating spacer. These spacers contained vancomycin in 82% (median 4.0 g, range 1–16 g) and tobramycin in 94% (median 4.8 g, range 1–12 g), among others in small percentages. The dose of tobramycin in the spacer considered either as a dichotomous variable (> 4.8 g, OR 5.87) or linearly (OR 1.24 per 1-g increase) was significantly associated with acute kidney injury, although systemic administration of aminoglycosides or vancomycin was not.

Additional single-center series that were published subsequent to this review have generally used more current diagnostic criteria.

Noto et al⁷² found that 10 of 46 patients treated with antibiotic-loaded cement spacers had a greater than 50% rise in serum creatinine (average increase 260%). All spacers contained tobramycin (mean dose 8.2 g), and 9 of 10 also contained vancomycin (mean 7.6 g). All of the 9 patients with acute kidney injury with follow-up data recovered renal function.

Reed et al⁷⁵ found 26 cases of acute kidney injury (based on RIFLE creatinine criteria) in 306 patients with antibiotic-loaded cement spacers treating various periprosthetic joint infections (including hips, knees, shoulders, and digits) and compared them with 74 controls who did not develop acute kidney injury. By multivariable analysis, receipt of an ACE inhibitor within 7 days of surgery and receipt of piperacillin-tazobactam within 7 days after surgery were both significantly more common in cases with acute kidney injury than in controls without acute kidney injury.

Aeng et al⁷³ prospectively studied 50 consecutive patients receiving antibiotic-loaded spacers containing tobramycin (with or without vancomycin) for treatment of infected hip or knee replacements. Using RIFLE creatinine criteria, they found an incidence of acute kidney injury of 20% (10 of 50). Factors significantly associated with acute kidney injury included cement premixed by the manufacturer with gentamicin (0.5 g per 40-g bag) in addition to the tobramycin they added, intraoperative blood transfusions, and postoperative use of nonsteroidal anti-inflammatory drugs.

Geller et al,⁷⁴ in a multicenter retrospective study of 247 patients with prosthetic joint infections (156 knees and 91 hips) undergoing antibiotic-loaded cement spacer placement, found an incidence of acute kidney injury of 26% based on KDIGO creatinine criteria. Significant risk factors included higher body mass index, lower preoperative hemoglobin level, drop in hemoglobin after surgery, and comorbidity (hypertension, diabetes, chronic kidney disease, or cardiovascular disease). Most of the spacers contained a combination of vancomycin and either tobramycin (81%) or gentamicin (13%). The spacers contained an average of 5.3 g (range 0.6–18 g) of vancomycin (average 2.65 g per 40-g bag) and an average of 5.2 g (range 0.5–16.4 g) of tobramycin (average 2.6 g per bag).

As in Menge et al,⁶⁹ this study illustrates the wide range of antibiotic dosages in use and the lack of standardization. In contrast to the study by Menge et al, however, development of acute kidney injury was not related to the amount of vancomycin or tobramycin contained in the spacers. Eventual clearance of infection (at 1 and 2 years) was significantly related to increasing amounts of vancomycin. Multiple different systemic antibiotics were used, most commonly vancomycin (44%), and systemic vancomycin was not associated with acute kidney injury.

Yadav et al,⁸¹ in a study of 3,129 consecutive revision procedures of the knee or hip, found an incidence of acute kidney injury by RIFLE creatinine criteria of 29% in the 197 patients who received antibiotic-loaded cement spacers for periprosthetic joint infection compared with 3.4% in the 2,848 who underwent revision for aseptic reasons. In 84 patients with prosthetic joint infection having various surgeries not including placement of a spacer, the acute kidney injury rate at some point in their course was an alarmingly high 82%. In the group that received spacers, only age and comorbidity as assessed by Charlson comorbidity index were independently associated with acute kidney injury by multivariate analysis. Surprisingly, modest renal impairment was protective, possibly because physicians of patients with chronic kidney disease were more vigilant and took appropriate measures to prevent acute kidney injury.

Overall, the risk of acute kidney injury appears to be much higher during treatment of prosthetic joint infection with a 2-stage procedure using an antibiotic-loaded cement spacer than after primary total joint arthroplasty or revision for aseptic reasons, and may complicate up to one-third of cases.

As in primary total joint arthroplasty in general, higher-risk cases should be identified based on age, body mass index, chronic kidney disease, comorbidities (hypertension, diabetes, established cardiovascular disease), and anemia.

Preoperative transfusion can be considered case by case depending on degree of anemia and associated risk factors.

All renin-angiotensin-aldosterone system inhibitors should be withheld starting 1 week before surgery.

Both nonselective and cyclooxygenase-2 selective nonsteroidal anti-inflammatory drugs should be avoided, if possible.

Strict attention should be paid to adequate intraoperative and postoperative fluid resuscitation.

Kidney function should be monitored closely in the early postoperative period, including urine output and daily creatinine for at least 72 hours.

Systemic administration of potentially nephrotoxic antibiotics should be minimized, especially the combination of vancomycin with piperacillin-tazobactam.⁸⁴ Daptomycin is a consideration.⁴³

If acute kidney injury should develop, serum levels of vancomycin or aminoglycosides should be measured if the spacer contains these antibiotics. The spacer may need to be removed if toxic serum levels persist.

TAKE-HOME POINTS

Acute kidney injury may complicate up to 10% of primary lower-extremity total joint arthroplasties and up to 25% of periprosthetic joint infections treated with a 2-stage procedure including placement of an antibiotic-loaded cement spacer in the first stage.

Risk factors for acute kidney injury include older age, obesity, chronic kidney disease, and overall comorbidity. Potentially modifiable risk factors include anemia, perioperative transfusions, aminoglycoside prophylaxis, perioperative renin-angiotensin system blockade, and postoperative nonsteroidal anti-inflammatory drugs. These should be mitigated when possible.

In patients with periprosthetic joint infection who receive antibiotic-loaded cement spacers, especially patients  with additional risk factors for acute kidney injury, strict attention should be paid to the dose of antibiotic in the spacer, with levels checked postoperatively if necessary. Nonnephrotoxic antibiotics should be chosen for systemic administration when possible.

Prospective randomized controlled trials are needed to guide therapy after total joint arthroplasty, and to verify the adverse long-term outcomes of acute kidney injury in this setting.

Total hip or knee replacement (also called total joint arthroplasty) is highly successful at relieving pain and restoring function, but at the risk of acute kidney injury, which is a sudden loss of renal function. Various factors have been associated with this risk, some of which are potentially modifiable, notably, the use of nephrotoxic antibiotics and other drugs.

This review examines the incidence of acute kidney injury using current criteria in total joint arthroplasty of the hip or knee in general, and in the setting of revision surgery for prosthetic joint infection in particular, in which the risk is higher. We identify risk factors for acute kidney injury and propose ways to lower the risk.

MILLIONS OF PROCEDURES ANNUALLY

Total replacement of the hip^1,2 or knee³ is being done more and more. Kurtz et al⁴ estimate that by the year 2030, we will see approximately 3.5 million primary total knee and 500,000 primary total hip replacements every year. In addition, revision total knee procedures are expected to exceed 250,000 per year, and revision total hip procedures are expected to exceed 90,000 per year.⁴

Chronic infection may complicate up to 2% of these procedures and is associated with significant morbidity, death, and financial costs. Currently, it may be the reason for 25% of total joint arthroplasty revisions,⁵ but by the year 2030, it is projected to account for 66% of revision total knee arthroplasties and 48% of revision total hip arthroplasties.⁶

PRIMARY TOTAL JOINT ARTHROPLASTY AND ACUTE KIDNEY INJURY

Study designs, findings varied widely

The incidence of acute kidney injury varied markedly among the studies of primary total joint arthroplasty or revision for aseptic reasons. Numerous factors explain this heterogeneity.

Designs ranged from single-center studies with relatively small numbers of patients to large regional and national samples based on administrative data.

Almost all of the studies were retrospective. We are not aware of any randomized controlled trials.

Discharge diagnosis may miss many cases

Several studies based the diagnosis of acute kidney injury on International Classification of Diseases, Ninth Revision (ICD-9) coding from hospital discharge summaries.

Nadkarni et al,²⁹ in the largest study published to date, used the nationwide inpatient sample database of more than 7 million total joint arthroplasties and found an incidence of acute kidney injury based on ICD-9 coding of 1.3% over the years 2002 to 2012, although this increased to 1.8% to 1.9% from 2010 to 2012.

Lopez-de-Andres et al,³⁰ in a similar study using the Spanish national hospital discharge database, evaluated 20,188 patients who underwent revision total hip or knee arthroplasty and found an overall incidence of acute kidney injury of 0.94%, also using ICD-9 coding.

Gharaibeh et al³¹ used similar methods to diagnose acute kidney injury in a single-center study of 8,949 patients and found an incidence of 1.1%.

Although these 3 studies suggest that the incidence of acute kidney injury is relatively low, Grams et al³⁵ found the sensitivity of ICD-9 coding from hospital records for the diagnosis of acute kidney injury to be only 11.7% compared with KDIGO serum creatinine and urine output criteria. This suggests that the true incidence in these studies may be many times higher, possibly near 10%.

Do all stages of kidney injury count?

Jafari et al,⁷ in a large series from a single medical center, used only the “I” (injury) and “F” (failure) levels of the RIFLE criteria (corresponding to stages 2 and 3 of the KDIGO criteria) and found an incidence of 0.55% in more than 17,000 total joint arthroplasties.

Jamsa et al⁸ used the same criteria for acute kidney injury (only “I” and “F”) and found 58 cases in 5,609 patients in whom postoperative serum creatinine was measured, for an incidence of 1%; the remaining 14,966 patients in their cohort did not have serum creatinine measured, and it was assumed they did not have acute kidney injury. Neither of these studies included the most common “R” (risk) stage of acute kidney injury.

Parr et al³⁶ recently studied a nationwide sample of 657,840 hospitalized veterans and found that of 90,614 who developed acute kidney injury based on KDIGO creatinine criteria, 84% reached only stage R. This suggests that if all stages were considered, the true incidence of acute kidney injury would have been higher—possibly 4% in the Jafari series and possibly 7% in the Jamsa series.

Smaller studies had higher rates

Smaller, single-center series reported much higher incidences of acute kidney injury.

Kimmel et al¹¹ found an incidence of 14.8% in 425 total joint arthroplasties using RIFLE creatinine criteria.

Johansson et al²⁵ found an incidence of 19.9% in 136 total joint arthroplasties using KDIGO creatinine criteria.

Sehgal et al⁹ found an incidence of 21.9% in 659 total joint arthroplasties using AKIN creatinine criteria.

Challagundla et al²⁴ found an incidence of 23.7% in 198 procedures using RIFLE creatinine criteria.

Weingarten et al,¹⁰ in a single-center series of 7,463 total joint arthroplasties, found an incidence of acute kidney injury of only 2.2% using AKIN criteria, although 12% of the patients with acute kidney injury did not return to their baseline serum creatinine levels by 3 months.

Our estimate: Nearly 10%

In total, in the 20 studies in Table 1 that included all stages of acute kidney injury, there were 1,909 cases of acute kidney injury in 34,337 patients, for an incidence of 5.6%. Considering that all studies but one were retrospective and none considered urine output criteria for acute kidney injury, we believe that using current KDIGO criteria, the true incidence of acute kidney injury complicating primary lower-extremity total joint arthroplasties is really closer to 10%.

Various factors have been associated with development of acute kidney injury by multivariate analysis in these studies. Some are modifiable, while others are not, at least in the short term.

Nonmodifiable risk factors

Older age is often significant in studies assessing primary total joint arthroplasty or revision total joint arthroplasty not specifically for infection.^{11,12,16,17,26,28}

Obesity is also a major factor in the development of acute kidney injury,^{7,10–12,17,18} and, along with age, is a major factor contributing to the need for joint replacement in the first place.

Male sex may increase risk.²⁹

Diabetes mellitus was identified as a risk factor in several studies,^10,12,17,20 and hypertension in a few.^7,10,24

Other comorbidities and factors such as cardiovascular disease,^7,10 liver disease,⁷ pulmonary disease,⁷ high American Society of Anesthesiology score,^8,19 and benign heart murmurs preoperatively by routine physical examination have also been linked to acute kidney injury after joint arthroplasty.²⁸

Chronic kidney disease as a risk factor

Chronic kidney disease at baseline was associated with acute kidney injury in several of these series.^{7,11–13,15,19,29}

Warth et al¹² studied 1,038 patients and found an incidence of acute kidney injury of 11% in the 135 with chronic kidney disease (defined as serum creatinine > 1.2 mg/dL) and who received acetaminophen or narcotics for pain control, compared with 4.8% in the remaining 903 patients without chronic kidney disease, who received ketorolac or celecoxib.

Perregaard et al¹³ studied 3,410 patients who underwent total hip arthroplasty and found an incidence of acute kidney injury (per KDIGO creatinine criteria) of 2.2% overall, but 7% in the 134 patients with chronic kidney disease based on KDIGO creatinine criteria.

Nowicka et al¹⁵ found an incidence of acute kidney injury of 16.7% in the 48 patients with chronic kidney disease (defined as a glomerular filtration rate estimated by the Cockroft-Gault formula of less than 60 mL/min/1.73 m²), compared with 4.5% in the remaining 289.

Modifiable risk factors

Modifiable risk factors that should be considered in high-risk cases include anemia, perioperative blood transfusion, perioperative use of renin-angiotensin-aldosterone system inhibitors such as angiotensin-converting enzyme (ACE) inhibitors and angiotensin II receptor blockers (ARBs), particular antibiotics used for prophylaxis, and nonsteroidal anti-inflammatory drugs used postoperatively.

Anemia and blood transfusion

Preoperative anemia has been associated with postoperative acute kidney injury in various surgical settings such as cardiac surgery.^37,38 Perioperative red blood cell transfusions have also been associated with acute kidney injury in cardiac surgery; similar results may apply to total joint arthroplasty.

Choi et al,¹⁷ in 2,467 patients undergoing hip replacement, found a significant risk for acute kidney injury if postoperative hemoglobin was consistently below 10 g/dL compared with consistently above this level, with an inverse probability-of-treatment weighted odds ratio of 1.817 (P = .011).

Others have found a significant association of perioperative blood transfusion with acute kidney injury in total joint arthroplasty.^10,29

Nadkarni et al,²⁹ for example, used the nationwide inpatient sample database and found by multivariate analysis that perioperative blood transfusion was strongly associated with acute kidney injury, with an adjusted odds ratio of 2.28 (95% confidence interval [CI] 2.15–2.42, P < .0001).

Comment. A higher incidence of acute kidney injury may represent confounding by indication bias, as sicker patients or complicated surgeries may require transfusion, and this risk may not be completely accounted for by multivariate analysis. It is also possible, however, that transfusions per se may contribute to acute kidney injury. Possible direct or indirect mechanisms mediating acute kidney injury include hemolytic reactions, circulatory overload, acute lung injury, and immunomodulatory effects.³⁹

Preoperative transfusion in anemic patients undergoing cardiac surgery may also reduce the incidence of postoperative acute kidney injury both by correcting the anemia and by limiting the need for perioperative transfusions.⁴⁰ It remains to be determined whether elective preoperative transfusion to correct anemia would reduce postoperative development of acute kidney injury in total joint arthroplasty. As an aside, perioperative transfusion has also been linked to development of periprosthetic joint infection.⁴¹

Renin-angiotensin-aldosterone system inhibitors

Several studies found perioperative use of renin-angiotensin-aldosterone system inhibitors to be a risk factor for acute kidney injury.

Kimmel et al¹¹ reported adjusted odds ratios of 2.70 (95% CI 1.12–6.48) for ACE inhibitor use and 2.64 (95% CI 1.18–5.93) for ARB use in a study of 425 primary total joint arthroplasties.

Challagundla et al²⁴ found an odds ratio of 3.07 (95% CI 1.40–6.74) with ACE inhibitor or ARB use by multivariate analysis in 198 total joint arthroplasties.

Nielson et al¹⁸ studied 798 patients who underwent total joint arthroplasty and found that preoperative use of renin-angiotensin system inhibitors was associated with a significantly higher rate of postoperative acute kidney injury (8.3% vs 1.7% without inhibition), which was statistically significant by multivariate analysis (odds ratio 2.6, 95% CI 1.04–6.51).

We recommend holding renin-angiotensin-aldosterone system inhibitors 7 days before surgery through the postoperative period in high-risk cases.

Aminoglycoside use as a risk factor

Prophylactic administration of systemic antibiotics is the standard of care. In a systematic review of 26 studies and meta-analysis of 7 studies (3,065 patients), prophylactic antibiotics reduced the relative risk of wound infection by 81% with an absolute risk reduction of 8%.⁴²

A modifiable risk factor for acute kidney injury is the specific antibiotic used for prophylaxis. Multiple studies assessed the risk of acute kidney injury comparing regimens containing an aminoglycoside (typically gentamicin) with regimens lacking these agents.^20–26 In general, these studies found a significantly higher risk of acute kidney injury when gentamicin was used.

Challagundla et al²⁴ found an incidence of acute kidney injury of 52% using RIFLE creatinine criteria in 52 patients receiving 8 g total of flucloxacillin plus 160 mg of gentamicin (120 mg if they weighed less than 60 kg) compared with 8% in 48 patients given cefuroxime (3 g total) and 14% in an additional 52 patients also given cefuroxime.

Johansson et al²⁵ found an incidence of KDIGO creatinine-based acute kidney injury of 13% in 70 patients given dicloxacillin alone prophylactically compared with 27% given dicloxacillin and gentamicin, with a relative risk of 3.

Bell et al,²¹ in a large registry-based analysis from Scotland involving 7,666 elective orthopedic procedures, found that use of flucloxacillin 2 g plus a single dose of gentamicin 4 mg/kg was significantly associated with a 94% higher risk of acute kidney injury (KDIGO creatinine criteria) compared with a cefuroxime-based regimen, with absolute rates increasing from 6.2% to 10.8%.

Dubrovskaya et al²⁰ and Ferguson et al,²⁶ in contrast, found no increased risk with addition of gentamicin.

We recommend avoiding aminoglycosides for prophylaxis in primary lower-extremity total joint arthroplasty in patients at higher risk unless required for specific microbiologic reasons.

Vancomycin may also increase risk

Courtney et al¹⁹ assessed the risk of adding vancomycin to cefazolin for routine prophylaxis in a retrospective series of 1,828 total hip or knee arthroplasties and found a significantly higher rate of acute kidney injury, using AKIN criteria (13% vs 8%, odds ratio by multivariate analysis 1.82, P = .002).¹⁹

Other agents shown to be effective in treating periprosthetic joint infections or complicated skin and soft-tissue infections with resistant organisms include daptomycin⁴³ and linezolid.⁴⁴ These nonnephrotoxic alternatives to vancomycin may be a consideration if prophylaxis for methicillin-resistant Staphylococcus aureus is deemed necessary in patients at risk for acute kidney injury.

Deep infection may complicate nearly 1% of total hip⁴⁵ and 2% of total knee arthroplasties.⁴⁶ Kurtz et al^4,6 have projected that by 2030, infection will be the cause of two-thirds of the estimated 268,000 revision total knee arthroplasties and about half of the estimated 96,700 revision total hip arthroplasties.

The most common method of treating a chronically infected replacement joint is a 2-stage procedure.⁵ First, the prosthesis is removed, all infected bone and soft tissue is debrided, and an antibiotic-loaded cement spacer is implanted. Systemic antibiotics are given concurrently, typically for about 6 weeks. After the infection is brought under control, perhaps 2 to 3 months later, the spacer is removed and a new joint is implanted with antibiotic-loaded cement. A 1-stage procedure may be an option in selected cases and would obviate the need for an antibiotic-loaded cement spacer.^47,48

Of obvious relevance to development of acute kidney injury is the choice and amount of antibiotics embedded in the cement used for spacers and in implantation. Very high antibiotic levels are achieved within the joint space, usually with little systemic absorption, although significant systemic exposure has been documented in some cases.

The polymethylmethacrylate cement used for these purposes comes in 40-g bags. Multiple bags are typically required per joint, perhaps 2 to 4.⁴⁹

The rate of elution of antibiotics is determined by several factors, including surface area, porosity, and the number of antibiotics. In general, elution is greatest early on, with exponential decline lasting perhaps 1 week,  followed by slow, sustained release over weeks to months.⁵⁰ However, several in vitro studies have indicated that only about 5%^50,51 of the total antibiotic actually elutes over time.

Initially, multiple antibiotic-laden cement beads were used to fill the joint space, but this significantly limited function and mobility.⁵² Now, cement spacers are used, and they can be nonarticulating or articulating for maximal joint mobility.⁵³ Although much greater antibiotic elution occurs from beads due to their high surface area-to-volume ratio, spacers still provide an adequate dose.

ANTIBIOTIC-LOADED CEMENT: DOSAGE AND ELUTION CHARACTERISTICS

Antibiotic-loaded cement can be either low-dose or high-dose.

Low-dose cement

Low-dose cement typically consists of 0.5 to 1.0 g of antibiotic per 40-g bag of cement, usually an aminoglycoside (gentamicin or tobramycin) or vancomycin, and can be purchased premixed by the manufacturer. Such cement is only used prophylactically with primary total joint arthroplasty or revision for aseptic reasons, a practice common in Europe but less so in the United States. Some American authors propose antibiotic-loaded cement prophylaxis for patients at high risk, eg, those with immunosuppression, inflammatory cause of arthritis, or diabetes.⁵⁴

Vrabec et al,⁵⁵ in a study of low-dose tobramycin-loaded cement used for primary total knee arthroplasty, found a peak median intra-articular tobramycin concentration of 32 mg/L at 6 hours, declining to 6 mg/L at 48 hours with all serum levels 0.3 mg/L or less (unmeasureable) at similar time points.

Sterling et al,⁵⁶ studying primary total hip arthroplasties with low-dose tobramycin-loaded cement, found mean levels in drainage fluid of 103 mg/L at 6 hours, declining to 15 mg/L at 48 hours. Serum levels peaked at 0.94 mg/L at 3 hours, declining to 0.2 mg/L by 48 hours.

Although most of the antibiotic elution occurs early (within the first week), antibiotic can be found in joint aspirates up to 20 years later.⁵⁷ We are unaware of any well-documented cases of acute kidney injury ascribable to low-dose antibiotic-loaded cement used prophylactically. One case report making this assertion did not determine serum levels of aminoglycoside.⁵⁸

High-dose cement

High-dose antibiotic-loaded cement typically contains about 4 to 8 g of antibiotic per 40-g bag of cement and is used in the treatment of prosthetic joint infection to form the spacers. The antibiotic must be mixed into the cement powder by the surgeon in the operating room.

There is no standard combination or dosage. The choice of antibiotic can be tailored to the infecting organism if known. Otherwise, gram-positive organisms are most common, and vancomycin and aminoglycosides are often used together. This particular combination will enhance the elution of both antibiotics when studied in vitro, a process termed “passive opportunism.”⁵⁹ Other antibiotics in use include aztreonam, piperacillin, teicoplanin, fluoroquinolones, cephalosporins, and daptomycin, among others.

About 8 g of antibiotic total per 40-g bag is the maximum to allow easy molding.⁵² As an example, this may include 4 g of vancomycin and 3.6 g of tobramycin per 40 g. Given that 3 to 4 such bags are often used per joint, there is significant risk of systemic exposure.

Kalil et al⁶⁰ studied 8 patients who received high-dose tobramycin-loaded cement to treat periprosthetic joint infections of the hip or knee and found that 7 had detectable serum levels (mean 0.84 mg/L, highest 2.0 mg/L), including 1 with a level of 0.9 mg/L on day 38; 4 of these 8 developed acute kidney injury by AKIN criteria, although other risk factors for acute kidney injury existed. Nearly all had concomitant vancomycin (3 to 8 g) added to the cement as well.

Hsieh et al⁶¹ studied 46 patients with infected total hip arthroplasties treated with high-dose antibiotic-loaded cement spacers (vancomycin 4 g and aztreonam 4 g per 40-g bag) and found vancomycin levels in joint drainage higher than 1,500 mg/L on day 1, decreasing to 571 mg/L on day 7; serum levels were low (range 0.1–1.6 mg/L at 24 hours), falling to undetectable by 72 hours.

Case reports have associated high-dose antibiotic-loaded cement spacers with acute kidney injury.

Curtis et al⁶² described an 85-year-old patient with stage 3 chronic kidney disease who was treated for an infected total knee arthroplasty with an antibiotic-loaded cement spacer (containing 3.6 g of tobramycin and 3 g of cefazolin per 40-g bag, 3 bags total) and developed stage 3 acute kidney injury. After 16 days and 3 hemodialysis sessions, the patient’s serum tobramycin level was still 2 mg/L despite receiving no systemic tobramycin.

Wu et al⁶³ reported a case of acute kidney injury that required dialysis after implantation of a tobramycin- and vancomycin-loaded spacer, with persistent serum tobramycin levels despite repeated hemodialysis sessions until the spacer was removed.

Chalmers et al⁶⁴ described 2 patients with acute kidney injury and persistently elevated serum tobramycin levels (3.9 mg/L on day 39 in 1 patient and 2.0 mg/L on day 24 in the other patient) despite no systemic administration.

In these and other case reports,^65–67 dialysis and spacer explantation were usually required. 


Comment. It is intuitive that acute kidney injury would more likely complicate revision total joint arthroplasties for infection than for primary total joint arthroplasties or revisions for aseptic reasons, given the systemic effects of infection and exposure to nephrotoxic or allergenic antibiotics. And the available data suggest that the risk of acute kidney injury is higher with revision for prosthetic joint infection than with revision for aseptic reasons. However, many of the studies were retrospective, relatively small, single-center series and used different definitions of acute kidney injury.

Luu et al⁸³ performed a systematic review of studies published between January 1989 and June 2012 reporting systemic complications (including acute kidney injury) of 2-stage revision arthroplasties including placement of an antibiotic-loaded cement spacer for treatment of periprosthetic joint infection. Overall, 10 studies were identified with 544 total patients. Five of these studies, with 409 patients, reported at least 1 case of acute kidney injury for a total of 27 patients, giving an incidence of 6.6% in these studies.^68–71 The remaining 5 studies, totaling 135 patients, did not report any cases of acute kidney injury,^{50,61,76–78} although that was not the primary focus of any of those trials.

Most notable from this systematic review, the study of Menge et al⁶⁹ retrospectively determined the incidence of acute kidney injury (defined as a 50% rise in serum creatinine to > 1.4 mg/dL within 90 days of surgery) to be 17% in 84 patients with infected total knee arthroplasties treated with antibiotic-loaded cement spacers. A mean of 3.5 bags of cement per spacer were used in the 35 articulating spacers, compared with 2.9 per nonarticulating spacer. These spacers contained vancomycin in 82% (median 4.0 g, range 1–16 g) and tobramycin in 94% (median 4.8 g, range 1–12 g), among others in small percentages. The dose of tobramycin in the spacer considered either as a dichotomous variable (> 4.8 g, OR 5.87) or linearly (OR 1.24 per 1-g increase) was significantly associated with acute kidney injury, although systemic administration of aminoglycosides or vancomycin was not.

Additional single-center series that were published subsequent to this review have generally used more current diagnostic criteria.

Noto et al⁷² found that 10 of 46 patients treated with antibiotic-loaded cement spacers had a greater than 50% rise in serum creatinine (average increase 260%). All spacers contained tobramycin (mean dose 8.2 g), and 9 of 10 also contained vancomycin (mean 7.6 g). All of the 9 patients with acute kidney injury with follow-up data recovered renal function.

Reed et al⁷⁵ found 26 cases of acute kidney injury (based on RIFLE creatinine criteria) in 306 patients with antibiotic-loaded cement spacers treating various periprosthetic joint infections (including hips, knees, shoulders, and digits) and compared them with 74 controls who did not develop acute kidney injury. By multivariable analysis, receipt of an ACE inhibitor within 7 days of surgery and receipt of piperacillin-tazobactam within 7 days after surgery were both significantly more common in cases with acute kidney injury than in controls without acute kidney injury.

Aeng et al⁷³ prospectively studied 50 consecutive patients receiving antibiotic-loaded spacers containing tobramycin (with or without vancomycin) for treatment of infected hip or knee replacements. Using RIFLE creatinine criteria, they found an incidence of acute kidney injury of 20% (10 of 50). Factors significantly associated with acute kidney injury included cement premixed by the manufacturer with gentamicin (0.5 g per 40-g bag) in addition to the tobramycin they added, intraoperative blood transfusions, and postoperative use of nonsteroidal anti-inflammatory drugs.

Geller et al,⁷⁴ in a multicenter retrospective study of 247 patients with prosthetic joint infections (156 knees and 91 hips) undergoing antibiotic-loaded cement spacer placement, found an incidence of acute kidney injury of 26% based on KDIGO creatinine criteria. Significant risk factors included higher body mass index, lower preoperative hemoglobin level, drop in hemoglobin after surgery, and comorbidity (hypertension, diabetes, chronic kidney disease, or cardiovascular disease). Most of the spacers contained a combination of vancomycin and either tobramycin (81%) or gentamicin (13%). The spacers contained an average of 5.3 g (range 0.6–18 g) of vancomycin (average 2.65 g per 40-g bag) and an average of 5.2 g (range 0.5–16.4 g) of tobramycin (average 2.6 g per bag).

As in Menge et al,⁶⁹ this study illustrates the wide range of antibiotic dosages in use and the lack of standardization. In contrast to the study by Menge et al, however, development of acute kidney injury was not related to the amount of vancomycin or tobramycin contained in the spacers. Eventual clearance of infection (at 1 and 2 years) was significantly related to increasing amounts of vancomycin. Multiple different systemic antibiotics were used, most commonly vancomycin (44%), and systemic vancomycin was not associated with acute kidney injury.

Yadav et al,⁸¹ in a study of 3,129 consecutive revision procedures of the knee or hip, found an incidence of acute kidney injury by RIFLE creatinine criteria of 29% in the 197 patients who received antibiotic-loaded cement spacers for periprosthetic joint infection compared with 3.4% in the 2,848 who underwent revision for aseptic reasons. In 84 patients with prosthetic joint infection having various surgeries not including placement of a spacer, the acute kidney injury rate at some point in their course was an alarmingly high 82%. In the group that received spacers, only age and comorbidity as assessed by Charlson comorbidity index were independently associated with acute kidney injury by multivariate analysis. Surprisingly, modest renal impairment was protective, possibly because physicians of patients with chronic kidney disease were more vigilant and took appropriate measures to prevent acute kidney injury.

Overall, the risk of acute kidney injury appears to be much higher during treatment of prosthetic joint infection with a 2-stage procedure using an antibiotic-loaded cement spacer than after primary total joint arthroplasty or revision for aseptic reasons, and may complicate up to one-third of cases.

As in primary total joint arthroplasty in general, higher-risk cases should be identified based on age, body mass index, chronic kidney disease, comorbidities (hypertension, diabetes, established cardiovascular disease), and anemia.

Preoperative transfusion can be considered case by case depending on degree of anemia and associated risk factors.

All renin-angiotensin-aldosterone system inhibitors should be withheld starting 1 week before surgery.

Both nonselective and cyclooxygenase-2 selective nonsteroidal anti-inflammatory drugs should be avoided, if possible.

Strict attention should be paid to adequate intraoperative and postoperative fluid resuscitation.

Kidney function should be monitored closely in the early postoperative period, including urine output and daily creatinine for at least 72 hours.

Systemic administration of potentially nephrotoxic antibiotics should be minimized, especially the combination of vancomycin with piperacillin-tazobactam.⁸⁴ Daptomycin is a consideration.⁴³

If acute kidney injury should develop, serum levels of vancomycin or aminoglycosides should be measured if the spacer contains these antibiotics. The spacer may need to be removed if toxic serum levels persist.

TAKE-HOME POINTS

Acute kidney injury may complicate up to 10% of primary lower-extremity total joint arthroplasties and up to 25% of periprosthetic joint infections treated with a 2-stage procedure including placement of an antibiotic-loaded cement spacer in the first stage.

Risk factors for acute kidney injury include older age, obesity, chronic kidney disease, and overall comorbidity. Potentially modifiable risk factors include anemia, perioperative transfusions, aminoglycoside prophylaxis, perioperative renin-angiotensin system blockade, and postoperative nonsteroidal anti-inflammatory drugs. These should be mitigated when possible.

In patients with periprosthetic joint infection who receive antibiotic-loaded cement spacers, especially patients  with additional risk factors for acute kidney injury, strict attention should be paid to the dose of antibiotic in the spacer, with levels checked postoperatively if necessary. Nonnephrotoxic antibiotics should be chosen for systemic administration when possible.

Prospective randomized controlled trials are needed to guide therapy after total joint arthroplasty, and to verify the adverse long-term outcomes of acute kidney injury in this setting.

To the Editor: Regarding the article about a man with rapidly progressive pleural effusion by Zoumot et al in the January 2019 issue,¹ there was some inconsistency between the teaching points and the actions taken.

Question 1 asked what was the most likely cause of the patient’s pleuritic chest pain. Pulmonary embolism was an unlikely diagnosis, given the patient’s presentation and his normal D-dimer level, which the text acknowledges, but then proceeds to state that computed tomographic angiography of the chest was done anyway.

After pleural effusion was diagnosed, question 2 asked what was the best management strategy for the patient at that time. The best management strategy was to give oral antibiotics with close follow-up because the patient was at low risk of a poor outcome, but he was advised to be admitted for intravenous antibiotics anyway.

I’m not quite sure of the point of the didactic exercise when actions are not consistent with the analytic rationale for testing and treatment.

To the Editor: Regarding the article about a man with rapidly progressive pleural effusion by Zoumot et al in the January 2019 issue,¹ there was some inconsistency between the teaching points and the actions taken.

Question 1 asked what was the most likely cause of the patient’s pleuritic chest pain. Pulmonary embolism was an unlikely diagnosis, given the patient’s presentation and his normal D-dimer level, which the text acknowledges, but then proceeds to state that computed tomographic angiography of the chest was done anyway.

After pleural effusion was diagnosed, question 2 asked what was the best management strategy for the patient at that time. The best management strategy was to give oral antibiotics with close follow-up because the patient was at low risk of a poor outcome, but he was advised to be admitted for intravenous antibiotics anyway.

I’m not quite sure of the point of the didactic exercise when actions are not consistent with the analytic rationale for testing and treatment.

To the Editor: Regarding the article about a man with rapidly progressive pleural effusion by Zoumot et al in the January 2019 issue,¹ there was some inconsistency between the teaching points and the actions taken.

Question 1 asked what was the most likely cause of the patient’s pleuritic chest pain. Pulmonary embolism was an unlikely diagnosis, given the patient’s presentation and his normal D-dimer level, which the text acknowledges, but then proceeds to state that computed tomographic angiography of the chest was done anyway.

After pleural effusion was diagnosed, question 2 asked what was the best management strategy for the patient at that time. The best management strategy was to give oral antibiotics with close follow-up because the patient was at low risk of a poor outcome, but he was advised to be admitted for intravenous antibiotics anyway.

I’m not quite sure of the point of the didactic exercise when actions are not consistent with the analytic rationale for testing and treatment.

In Reply: We thank Dr. Davidson for his comments. Indeed, the teaching points may appear inconsistent with the actual patient journey in this case. In the real world, physicians from different teams and specialties are involved in the care of a patient, and medical practice may not strictly adhere to guidelines.

In question 1, the emergency department physician decided to proceed with computed tomographic pulmonary angiography to rule out pulmonary embolism. Based on best practice guidelines, pulmonary angiography was not indicated, as the clinical pretest probability of pulmonary embolism was low, supported by the patient’s negative D-dimer test. When we wrote the article, as we already had the scan, we used it to support the learning points in terms of findings on computed tomography at the early stage of a developing empyema, and also to support that the scan was in fact not indicated (not the other way around).

As for question 2, specific data-driven guidelines do not exist on how best to manage patients with bronchopneumonia with an early evolving parapneumonic effusion. In the text that follows question 2, we stated that management as an inpatient or outpatient would have been reasonable. Although we considered the patient at low risk for a poor outcome, we offered inpatient admission at the time for better control of his severe pleuritic pain (this could have been made clearer in the text), as well as close monitoring of his evolving parapneumonic effusion, and we do not believe that this contradicts the teaching points of this case.

In Reply: We thank Dr. Davidson for his comments. Indeed, the teaching points may appear inconsistent with the actual patient journey in this case. In the real world, physicians from different teams and specialties are involved in the care of a patient, and medical practice may not strictly adhere to guidelines.

In question 1, the emergency department physician decided to proceed with computed tomographic pulmonary angiography to rule out pulmonary embolism. Based on best practice guidelines, pulmonary angiography was not indicated, as the clinical pretest probability of pulmonary embolism was low, supported by the patient’s negative D-dimer test. When we wrote the article, as we already had the scan, we used it to support the learning points in terms of findings on computed tomography at the early stage of a developing empyema, and also to support that the scan was in fact not indicated (not the other way around).

As for question 2, specific data-driven guidelines do not exist on how best to manage patients with bronchopneumonia with an early evolving parapneumonic effusion. In the text that follows question 2, we stated that management as an inpatient or outpatient would have been reasonable. Although we considered the patient at low risk for a poor outcome, we offered inpatient admission at the time for better control of his severe pleuritic pain (this could have been made clearer in the text), as well as close monitoring of his evolving parapneumonic effusion, and we do not believe that this contradicts the teaching points of this case.

In Reply: We thank Dr. Davidson for his comments. Indeed, the teaching points may appear inconsistent with the actual patient journey in this case. In the real world, physicians from different teams and specialties are involved in the care of a patient, and medical practice may not strictly adhere to guidelines.

In question 1, the emergency department physician decided to proceed with computed tomographic pulmonary angiography to rule out pulmonary embolism. Based on best practice guidelines, pulmonary angiography was not indicated, as the clinical pretest probability of pulmonary embolism was low, supported by the patient’s negative D-dimer test. When we wrote the article, as we already had the scan, we used it to support the learning points in terms of findings on computed tomography at the early stage of a developing empyema, and also to support that the scan was in fact not indicated (not the other way around).

As for question 2, specific data-driven guidelines do not exist on how best to manage patients with bronchopneumonia with an early evolving parapneumonic effusion. In the text that follows question 2, we stated that management as an inpatient or outpatient would have been reasonable. Although we considered the patient at low risk for a poor outcome, we offered inpatient admission at the time for better control of his severe pleuritic pain (this could have been made clearer in the text), as well as close monitoring of his evolving parapneumonic effusion, and we do not believe that this contradicts the teaching points of this case.