Promotion through the ranks is the hallmark of success in academia. The support and infrastructure necessary to develop junior faculty members at academic medical centers may be inadequate.1, 2 Academic hospitalists are particularly vulnerable and at high risk for failure because of their heavy clinical commitment and limited time to pursue scholarly interests. Further, relatively few have pursued fellowship training, which means that many hospitalists must learn research‐related skills and the nuances of academia after joining the faculty.

Top‐notch mentors are believed to be integral to the success of the academic physician.36 Among other responsibilities, mentors (1) direct mentees toward promising opportunities, (2) serve as advocates for mentees, and (3) lend expertise to mentees' studies and scholarship. In general, there is concern that the cadre of talented, committed, and capable mentors is dwindling such that they are insufficient in number to satisfy and support the needs of the faculty.7, 8 In hospital medicine, experienced mentorship is particularly in short supply because the field is relatively new and there has been tremendous growth in the number of academic hospitalists, producing a large demand.

Like many hospitalist groups, our hospitalist division, the Collaborative Inpatient Medicine Service (CIMS), has experienced significant growth. It became apparent that the faculty needed and deserved a well‐designed academic support program to foster the development of skills necessary for academic success. The remainder of this article discusses our approach toward fulfilling these needs and the results to date.

DEVELOPING THE HOSPITALIST ACADEMIC SUPPORT PROGRAM

EVALUATION DATA

In the 2 years since implementation of the scholarly support program, individual faculty in the CIMS have been meeting the above‐mentioned goals. Specifically, with respect to acquiring knowledge and skills, 2 faculty members have completed their master's degrees, and 6 others have made use of select courses to augment their knowledge and skills. All faculty members (100%) have a scholarly project they are leading, and most have reached out to a colleague in the CIMS to assist them, such that nearly all are team members on at least 1 other scholarly project. Through informal mentoring sessions and a once‐yearly formal meeting related to academic promotion, all members (100%) of the faculty are aware of the expectations and requirements for promotion.

Table 1 shows the accomplishment of the 5 faculty members in the academic track who have been division members for 3 years or more. Among the 5 faculty in the academic track, publications and extramural funding are improving. In the 5 years before the initiative, CIMS faculty averaged approximately 0.5 publications per person per year; in the first 2 years of this initiative, that number has increased to 1.3 publications per person per year. The 1 physician who has not yet been published has completed projects and has several article in process. External funding (largely in the form of 3 extramural grants from private foundations) has increased dramatically from an average of 4% per FTE before the intervention to approximately 15% per FTE afterward. In addition, all faculty members have secured a source of additional funding to reduce their clinical efforts since the implementation of this program. One foundation funded project that involved all division members, whose goal was to develop mechanisms to improve the discharge process of elderly patients to their homes, won the award at the SGIM 2007 National Meeting for the best clinical innovation. As illustrated in Table 1, 1 of the founding CIMS members transferred out of the academic track in 2003 in alignment with this physician's personal and professional goals and preferences. Two faculty members have moved up an academic rank, and several others are poised to do so.

Thus, the divisional objectives (increasing number of publications, securing funding to increase the time devoted to scholarship, new leadership roles, and progression toward promotion) are being met as well.

CONCLUSIONS

Our rapidly growing hospitalist division recognized that several factors threatened the ability of the division and individuals to succeed academically. Divisional, departmental, and medical center leadership was committed to creating a supportive structure that would be available to all hospitalists as opposed to expecting each individual to unearth the necessary resources on their own. The innovative approach to foster individual, and therefore divisional, academic and scholarly success was designed around the following strategies: retention of an expert mentor (who is a not a hospitalist) and securing 20% of his time, investing in scholarship by protecting 30% nonclinical time for academic pursuits, and attempting to seek out fellowship‐trained hospitalists when hiring.

Although quality mentorship, protected time, and recruiting the best‐available talent to fill needs may not seem all that innovative, we believe the systematic approach to the problem and our steadfast application of the strategic plan is unique, innovative, and may present a model to be emulated by other divisions. Some may contend that it is impossible to protect 30% FTE of academic hospitalists indefinitely. Our group has made substantial investment in supporting the academic pursuits of our physicians, and we believe this is essential to maintaining their satisfaction and commitment to scholarship. This amount of protected time is offered to the entire physician faculty and continues even as our division has almost tripled in size. This initiative represents a carefully calculated investment that has influenced our ability to recruit and retain excellent people. Ongoing prospective study of this intervention over time will provide additional perspective on its value and shortcomings. Nonetheless, early data suggest that the plan is indeed working and that our group is satisfied with the return on investment to date.

Promotion through the ranks is the hallmark of success in academia. The support and infrastructure necessary to develop junior faculty members at academic medical centers may be inadequate.1, 2 Academic hospitalists are particularly vulnerable and at high risk for failure because of their heavy clinical commitment and limited time to pursue scholarly interests. Further, relatively few have pursued fellowship training, which means that many hospitalists must learn research‐related skills and the nuances of academia after joining the faculty.

Top‐notch mentors are believed to be integral to the success of the academic physician.36 Among other responsibilities, mentors (1) direct mentees toward promising opportunities, (2) serve as advocates for mentees, and (3) lend expertise to mentees' studies and scholarship. In general, there is concern that the cadre of talented, committed, and capable mentors is dwindling such that they are insufficient in number to satisfy and support the needs of the faculty.7, 8 In hospital medicine, experienced mentorship is particularly in short supply because the field is relatively new and there has been tremendous growth in the number of academic hospitalists, producing a large demand.

Like many hospitalist groups, our hospitalist division, the Collaborative Inpatient Medicine Service (CIMS), has experienced significant growth. It became apparent that the faculty needed and deserved a well‐designed academic support program to foster the development of skills necessary for academic success. The remainder of this article discusses our approach toward fulfilling these needs and the results to date.

DEVELOPING THE HOSPITALIST ACADEMIC SUPPORT PROGRAM

EVALUATION DATA

In the 2 years since implementation of the scholarly support program, individual faculty in the CIMS have been meeting the above‐mentioned goals. Specifically, with respect to acquiring knowledge and skills, 2 faculty members have completed their master's degrees, and 6 others have made use of select courses to augment their knowledge and skills. All faculty members (100%) have a scholarly project they are leading, and most have reached out to a colleague in the CIMS to assist them, such that nearly all are team members on at least 1 other scholarly project. Through informal mentoring sessions and a once‐yearly formal meeting related to academic promotion, all members (100%) of the faculty are aware of the expectations and requirements for promotion.

Table 1 shows the accomplishment of the 5 faculty members in the academic track who have been division members for 3 years or more. Among the 5 faculty in the academic track, publications and extramural funding are improving. In the 5 years before the initiative, CIMS faculty averaged approximately 0.5 publications per person per year; in the first 2 years of this initiative, that number has increased to 1.3 publications per person per year. The 1 physician who has not yet been published has completed projects and has several article in process. External funding (largely in the form of 3 extramural grants from private foundations) has increased dramatically from an average of 4% per FTE before the intervention to approximately 15% per FTE afterward. In addition, all faculty members have secured a source of additional funding to reduce their clinical efforts since the implementation of this program. One foundation funded project that involved all division members, whose goal was to develop mechanisms to improve the discharge process of elderly patients to their homes, won the award at the SGIM 2007 National Meeting for the best clinical innovation. As illustrated in Table 1, 1 of the founding CIMS members transferred out of the academic track in 2003 in alignment with this physician's personal and professional goals and preferences. Two faculty members have moved up an academic rank, and several others are poised to do so.

Thus, the divisional objectives (increasing number of publications, securing funding to increase the time devoted to scholarship, new leadership roles, and progression toward promotion) are being met as well.

CONCLUSIONS

Our rapidly growing hospitalist division recognized that several factors threatened the ability of the division and individuals to succeed academically. Divisional, departmental, and medical center leadership was committed to creating a supportive structure that would be available to all hospitalists as opposed to expecting each individual to unearth the necessary resources on their own. The innovative approach to foster individual, and therefore divisional, academic and scholarly success was designed around the following strategies: retention of an expert mentor (who is a not a hospitalist) and securing 20% of his time, investing in scholarship by protecting 30% nonclinical time for academic pursuits, and attempting to seek out fellowship‐trained hospitalists when hiring.

Although quality mentorship, protected time, and recruiting the best‐available talent to fill needs may not seem all that innovative, we believe the systematic approach to the problem and our steadfast application of the strategic plan is unique, innovative, and may present a model to be emulated by other divisions. Some may contend that it is impossible to protect 30% FTE of academic hospitalists indefinitely. Our group has made substantial investment in supporting the academic pursuits of our physicians, and we believe this is essential to maintaining their satisfaction and commitment to scholarship. This amount of protected time is offered to the entire physician faculty and continues even as our division has almost tripled in size. This initiative represents a carefully calculated investment that has influenced our ability to recruit and retain excellent people. Ongoing prospective study of this intervention over time will provide additional perspective on its value and shortcomings. Nonetheless, early data suggest that the plan is indeed working and that our group is satisfied with the return on investment to date.

Promotion through the ranks is the hallmark of success in academia. The support and infrastructure necessary to develop junior faculty members at academic medical centers may be inadequate.1, 2 Academic hospitalists are particularly vulnerable and at high risk for failure because of their heavy clinical commitment and limited time to pursue scholarly interests. Further, relatively few have pursued fellowship training, which means that many hospitalists must learn research‐related skills and the nuances of academia after joining the faculty.

Top‐notch mentors are believed to be integral to the success of the academic physician.36 Among other responsibilities, mentors (1) direct mentees toward promising opportunities, (2) serve as advocates for mentees, and (3) lend expertise to mentees' studies and scholarship. In general, there is concern that the cadre of talented, committed, and capable mentors is dwindling such that they are insufficient in number to satisfy and support the needs of the faculty.7, 8 In hospital medicine, experienced mentorship is particularly in short supply because the field is relatively new and there has been tremendous growth in the number of academic hospitalists, producing a large demand.

Like many hospitalist groups, our hospitalist division, the Collaborative Inpatient Medicine Service (CIMS), has experienced significant growth. It became apparent that the faculty needed and deserved a well‐designed academic support program to foster the development of skills necessary for academic success. The remainder of this article discusses our approach toward fulfilling these needs and the results to date.

DEVELOPING THE HOSPITALIST ACADEMIC SUPPORT PROGRAM

EVALUATION DATA

In the 2 years since implementation of the scholarly support program, individual faculty in the CIMS have been meeting the above‐mentioned goals. Specifically, with respect to acquiring knowledge and skills, 2 faculty members have completed their master's degrees, and 6 others have made use of select courses to augment their knowledge and skills. All faculty members (100%) have a scholarly project they are leading, and most have reached out to a colleague in the CIMS to assist them, such that nearly all are team members on at least 1 other scholarly project. Through informal mentoring sessions and a once‐yearly formal meeting related to academic promotion, all members (100%) of the faculty are aware of the expectations and requirements for promotion.

Table 1 shows the accomplishment of the 5 faculty members in the academic track who have been division members for 3 years or more. Among the 5 faculty in the academic track, publications and extramural funding are improving. In the 5 years before the initiative, CIMS faculty averaged approximately 0.5 publications per person per year; in the first 2 years of this initiative, that number has increased to 1.3 publications per person per year. The 1 physician who has not yet been published has completed projects and has several article in process. External funding (largely in the form of 3 extramural grants from private foundations) has increased dramatically from an average of 4% per FTE before the intervention to approximately 15% per FTE afterward. In addition, all faculty members have secured a source of additional funding to reduce their clinical efforts since the implementation of this program. One foundation funded project that involved all division members, whose goal was to develop mechanisms to improve the discharge process of elderly patients to their homes, won the award at the SGIM 2007 National Meeting for the best clinical innovation. As illustrated in Table 1, 1 of the founding CIMS members transferred out of the academic track in 2003 in alignment with this physician's personal and professional goals and preferences. Two faculty members have moved up an academic rank, and several others are poised to do so.

Thus, the divisional objectives (increasing number of publications, securing funding to increase the time devoted to scholarship, new leadership roles, and progression toward promotion) are being met as well.

CONCLUSIONS

Our rapidly growing hospitalist division recognized that several factors threatened the ability of the division and individuals to succeed academically. Divisional, departmental, and medical center leadership was committed to creating a supportive structure that would be available to all hospitalists as opposed to expecting each individual to unearth the necessary resources on their own. The innovative approach to foster individual, and therefore divisional, academic and scholarly success was designed around the following strategies: retention of an expert mentor (who is a not a hospitalist) and securing 20% of his time, investing in scholarship by protecting 30% nonclinical time for academic pursuits, and attempting to seek out fellowship‐trained hospitalists when hiring.

Although quality mentorship, protected time, and recruiting the best‐available talent to fill needs may not seem all that innovative, we believe the systematic approach to the problem and our steadfast application of the strategic plan is unique, innovative, and may present a model to be emulated by other divisions. Some may contend that it is impossible to protect 30% FTE of academic hospitalists indefinitely. Our group has made substantial investment in supporting the academic pursuits of our physicians, and we believe this is essential to maintaining their satisfaction and commitment to scholarship. This amount of protected time is offered to the entire physician faculty and continues even as our division has almost tripled in size. This initiative represents a carefully calculated investment that has influenced our ability to recruit and retain excellent people. Ongoing prospective study of this intervention over time will provide additional perspective on its value and shortcomings. Nonetheless, early data suggest that the plan is indeed working and that our group is satisfied with the return on investment to date.

Clinicians commonly use renin‐angiotensin‐aldosterone‐system (RAAS) antagonists such as angiotensin‐converting enzyme inhibitors (ACEIs) and angiotensin II receptor subtype 1 antagonists (ARAs) to treat hypertension, congestive heart failure, and diabetic nephropathy. Hospitalists and other clinicians involved in the preoperative care of patients treated chronically with these agents are faced with the uncertainty of whether to continue these medications immediately prior to surgery.

The concern among those who recommend holding therapy is that pharmacologic suppression of the RAAS in patients undergoing general anesthesia may lead to severe or refractory (to intravenous fluid support) hypotension requiring vasopressors. On the other hand, if complications are no more likely when continuing one of these agents up to the day of surgery, withholding it could represent an unnecessary and potentially harmful intervention (eg, when a clinician caring postoperatively for a patient forgets to restart it). Although several studies have attempted to address this dilemma, a systematic and comprehensive summary of the pertinent evidence has not been published.

In this systematic review and meta‐analysis, we sought to summarize the best available evidence about the relative incidence of patient‐important outcomes1 in patients who do or do not receive ACEI/ARA therapy on the day of their nonemergent surgery.

METHODS

We report this protocol‐driven review in accordance with the Quality of Reporting of Meta‐analyses (QUOROM) standards for reporting systematic reviews of randomized trials.2

RESULTS

Search Results

The 509 titles reviewed included 410 titles produced by electronic searches and an additional 99 titles from other sources (Fig. 1).

Figure 1

Meta‐analyses

Pooled results suggested that patients receiving the immediate preoperative ACEI/ARA dose were more likely (RR 1.51, 95% CI 1.14‐2.01) to develop hypotension requiring vasopressors at or shortly after induction of anesthesia (Fig. 2A). There was important inconsistency between studies (I² = 59%). The pooled effect derived from randomized trials (RR = 2.26, 95% CI 0.84‐6.12) seemed greater than that derived from the 2 observational studies (RR = 1.33, 95% CI 1.02‐1.73), but the treatment‐study design interaction was not significant (P = .3). Similarly, other subgroup explorations were not contributory.

Figure 2

The incidence of perioperative myocardial infarction was not significantly different between continuing and withheld groups (Fig. 2B); the results were consistent across trials (I² = 0%) but were imprecise (RR = 0.41, 95% CI 0.07‐2.53). Data were insufficient for subgroup analyses.

DISCUSSION

Statement of Principle Findings

Our systematic review identified 3 randomized trials and 2 observational studies examining the clinical consequences of continuing versus deliberately withholding the immediate preoperative dose of a renin‐angiotensin‐aldosterone system antagonist in patients treated chronically with these agents and scheduled to undergo nonemergent surgery.

Results from pooled estimates suggest that continuing chronic therapy up until surgery may increase the risk of perioperative hypotension requiring vasopressors (Fig. 3). Otherwise, this systematic review did not identify any clinically significant consequences associated either with preoperatively withholding or continuing RAAS antagonists. We do note that all 3 of the myocardial infarctions reported occurred in patients from whom the immediate preoperative ACEI/ARA dose was withheld, although no meaningful conclusion can be inferred from so few data points.

Figure 3

Clinicians commonly use renin‐angiotensin‐aldosterone‐system (RAAS) antagonists such as angiotensin‐converting enzyme inhibitors (ACEIs) and angiotensin II receptor subtype 1 antagonists (ARAs) to treat hypertension, congestive heart failure, and diabetic nephropathy. Hospitalists and other clinicians involved in the preoperative care of patients treated chronically with these agents are faced with the uncertainty of whether to continue these medications immediately prior to surgery.

The concern among those who recommend holding therapy is that pharmacologic suppression of the RAAS in patients undergoing general anesthesia may lead to severe or refractory (to intravenous fluid support) hypotension requiring vasopressors. On the other hand, if complications are no more likely when continuing one of these agents up to the day of surgery, withholding it could represent an unnecessary and potentially harmful intervention (eg, when a clinician caring postoperatively for a patient forgets to restart it). Although several studies have attempted to address this dilemma, a systematic and comprehensive summary of the pertinent evidence has not been published.

In this systematic review and meta‐analysis, we sought to summarize the best available evidence about the relative incidence of patient‐important outcomes1 in patients who do or do not receive ACEI/ARA therapy on the day of their nonemergent surgery.

METHODS

We report this protocol‐driven review in accordance with the Quality of Reporting of Meta‐analyses (QUOROM) standards for reporting systematic reviews of randomized trials.2

RESULTS

Search Results

The 509 titles reviewed included 410 titles produced by electronic searches and an additional 99 titles from other sources (Fig. 1).

Figure 1

Meta‐analyses

Pooled results suggested that patients receiving the immediate preoperative ACEI/ARA dose were more likely (RR 1.51, 95% CI 1.14‐2.01) to develop hypotension requiring vasopressors at or shortly after induction of anesthesia (Fig. 2A). There was important inconsistency between studies (I² = 59%). The pooled effect derived from randomized trials (RR = 2.26, 95% CI 0.84‐6.12) seemed greater than that derived from the 2 observational studies (RR = 1.33, 95% CI 1.02‐1.73), but the treatment‐study design interaction was not significant (P = .3). Similarly, other subgroup explorations were not contributory.

Figure 2

The incidence of perioperative myocardial infarction was not significantly different between continuing and withheld groups (Fig. 2B); the results were consistent across trials (I² = 0%) but were imprecise (RR = 0.41, 95% CI 0.07‐2.53). Data were insufficient for subgroup analyses.

DISCUSSION

Statement of Principle Findings

Our systematic review identified 3 randomized trials and 2 observational studies examining the clinical consequences of continuing versus deliberately withholding the immediate preoperative dose of a renin‐angiotensin‐aldosterone system antagonist in patients treated chronically with these agents and scheduled to undergo nonemergent surgery.

Results from pooled estimates suggest that continuing chronic therapy up until surgery may increase the risk of perioperative hypotension requiring vasopressors (Fig. 3). Otherwise, this systematic review did not identify any clinically significant consequences associated either with preoperatively withholding or continuing RAAS antagonists. We do note that all 3 of the myocardial infarctions reported occurred in patients from whom the immediate preoperative ACEI/ARA dose was withheld, although no meaningful conclusion can be inferred from so few data points.

Figure 3

Clinicians commonly use renin‐angiotensin‐aldosterone‐system (RAAS) antagonists such as angiotensin‐converting enzyme inhibitors (ACEIs) and angiotensin II receptor subtype 1 antagonists (ARAs) to treat hypertension, congestive heart failure, and diabetic nephropathy. Hospitalists and other clinicians involved in the preoperative care of patients treated chronically with these agents are faced with the uncertainty of whether to continue these medications immediately prior to surgery.

The concern among those who recommend holding therapy is that pharmacologic suppression of the RAAS in patients undergoing general anesthesia may lead to severe or refractory (to intravenous fluid support) hypotension requiring vasopressors. On the other hand, if complications are no more likely when continuing one of these agents up to the day of surgery, withholding it could represent an unnecessary and potentially harmful intervention (eg, when a clinician caring postoperatively for a patient forgets to restart it). Although several studies have attempted to address this dilemma, a systematic and comprehensive summary of the pertinent evidence has not been published.

In this systematic review and meta‐analysis, we sought to summarize the best available evidence about the relative incidence of patient‐important outcomes1 in patients who do or do not receive ACEI/ARA therapy on the day of their nonemergent surgery.

METHODS

We report this protocol‐driven review in accordance with the Quality of Reporting of Meta‐analyses (QUOROM) standards for reporting systematic reviews of randomized trials.2

RESULTS

Search Results

The 509 titles reviewed included 410 titles produced by electronic searches and an additional 99 titles from other sources (Fig. 1).

Figure 1

Meta‐analyses

Pooled results suggested that patients receiving the immediate preoperative ACEI/ARA dose were more likely (RR 1.51, 95% CI 1.14‐2.01) to develop hypotension requiring vasopressors at or shortly after induction of anesthesia (Fig. 2A). There was important inconsistency between studies (I² = 59%). The pooled effect derived from randomized trials (RR = 2.26, 95% CI 0.84‐6.12) seemed greater than that derived from the 2 observational studies (RR = 1.33, 95% CI 1.02‐1.73), but the treatment‐study design interaction was not significant (P = .3). Similarly, other subgroup explorations were not contributory.

Figure 2

The incidence of perioperative myocardial infarction was not significantly different between continuing and withheld groups (Fig. 2B); the results were consistent across trials (I² = 0%) but were imprecise (RR = 0.41, 95% CI 0.07‐2.53). Data were insufficient for subgroup analyses.

DISCUSSION

Statement of Principle Findings

Our systematic review identified 3 randomized trials and 2 observational studies examining the clinical consequences of continuing versus deliberately withholding the immediate preoperative dose of a renin‐angiotensin‐aldosterone system antagonist in patients treated chronically with these agents and scheduled to undergo nonemergent surgery.

Results from pooled estimates suggest that continuing chronic therapy up until surgery may increase the risk of perioperative hypotension requiring vasopressors (Fig. 3). Otherwise, this systematic review did not identify any clinically significant consequences associated either with preoperatively withholding or continuing RAAS antagonists. We do note that all 3 of the myocardial infarctions reported occurred in patients from whom the immediate preoperative ACEI/ARA dose was withheld, although no meaningful conclusion can be inferred from so few data points.

Figure 3

Reduction of undesirable variation in care has been a major focus of systematic efforts to improve the quality of the healthcare system.13 The emergence of hospitalists, physicians specializing in the care of hospitalized patients, was spurred by a desire to streamline care and reduce variability in hospital management of common diseases.4, 5 Over the past decade, hospitalist systems have become a leading vehicle for care delivery.4, 6, 7 It remains unclear, however, whether implementation of hospitalist systems has lessened undesirable variation in the inpatient management of common diseases.

While systematic reviews have found costs and hospital length of stay to be 10‐15% lower in both pediatric and internal medicine hospitalist systems, few studies have adequately assessed the processes or quality of care in hospitalist systems.8, 9 Two internal medicine studies have found decreased mortality in hospitalist systems, but the mechanism by which hospitalists apparently achieved these gains is unclear.10, 11 Even less is known about care processes or quality in pediatric hospitalist systems. Death is a rare occurrence in pediatric ward settings, and the seven studies conducted to date comparing pediatric hospitalist and traditional systems have been universally underpowered to detect differences in mortality.9, 1218 There is a need to better understand care processes as a first step in understanding and improving quality of care in hospitalist systems.19

The Pediatric Research in Inpatient Settings (PRIS) Network was formed to improve the quality of care for hospitalized children through collaborative clinical research. In this study, we sought to study variation in the care of common pediatric conditions among a cohort of pediatric hospitalists. We have previously reported that less variability exists in hospitalists' reported management of inpatient conditions than in the reported management of these same conditions by community‐based pediatricians,20 but we were concerned that substantial undesirable variation (ie, variation in practice due to uncertainty or unsubstantiated local practice traditions, rather than justified variation in care based on different risks of harms or benefits in different patients) may still exist among hospitalists. We therefore conducted a study: 1) to investigate variation in hospitalists' reported use of common inpatient therapies, and 2) to test the hypothesis that greater variation exists in hospitalists' reported use of inpatient therapies of unproven benefit than in those therapies proven to be beneficial.

METHODS

RESULTS

213 of the 320 individuals identified through the 3 lists of pediatric hospitalists (67%) responded to the survey. Of these, 198 (93%) identified themselves as hospitalists and were therefore included. As previously reported,20 53% of respondents were male, 55% worked in academic training environments, and 47% had completed advanced training (fellowship) beyond their core pediatric training (residency training); respondents reported completing residency training 11 9 (mean, standard deviation) years prior to the survey, and spending 176 72 days per year in the care of hospitalized patients.

Variation in reported management: asthma

(Figure 1, Panel A). Relatively little variation existed in reported use of the 4 asthma therapies studied. Only 4.4% (95% CI, 1.4‐7.4%) of respondents did not provide the reference response of using inhaled albuterol often or almost always in the care of inpatients with asthma, and only 6.0% (2.5‐9.5%) of respondents did not report using systemic corticosteroids often or almost always. Variation in reported use of ipratropium was somewhat higher.

Figure 1

Proven vs. Unproven Therapies

(Figure 2). Variation in reported use of therapies of unproven benefit was significantly higher than variation in reported use of the 4 proven therapies (albuterol, corticosteroids, and ipratropium in the first 24 h for asthma; IV re‐hydration for gastroenteritis). The mean variation in reported use of unproven therapies was 44.6 20.5%, compared with 15.5 12.5% variation in reported use of therapies of proven benefit (p = 0.02).

Figure 2

As a sensitivity analysis, the use of thickened feeds as a therapy for GERD was re‐categorized as proven and the above analysis repeated, for the reasons outlined in the methods section. This did not alter the identified relationship between variability and the evidence base fundamentally; hospitalists' reported variation in use of therapies of unproven benefit in this sensitivity analysis was 44.6 21.7%, compared with 21.4 17.0% variation in reported use of proven therapies (p = 0.05).

DISCUSSION

Substantial variation exists in the inpatient management of common pediatric diseases. Although we have previously found less reported variability in pediatric hospitalists' practices than in those of community‐based pediatricians,20 the current study demonstrates a high degree of reported variation even among a cohort of inpatient specialists. Importantly, however, reported variation was found to be significantly less for those inpatient therapies supported by a robust evidence base.

Bronchiolitis, gastroenteritis, asthma, and GERD are extremely common causes of pediatric hospitalization throughout the developed world.2125 Our finding of high reported variability in the routine care of all of these conditions except asthma is concerning, as it suggests that experts do not agree on how to manage children hospitalized with even the most common childhood diseases. While we hypothesized that there would be some variation in the use of therapies whose benefit has not been well established, the high degree of variation observed is of concern because it indicates that an insufficient evidentiary base exists to support much of our day‐to‐day practice. Some variation in practice in response to differing clinical presentations is both expected and desirable, but it is remarkable that variance in practice was significantly less for the most evidence‐based therapies than for those grounded less firmly in science, suggesting that the variation identified here is not justifiable variation based on appropriate responses to atypical clinical presentations, but uncertainty in the absence of clear data. Such undesired variability may decrease system reliability (introducing avoidable opportunity for error),26 and lead to under‐use of needed therapies as well as overuse of unnecessary therapies.1

Our work extends prior research that has identified wide variation in patterns of hospital admission, use of hospital resources, and processes of inpatient care,2732 by documenting reported variation in the use of common inpatient therapies. Rates of hospital admission may vary by as much as 7‐fold across regions.33 Our study demonstrates that wide variation exists not only in admission rates, but in reported inpatient care processes for some of the most common diseases seen in pediatric hospitals. Our study also supports the hypothesis that variation in care may be driven by gaps in knowledge.32 Among hospitalists, we found the strength of the evidence base to be a major determinant of reported variability.

Our study has several limitations. First, the data presented here are derived from provider self‐reports, which may not fully reflect actual practice. In the case of the few proven therapies studied, reporting bias could lead to an over‐reporting of adherence to evidence‐based standards of care. Like our study, however, prior studies have found that hospital‐based providers fairly consistently comply with evidence‐based practice recommendations for acute asthma care,34, 35 supporting our finding that variation in acute asthma care (which represented 3 of our 4 proven therapies) is low in this setting.

Another limitation is that classifications of therapies as proven or unproven change as the evidence base evolves. Of particular relevance to this study, the use of thickened feeds as a therapy for GERD, originally classified as being of unknown effectiveness, was reclassified by Clinical Evidence during the course of the study as likely to be beneficial. The relationship we identified between proven therapies and degree of variability in care did not change when we conducted a sensitivity analysis re‐categorizing this therapy as proven, but precisely quantifying variation is complicated by continuous changes in the state of the evidence.

Pediatric hospitalist systems have been found consistently to improve the efficiency of care,9 yet this study suggests that considerable variation in hospitalists' management of key conditions remains. The Pediatric Research in Inpatient Settings (PRIS) Network was formed in 2002 to improve the care of hospitalized children and the quality of inpatient practice by developing an evidence base for inpatient pediatric care. Ongoing multi‐center research efforts through PRIS and other research networks are beginning to critically evaluate therapies used in the management of common pediatric conditions. Rigorous studies of the processes and outcomes of pediatric hospital care will inform inpatient pediatric practice, and ultimately improve the care of hospitalized children. The current study strongly affirms the urgent need to establish such an evidence base. Without data to inform optimal care, efforts to reduce undesirable variation in care and improve care quality cannot be fully realized.

Reduction of undesirable variation in care has been a major focus of systematic efforts to improve the quality of the healthcare system.13 The emergence of hospitalists, physicians specializing in the care of hospitalized patients, was spurred by a desire to streamline care and reduce variability in hospital management of common diseases.4, 5 Over the past decade, hospitalist systems have become a leading vehicle for care delivery.4, 6, 7 It remains unclear, however, whether implementation of hospitalist systems has lessened undesirable variation in the inpatient management of common diseases.

While systematic reviews have found costs and hospital length of stay to be 10‐15% lower in both pediatric and internal medicine hospitalist systems, few studies have adequately assessed the processes or quality of care in hospitalist systems.8, 9 Two internal medicine studies have found decreased mortality in hospitalist systems, but the mechanism by which hospitalists apparently achieved these gains is unclear.10, 11 Even less is known about care processes or quality in pediatric hospitalist systems. Death is a rare occurrence in pediatric ward settings, and the seven studies conducted to date comparing pediatric hospitalist and traditional systems have been universally underpowered to detect differences in mortality.9, 1218 There is a need to better understand care processes as a first step in understanding and improving quality of care in hospitalist systems.19

The Pediatric Research in Inpatient Settings (PRIS) Network was formed to improve the quality of care for hospitalized children through collaborative clinical research. In this study, we sought to study variation in the care of common pediatric conditions among a cohort of pediatric hospitalists. We have previously reported that less variability exists in hospitalists' reported management of inpatient conditions than in the reported management of these same conditions by community‐based pediatricians,20 but we were concerned that substantial undesirable variation (ie, variation in practice due to uncertainty or unsubstantiated local practice traditions, rather than justified variation in care based on different risks of harms or benefits in different patients) may still exist among hospitalists. We therefore conducted a study: 1) to investigate variation in hospitalists' reported use of common inpatient therapies, and 2) to test the hypothesis that greater variation exists in hospitalists' reported use of inpatient therapies of unproven benefit than in those therapies proven to be beneficial.

METHODS

RESULTS

213 of the 320 individuals identified through the 3 lists of pediatric hospitalists (67%) responded to the survey. Of these, 198 (93%) identified themselves as hospitalists and were therefore included. As previously reported,20 53% of respondents were male, 55% worked in academic training environments, and 47% had completed advanced training (fellowship) beyond their core pediatric training (residency training); respondents reported completing residency training 11 9 (mean, standard deviation) years prior to the survey, and spending 176 72 days per year in the care of hospitalized patients.

Variation in reported management: asthma

(Figure 1, Panel A). Relatively little variation existed in reported use of the 4 asthma therapies studied. Only 4.4% (95% CI, 1.4‐7.4%) of respondents did not provide the reference response of using inhaled albuterol often or almost always in the care of inpatients with asthma, and only 6.0% (2.5‐9.5%) of respondents did not report using systemic corticosteroids often or almost always. Variation in reported use of ipratropium was somewhat higher.

Figure 1

Proven vs. Unproven Therapies

(Figure 2). Variation in reported use of therapies of unproven benefit was significantly higher than variation in reported use of the 4 proven therapies (albuterol, corticosteroids, and ipratropium in the first 24 h for asthma; IV re‐hydration for gastroenteritis). The mean variation in reported use of unproven therapies was 44.6 20.5%, compared with 15.5 12.5% variation in reported use of therapies of proven benefit (p = 0.02).

Figure 2

As a sensitivity analysis, the use of thickened feeds as a therapy for GERD was re‐categorized as proven and the above analysis repeated, for the reasons outlined in the methods section. This did not alter the identified relationship between variability and the evidence base fundamentally; hospitalists' reported variation in use of therapies of unproven benefit in this sensitivity analysis was 44.6 21.7%, compared with 21.4 17.0% variation in reported use of proven therapies (p = 0.05).

DISCUSSION

Substantial variation exists in the inpatient management of common pediatric diseases. Although we have previously found less reported variability in pediatric hospitalists' practices than in those of community‐based pediatricians,20 the current study demonstrates a high degree of reported variation even among a cohort of inpatient specialists. Importantly, however, reported variation was found to be significantly less for those inpatient therapies supported by a robust evidence base.

Bronchiolitis, gastroenteritis, asthma, and GERD are extremely common causes of pediatric hospitalization throughout the developed world.2125 Our finding of high reported variability in the routine care of all of these conditions except asthma is concerning, as it suggests that experts do not agree on how to manage children hospitalized with even the most common childhood diseases. While we hypothesized that there would be some variation in the use of therapies whose benefit has not been well established, the high degree of variation observed is of concern because it indicates that an insufficient evidentiary base exists to support much of our day‐to‐day practice. Some variation in practice in response to differing clinical presentations is both expected and desirable, but it is remarkable that variance in practice was significantly less for the most evidence‐based therapies than for those grounded less firmly in science, suggesting that the variation identified here is not justifiable variation based on appropriate responses to atypical clinical presentations, but uncertainty in the absence of clear data. Such undesired variability may decrease system reliability (introducing avoidable opportunity for error),26 and lead to under‐use of needed therapies as well as overuse of unnecessary therapies.1

Our work extends prior research that has identified wide variation in patterns of hospital admission, use of hospital resources, and processes of inpatient care,2732 by documenting reported variation in the use of common inpatient therapies. Rates of hospital admission may vary by as much as 7‐fold across regions.33 Our study demonstrates that wide variation exists not only in admission rates, but in reported inpatient care processes for some of the most common diseases seen in pediatric hospitals. Our study also supports the hypothesis that variation in care may be driven by gaps in knowledge.32 Among hospitalists, we found the strength of the evidence base to be a major determinant of reported variability.

Our study has several limitations. First, the data presented here are derived from provider self‐reports, which may not fully reflect actual practice. In the case of the few proven therapies studied, reporting bias could lead to an over‐reporting of adherence to evidence‐based standards of care. Like our study, however, prior studies have found that hospital‐based providers fairly consistently comply with evidence‐based practice recommendations for acute asthma care,34, 35 supporting our finding that variation in acute asthma care (which represented 3 of our 4 proven therapies) is low in this setting.

Another limitation is that classifications of therapies as proven or unproven change as the evidence base evolves. Of particular relevance to this study, the use of thickened feeds as a therapy for GERD, originally classified as being of unknown effectiveness, was reclassified by Clinical Evidence during the course of the study as likely to be beneficial. The relationship we identified between proven therapies and degree of variability in care did not change when we conducted a sensitivity analysis re‐categorizing this therapy as proven, but precisely quantifying variation is complicated by continuous changes in the state of the evidence.

Pediatric hospitalist systems have been found consistently to improve the efficiency of care,9 yet this study suggests that considerable variation in hospitalists' management of key conditions remains. The Pediatric Research in Inpatient Settings (PRIS) Network was formed in 2002 to improve the care of hospitalized children and the quality of inpatient practice by developing an evidence base for inpatient pediatric care. Ongoing multi‐center research efforts through PRIS and other research networks are beginning to critically evaluate therapies used in the management of common pediatric conditions. Rigorous studies of the processes and outcomes of pediatric hospital care will inform inpatient pediatric practice, and ultimately improve the care of hospitalized children. The current study strongly affirms the urgent need to establish such an evidence base. Without data to inform optimal care, efforts to reduce undesirable variation in care and improve care quality cannot be fully realized.

Reduction of undesirable variation in care has been a major focus of systematic efforts to improve the quality of the healthcare system.13 The emergence of hospitalists, physicians specializing in the care of hospitalized patients, was spurred by a desire to streamline care and reduce variability in hospital management of common diseases.4, 5 Over the past decade, hospitalist systems have become a leading vehicle for care delivery.4, 6, 7 It remains unclear, however, whether implementation of hospitalist systems has lessened undesirable variation in the inpatient management of common diseases.

While systematic reviews have found costs and hospital length of stay to be 10‐15% lower in both pediatric and internal medicine hospitalist systems, few studies have adequately assessed the processes or quality of care in hospitalist systems.8, 9 Two internal medicine studies have found decreased mortality in hospitalist systems, but the mechanism by which hospitalists apparently achieved these gains is unclear.10, 11 Even less is known about care processes or quality in pediatric hospitalist systems. Death is a rare occurrence in pediatric ward settings, and the seven studies conducted to date comparing pediatric hospitalist and traditional systems have been universally underpowered to detect differences in mortality.9, 1218 There is a need to better understand care processes as a first step in understanding and improving quality of care in hospitalist systems.19

The Pediatric Research in Inpatient Settings (PRIS) Network was formed to improve the quality of care for hospitalized children through collaborative clinical research. In this study, we sought to study variation in the care of common pediatric conditions among a cohort of pediatric hospitalists. We have previously reported that less variability exists in hospitalists' reported management of inpatient conditions than in the reported management of these same conditions by community‐based pediatricians,20 but we were concerned that substantial undesirable variation (ie, variation in practice due to uncertainty or unsubstantiated local practice traditions, rather than justified variation in care based on different risks of harms or benefits in different patients) may still exist among hospitalists. We therefore conducted a study: 1) to investigate variation in hospitalists' reported use of common inpatient therapies, and 2) to test the hypothesis that greater variation exists in hospitalists' reported use of inpatient therapies of unproven benefit than in those therapies proven to be beneficial.

METHODS

RESULTS

213 of the 320 individuals identified through the 3 lists of pediatric hospitalists (67%) responded to the survey. Of these, 198 (93%) identified themselves as hospitalists and were therefore included. As previously reported,20 53% of respondents were male, 55% worked in academic training environments, and 47% had completed advanced training (fellowship) beyond their core pediatric training (residency training); respondents reported completing residency training 11 9 (mean, standard deviation) years prior to the survey, and spending 176 72 days per year in the care of hospitalized patients.

Variation in reported management: asthma

(Figure 1, Panel A). Relatively little variation existed in reported use of the 4 asthma therapies studied. Only 4.4% (95% CI, 1.4‐7.4%) of respondents did not provide the reference response of using inhaled albuterol often or almost always in the care of inpatients with asthma, and only 6.0% (2.5‐9.5%) of respondents did not report using systemic corticosteroids often or almost always. Variation in reported use of ipratropium was somewhat higher.

Figure 1

Proven vs. Unproven Therapies

(Figure 2). Variation in reported use of therapies of unproven benefit was significantly higher than variation in reported use of the 4 proven therapies (albuterol, corticosteroids, and ipratropium in the first 24 h for asthma; IV re‐hydration for gastroenteritis). The mean variation in reported use of unproven therapies was 44.6 20.5%, compared with 15.5 12.5% variation in reported use of therapies of proven benefit (p = 0.02).

Figure 2

As a sensitivity analysis, the use of thickened feeds as a therapy for GERD was re‐categorized as proven and the above analysis repeated, for the reasons outlined in the methods section. This did not alter the identified relationship between variability and the evidence base fundamentally; hospitalists' reported variation in use of therapies of unproven benefit in this sensitivity analysis was 44.6 21.7%, compared with 21.4 17.0% variation in reported use of proven therapies (p = 0.05).

DISCUSSION

Substantial variation exists in the inpatient management of common pediatric diseases. Although we have previously found less reported variability in pediatric hospitalists' practices than in those of community‐based pediatricians,20 the current study demonstrates a high degree of reported variation even among a cohort of inpatient specialists. Importantly, however, reported variation was found to be significantly less for those inpatient therapies supported by a robust evidence base.

Bronchiolitis, gastroenteritis, asthma, and GERD are extremely common causes of pediatric hospitalization throughout the developed world.2125 Our finding of high reported variability in the routine care of all of these conditions except asthma is concerning, as it suggests that experts do not agree on how to manage children hospitalized with even the most common childhood diseases. While we hypothesized that there would be some variation in the use of therapies whose benefit has not been well established, the high degree of variation observed is of concern because it indicates that an insufficient evidentiary base exists to support much of our day‐to‐day practice. Some variation in practice in response to differing clinical presentations is both expected and desirable, but it is remarkable that variance in practice was significantly less for the most evidence‐based therapies than for those grounded less firmly in science, suggesting that the variation identified here is not justifiable variation based on appropriate responses to atypical clinical presentations, but uncertainty in the absence of clear data. Such undesired variability may decrease system reliability (introducing avoidable opportunity for error),26 and lead to under‐use of needed therapies as well as overuse of unnecessary therapies.1

Our work extends prior research that has identified wide variation in patterns of hospital admission, use of hospital resources, and processes of inpatient care,2732 by documenting reported variation in the use of common inpatient therapies. Rates of hospital admission may vary by as much as 7‐fold across regions.33 Our study demonstrates that wide variation exists not only in admission rates, but in reported inpatient care processes for some of the most common diseases seen in pediatric hospitals. Our study also supports the hypothesis that variation in care may be driven by gaps in knowledge.32 Among hospitalists, we found the strength of the evidence base to be a major determinant of reported variability.

Our study has several limitations. First, the data presented here are derived from provider self‐reports, which may not fully reflect actual practice. In the case of the few proven therapies studied, reporting bias could lead to an over‐reporting of adherence to evidence‐based standards of care. Like our study, however, prior studies have found that hospital‐based providers fairly consistently comply with evidence‐based practice recommendations for acute asthma care,34, 35 supporting our finding that variation in acute asthma care (which represented 3 of our 4 proven therapies) is low in this setting.

Another limitation is that classifications of therapies as proven or unproven change as the evidence base evolves. Of particular relevance to this study, the use of thickened feeds as a therapy for GERD, originally classified as being of unknown effectiveness, was reclassified by Clinical Evidence during the course of the study as likely to be beneficial. The relationship we identified between proven therapies and degree of variability in care did not change when we conducted a sensitivity analysis re‐categorizing this therapy as proven, but precisely quantifying variation is complicated by continuous changes in the state of the evidence.

Pediatric hospitalist systems have been found consistently to improve the efficiency of care,9 yet this study suggests that considerable variation in hospitalists' management of key conditions remains. The Pediatric Research in Inpatient Settings (PRIS) Network was formed in 2002 to improve the care of hospitalized children and the quality of inpatient practice by developing an evidence base for inpatient pediatric care. Ongoing multi‐center research efforts through PRIS and other research networks are beginning to critically evaluate therapies used in the management of common pediatric conditions. Rigorous studies of the processes and outcomes of pediatric hospital care will inform inpatient pediatric practice, and ultimately improve the care of hospitalized children. The current study strongly affirms the urgent need to establish such an evidence base. Without data to inform optimal care, efforts to reduce undesirable variation in care and improve care quality cannot be fully realized.

A 60 year‐old woman with advanced kidney disease presented with one month of progressively worsening, sharp burning pain and decreased sensation in her left hand. Cold air exacerbated the pain. She noted decreasing ability to utilize her left fingers, a weakened grip and that the muscles in her hand looked smaller.

Localized sensory and motor symptoms in a discrete region of a single limb suggest neuropathy. The lack of symptoms in the face or ipsilateral lower extremity would dissuade a clinician from considering central etiologies; the presence of neuropathic pain is uncommon for cortical lesions. The involvement of motor and sensory nerves indicates peripheral nerve involvement.

The general approach to patients with peripheral neuropathy begins with identifying the neuropathy as a mononeuropathy (involving a single nerve), a polyneuropathy (symmetric involvement of multiple nerves) or a mononeuropathy multiplex (asymmetric involvement of multiple nerves). The patient, in this case, described subacute neuropathic pain, sensory loss, and weakness in her left hand in a distribution consistent with mononeuropathy or mononeuropathy multiplex.

This patient could have carpal tunnel syndrome given its prevalence in patients with advanced renal disease. The differential diagnosis is broad, however, and includes ulnar mononeuropathy, nerve ischemia due to vasculitis or vasculopathy, lower cervical radiculopathy (though the patient does not describe neck or radicular pain), lower brachial plexopathy, and complex regional pain syndrome.

The patient was diagnosed with advanced kidney disease one year ago when biopsy revealed focal segmental glomerulosclerosis secondary to lithium. Since her diagnosis, two grafts were placed in the left upper arm in anticipation of dialysis: the first, placed seven months prior to this admission, failed to mature; the second, placed one month prior to this admission, was complicated by bleeding at the fistula site and was not yet mature. Prior to this admission she had not required hemodialysis. Her past history included hypertension, dyslipidemia, hypothyroidism, secondary hyperparathyroidism, a remote history of cervical cancer (stage unknown, recent PAP smear negative), microcytosis and schizoaffective disorder. Her medications were furosemide, amlodipine, lisinopril, atenolol, atorvastatin, pantoprazole, olanzapine, levothyroxine, iron, darbepoetin, sevelamer, multivitamin and docusate.

Given the history of procedures in the left arm one should consider ischemic injury to the left median nerve. Other local complications could include compressive lesions such as an abscess or hematoma or direct nerve injury from the procedure. Carpal tunnel syndrome remains high on the differential due to its prevalence and because renal failure and hypothyroidism increase the risk of carpal tunnel syndrome.

A careful physical examination would localize the nerve or nerves involved. Examination findings that would be consistent with carpal tunnel syndrome include sensory loss in the distribution of the left median nerve, weakness of muscles innervated by the median nerve, including the abductor pollicus brevis and opponens muscles, and a Tinels and Phalens sign of the left wrist. A proximal median neuropathy resulting from ischemia or compression might also involve median‐innervated forearm muscle such as the pronator teres (forearm pronation) and flexor carpi radialis (hand flexion and abduction) muscles. Complex regional pain syndrome can be seen after a traumatic injury or surgery and is typified by severe neuropathic pain in a limb, often in combination with trophic changes in the affected extremity. A cervical radiculopathy would affect muscles supplied by the injured nerve root. For example, a C8 radiculopathy would affect all of the intrinsic hand muscles, the wrist and finger extensors, and the triceps brachii. A lower brachial plexopathy would present similarly to a lower cervical radiculopathy on clinical examination; electrodiagnostic evaluation would be necessary to distinguish these two disorders.

The patient appeared fatigued and her left hand was wrapped in blankets. Her vital signs were stable. There was no thyroid enlargement or lymphadenopathy. There was marked thenar, hypothenar and forearm atrophy on the left. Strength testing of her left hand demonstrated a grip strength of 2/5, finger extension and interosseous strength of 1/5, left wrist flexion and extension of 3/5; left biceps and triceps were 5/5. Sensation was mildly decreased to light touch, temperature, pain and proprioception throughout the left hand. Strength and sensation were intact in the right upper and bilateral lower extremities. Reflexes were 2+ throughout. The left radial pulse was diminished compared to the left ulnar that was 1+. The left hand was dry and cool. Laboratory evaluation revealed an elevated white blood cell count of 15,300/mm³, a hematocrit of 33 percent, mean corpuscular volume of 73 fL, and a normal platelet count. Electrolytes were consistent with advanced renal disease. The thyroid stimulating hormone level was normal.

The patient's examination reveals an injury to the sensory and motor components of the left ulnar, median, and distal radial nerves. The volar forearm wasting suggests proximal median motor nerve injury in the forearm. Because the triceps is spared, the weakness of finger and wrist extension implies distal radial motor nerve injury. The interosseous weakness is consistent with injury to the ulnar motor nerve. Weakness of grip and wrist flexion is less specific and it may be explained by injury to either the median or ulnar motor nerves. The diffuse sensory loss over the palmar and dorsal aspect of the left hand is consistent with neuropathic injury to the left median, ulnar, and radial sensory nerves. The preserved deep tendon reflexes are controlled by the musculocutaneous nerve (the biceps reflex) and branches arising from the proximal radial nerve (the triceps and brachioradialis reflexes), both of which are proximal to the apparent level of neuropathic insult.

Carpal tunnel syndrome is excluded since abnormalities extend beyond the median nerve distribution. The presentation is consistent with a mononeuropathy multiplex. Axonal etiologies of mononeuropathy multiplexincluding vasculitis, ischemia, neoplastic infiltration, and infectious etiologies such as Lyme diseaseare more common than demyelinating causes. Vasculitic neuropathy commonly involves the lower extremities and may have systemic symptoms, not present in this case. Neoplasm or other compressive lesions such as hematoma could explain these findings. Abscess must be considered given the leukocytosis. A lower brachial plexopathy, technically a mononeuropathy multiplex involving the proximal arm at the level of the brachial plexus, is also in the differential diagnosis. Another axonal disorder to consider would be neuralgic amyotrophy, an idiopathic form of acute brachial plexopathy associated with pain that is a complication of surgery that typically presents within a few hours to weeks of the procedure.

Demyelinating causes of mononeuropathy multiplex are less likely and include a variant of chronic inflammatory demyelinating polyneuropathy (Lewis‐Sumner syndrome) and hereditary neuropathy with liability to pressure palsies. Both processes are typically indolent and usually not painful, though the latter may present with fulminant numbness and weakness.

Nerve conduction and needle electromyography of the arm and cervical paraspinal muscles would differentiate axonal degeneration from demyelination. It would identify the affected nerves and any nerve root involvement. Given the concern for abscess or hematoma, MR neurography focused on the surgical site would also be important.

Electromyography and nerve conduction velocities demonstrated severe axonal loss of the left median, ulnar and distal radial sensory nerves consistent with acute denervation. A magnetic resonance neurogram following the course of these nerves revealed enlarged ulnar and median nerves with abnormal signal, but no compressive lesion (Fig. 1).

Figure 1

The electrodiagnostic testing is consistent with severe acute axonal injury to the left ulnar, median and radial nerves. This supports a diagnosis of mononeuropathy multiplex with axonal injury. Demyelinating causes are excluded at this point.

The MR neurogram demonstrates nonspecific nerve enlargement, which may be seen in ischemia, neoplastic processes (primary or metastatic), demyelinating disease, or, rarely, amyloidosis. Neoplastic involvement is unlikely in this case given the absence of a compressive mass lesion and the long segmental involvement of both the median and ulnar nerves. Compression from an abscess or hematoma is excluded. Neuralgic amyotrophy does not typically cause nerve enlargement.

Ischemia is the most likely diagnosis. Laboratory evaluation for vasculitis would be reasonable. Vasculitides that could present in this fashion include: polyarteritis nodosa, mixed connective tissue disease, Wegner's granulomatosis, Churg‐Strauss angiitis, Sjogren's, hepatitis C with serum cryoglobulinemia and possibly rheumatoid arthritis. Given the history of fistula placement in the affected limb, vascular sufficiency must be assessed.

Anti‐nuclear antibodies and anti‐neutrophilic cytoplasmic antibodies were negative, and a C‐ reactive protein was 5.9 mg/L (normal range, 0 to 10 mg/L). The erythrocyte sedimentation rate was 32mm/hr (normal range, 0 to 20) and serologies for hepatitis B and C were negative. There was no evidence of serum cryoglobulins.

A modestly elevated sedimentation rate and normal C‐reactive protein argue against a diagnosis of vasculitis. The negative ANA, ANCA, hepatitis serologies and cryoglobulin tests render unlikely the diagnoses of polyarteritis nodosa, Wegner's granulomatosis, Churg‐Strauss angiitis, or hepatitis related cryoglobulinemia. Eosinophilia (present in Churg‐Strauss), ENA (positive in mixed connective tissue disease), anti‐SSA and SSB (positive in Sjogren's) and a rheumatoid factor would round out this evaluation for vasculitis. A left radial sensory nerve biopsy could also be of value in diagnosing vasculitic neuropathy in this patient.

Given the evidence against vasculitis, the possibility of ischemia due to vascular insufficiency is concerning. Two ischemic complications of hemodialysis are known to cause distal multiple mononeuropathies. The first, ischemic monomelic neuropathy syndrome is seen almost exclusively in diabetics. It is characterized by the development of acute pain, weakness of the forearm and hand muscles, and sensory loss within minutes or hours of AV graft placement. Transient occlusion of the blood supply to the nerves of the forearm and hand induces nerve ischemia, but does not cause necrosis of other tissues. The nerve conduction findings in this patient are consistent with ischemic monomelic neuropathy syndrome. The delayed onset of her symptoms, however, makes this diagnosis unlikely.

The second ischemic complication of hemodialysis, and the likelier diagnosis, is vascular steal syndrome. This has a similar clinical and electrodiagnostic presentation to ischemic monomelic neuropathy syndrome, but has a latency period after surgery of days to months. Vascular steal occurs when a reversal of blood flow into the fistula steals flow from the palmar arch arteries and induces ischemia of the vasa nervorum. Vascular studies should be obtained urgently when this diagnosis is considered.

Evaluation of the arteriovenous graft and vascular surgery consultation were sought. Digital photoplethysmography revealed diminished waveforms in all fingers of the left hand. Arterial Doppler evaluation of the left upper extremity confirmed low‐velocity flow in the radial and ulnar arteries and failed to confirm flow in the brachial artery distal to the arteriovenous fistula. The patient underwent an angiogram of the left axillary and brachial arteries. There was normal flow until the level of the arteriovenous fistula but minimal flow distal to the fistula (Fig. 2).

Figure 2

The diminished waveforms on digital photoplethysmography are consistent with poor perfusion distally. The angiogram suggests that the multiple mononeuropathies are a consequence of ischemia from impaired blood flow.

Consulting the vascular surgeons in this setting is essential because restoring adequate blood flow to the affected nerves can prevent further loss of function. Prognosis is dependent on many factors, including the severity of the functional loss and the duration of the symptoms prior to the restoration of blood flow. The patient's severe weakness and substantial muscle atrophy, manifestations of axonal degeneration, imply a poorer prognosis for recovery of function.

Embolization of the arteriovenous fistula was performed by interventional radiology. Post embolization angiograms demonstrated improved peripheral arterial flow (Fig. 3). One day later, the patient's finger flexion and extension improved. She reported mildly decreased dysesthesias and on examination her fingers were warmer to the touch. One month after discharge, her strength continued to be impaired, though improved and she still experienced pain.

Figure 3

COMMENTARY

Hospitalists must be equipped to recognize urgent and potentially reversible causes of neuropathy. The hospitalist should maintain a high index of suspicion for ischemia (either due to vasculitis or vascular compromise), traumatic nerve injury, nerve compression or entrapment, lymphoma or metastatic infiltration, hepatitis C with cryoglobulinemia, Guillain‐Barre syndrome and toxic exposures. Table 1 highlights important causes of mononeuropathy multiplex and summarizes associated findings and indicated diagnostic tests for specific evaluation.

Another challenge for hospitalists is efficient evaluation of neuropathy. A systematic framework for creating a differential diagnosis and familiarity with available diagnostic tests is crucial. Hospitalists should be aware of three broad categories of neuropathy: mononeuropathy, polyneuropathy and mononeuropathy multiplex. Electrodiagnostic testing is essential to confirm the involved nerves and distinguishes axonal from demyelinating etiologies. Ultrasound, MR neurogram and, when indicated, nerve biopsy may be useful. Table 2 reviews these diagnostic tools as well as their indications and limitations.

Ischemic steal syndrome should be considered when neuropathy develops in a limb subsequent to arterio‐venous access procedures. Any vascular network, including the vertebral, carotid and coronary arteries, is at risk for steal. A feature common to all steal syndromes is the diversion of blood away from its original destination toward a lower pressure alternative. In some cases, this leads to a reversal of arterial flow and ischemia. Ischemic complications from AV access occur in 1‐9% of patients.1 Symptoms of steal can be mild, such as self‐limited dialysis induced pain, coldness and numbness, or severe, including severe pain, sensory and motor loss.2 If vascular compromise is sufficient, gangrene can ensue. Sensory deficits usually precede motor loss and the radial pulse is commonly absent or diminished. Other findings can include pallor of the fingers, muscle atrophy, resorption of the nail bed, and gangrene or ulcerations of the fingers. Risk factors for steal include atherosclerotic disease, female gender, age greater than 60 years, diabetes mellitus, previous surgery on the same arm, and use of the brachial artery as a donor.3 Symptoms of ischemic steal typically present within the first month after surgery, but can also be delayed; there is one report of a patient presenting one year postoperatively.4

Imaging studies such as doppler and angiography can be helpful in diagnosing ischemic steal syndrome. Fistulagrams may reveal a reversal of blood flow in the distal arm and hand, but these are reserved for cases with suspected proximal obstructive arterial disease.5 Vascular imaging studies can be misleading, however, as many patients will have physiologic but asymptomatic reversal of flow. Thus, a functional assessment such as digital plethysmography is recommended, especially in cases where clinical symptoms are vague. Digital pressures less than 60mmHg demonstrated 100% sensitivity and 87% specificity in one case control study of 40 patients.6 Treatment of ischemic steal syndrome is aimed at decreasing flow through the access shunt.

In conclusion, this case highlights the importance of timely and systematic evaluation of peripheral neuropathy in the hospital setting. Neuropathy with rapid progression and high potential for permanent damage necessitates early neurologic, or in this case, vascular consultation. Hospitalists should be facile in evaluating peripheral neuropathies and recognizing the appropriate indications for diagnostic tests and procedures.

A 60 year‐old woman with advanced kidney disease presented with one month of progressively worsening, sharp burning pain and decreased sensation in her left hand. Cold air exacerbated the pain. She noted decreasing ability to utilize her left fingers, a weakened grip and that the muscles in her hand looked smaller.

Localized sensory and motor symptoms in a discrete region of a single limb suggest neuropathy. The lack of symptoms in the face or ipsilateral lower extremity would dissuade a clinician from considering central etiologies; the presence of neuropathic pain is uncommon for cortical lesions. The involvement of motor and sensory nerves indicates peripheral nerve involvement.

The general approach to patients with peripheral neuropathy begins with identifying the neuropathy as a mononeuropathy (involving a single nerve), a polyneuropathy (symmetric involvement of multiple nerves) or a mononeuropathy multiplex (asymmetric involvement of multiple nerves). The patient, in this case, described subacute neuropathic pain, sensory loss, and weakness in her left hand in a distribution consistent with mononeuropathy or mononeuropathy multiplex.

This patient could have carpal tunnel syndrome given its prevalence in patients with advanced renal disease. The differential diagnosis is broad, however, and includes ulnar mononeuropathy, nerve ischemia due to vasculitis or vasculopathy, lower cervical radiculopathy (though the patient does not describe neck or radicular pain), lower brachial plexopathy, and complex regional pain syndrome.

The patient was diagnosed with advanced kidney disease one year ago when biopsy revealed focal segmental glomerulosclerosis secondary to lithium. Since her diagnosis, two grafts were placed in the left upper arm in anticipation of dialysis: the first, placed seven months prior to this admission, failed to mature; the second, placed one month prior to this admission, was complicated by bleeding at the fistula site and was not yet mature. Prior to this admission she had not required hemodialysis. Her past history included hypertension, dyslipidemia, hypothyroidism, secondary hyperparathyroidism, a remote history of cervical cancer (stage unknown, recent PAP smear negative), microcytosis and schizoaffective disorder. Her medications were furosemide, amlodipine, lisinopril, atenolol, atorvastatin, pantoprazole, olanzapine, levothyroxine, iron, darbepoetin, sevelamer, multivitamin and docusate.

Given the history of procedures in the left arm one should consider ischemic injury to the left median nerve. Other local complications could include compressive lesions such as an abscess or hematoma or direct nerve injury from the procedure. Carpal tunnel syndrome remains high on the differential due to its prevalence and because renal failure and hypothyroidism increase the risk of carpal tunnel syndrome.

A careful physical examination would localize the nerve or nerves involved. Examination findings that would be consistent with carpal tunnel syndrome include sensory loss in the distribution of the left median nerve, weakness of muscles innervated by the median nerve, including the abductor pollicus brevis and opponens muscles, and a Tinels and Phalens sign of the left wrist. A proximal median neuropathy resulting from ischemia or compression might also involve median‐innervated forearm muscle such as the pronator teres (forearm pronation) and flexor carpi radialis (hand flexion and abduction) muscles. Complex regional pain syndrome can be seen after a traumatic injury or surgery and is typified by severe neuropathic pain in a limb, often in combination with trophic changes in the affected extremity. A cervical radiculopathy would affect muscles supplied by the injured nerve root. For example, a C8 radiculopathy would affect all of the intrinsic hand muscles, the wrist and finger extensors, and the triceps brachii. A lower brachial plexopathy would present similarly to a lower cervical radiculopathy on clinical examination; electrodiagnostic evaluation would be necessary to distinguish these two disorders.

The patient appeared fatigued and her left hand was wrapped in blankets. Her vital signs were stable. There was no thyroid enlargement or lymphadenopathy. There was marked thenar, hypothenar and forearm atrophy on the left. Strength testing of her left hand demonstrated a grip strength of 2/5, finger extension and interosseous strength of 1/5, left wrist flexion and extension of 3/5; left biceps and triceps were 5/5. Sensation was mildly decreased to light touch, temperature, pain and proprioception throughout the left hand. Strength and sensation were intact in the right upper and bilateral lower extremities. Reflexes were 2+ throughout. The left radial pulse was diminished compared to the left ulnar that was 1+. The left hand was dry and cool. Laboratory evaluation revealed an elevated white blood cell count of 15,300/mm³, a hematocrit of 33 percent, mean corpuscular volume of 73 fL, and a normal platelet count. Electrolytes were consistent with advanced renal disease. The thyroid stimulating hormone level was normal.

The patient's examination reveals an injury to the sensory and motor components of the left ulnar, median, and distal radial nerves. The volar forearm wasting suggests proximal median motor nerve injury in the forearm. Because the triceps is spared, the weakness of finger and wrist extension implies distal radial motor nerve injury. The interosseous weakness is consistent with injury to the ulnar motor nerve. Weakness of grip and wrist flexion is less specific and it may be explained by injury to either the median or ulnar motor nerves. The diffuse sensory loss over the palmar and dorsal aspect of the left hand is consistent with neuropathic injury to the left median, ulnar, and radial sensory nerves. The preserved deep tendon reflexes are controlled by the musculocutaneous nerve (the biceps reflex) and branches arising from the proximal radial nerve (the triceps and brachioradialis reflexes), both of which are proximal to the apparent level of neuropathic insult.

Carpal tunnel syndrome is excluded since abnormalities extend beyond the median nerve distribution. The presentation is consistent with a mononeuropathy multiplex. Axonal etiologies of mononeuropathy multiplexincluding vasculitis, ischemia, neoplastic infiltration, and infectious etiologies such as Lyme diseaseare more common than demyelinating causes. Vasculitic neuropathy commonly involves the lower extremities and may have systemic symptoms, not present in this case. Neoplasm or other compressive lesions such as hematoma could explain these findings. Abscess must be considered given the leukocytosis. A lower brachial plexopathy, technically a mononeuropathy multiplex involving the proximal arm at the level of the brachial plexus, is also in the differential diagnosis. Another axonal disorder to consider would be neuralgic amyotrophy, an idiopathic form of acute brachial plexopathy associated with pain that is a complication of surgery that typically presents within a few hours to weeks of the procedure.

Demyelinating causes of mononeuropathy multiplex are less likely and include a variant of chronic inflammatory demyelinating polyneuropathy (Lewis‐Sumner syndrome) and hereditary neuropathy with liability to pressure palsies. Both processes are typically indolent and usually not painful, though the latter may present with fulminant numbness and weakness.

Nerve conduction and needle electromyography of the arm and cervical paraspinal muscles would differentiate axonal degeneration from demyelination. It would identify the affected nerves and any nerve root involvement. Given the concern for abscess or hematoma, MR neurography focused on the surgical site would also be important.

Electromyography and nerve conduction velocities demonstrated severe axonal loss of the left median, ulnar and distal radial sensory nerves consistent with acute denervation. A magnetic resonance neurogram following the course of these nerves revealed enlarged ulnar and median nerves with abnormal signal, but no compressive lesion (Fig. 1).

Figure 1

The electrodiagnostic testing is consistent with severe acute axonal injury to the left ulnar, median and radial nerves. This supports a diagnosis of mononeuropathy multiplex with axonal injury. Demyelinating causes are excluded at this point.

The MR neurogram demonstrates nonspecific nerve enlargement, which may be seen in ischemia, neoplastic processes (primary or metastatic), demyelinating disease, or, rarely, amyloidosis. Neoplastic involvement is unlikely in this case given the absence of a compressive mass lesion and the long segmental involvement of both the median and ulnar nerves. Compression from an abscess or hematoma is excluded. Neuralgic amyotrophy does not typically cause nerve enlargement.

Ischemia is the most likely diagnosis. Laboratory evaluation for vasculitis would be reasonable. Vasculitides that could present in this fashion include: polyarteritis nodosa, mixed connective tissue disease, Wegner's granulomatosis, Churg‐Strauss angiitis, Sjogren's, hepatitis C with serum cryoglobulinemia and possibly rheumatoid arthritis. Given the history of fistula placement in the affected limb, vascular sufficiency must be assessed.

Anti‐nuclear antibodies and anti‐neutrophilic cytoplasmic antibodies were negative, and a C‐ reactive protein was 5.9 mg/L (normal range, 0 to 10 mg/L). The erythrocyte sedimentation rate was 32mm/hr (normal range, 0 to 20) and serologies for hepatitis B and C were negative. There was no evidence of serum cryoglobulins.

A modestly elevated sedimentation rate and normal C‐reactive protein argue against a diagnosis of vasculitis. The negative ANA, ANCA, hepatitis serologies and cryoglobulin tests render unlikely the diagnoses of polyarteritis nodosa, Wegner's granulomatosis, Churg‐Strauss angiitis, or hepatitis related cryoglobulinemia. Eosinophilia (present in Churg‐Strauss), ENA (positive in mixed connective tissue disease), anti‐SSA and SSB (positive in Sjogren's) and a rheumatoid factor would round out this evaluation for vasculitis. A left radial sensory nerve biopsy could also be of value in diagnosing vasculitic neuropathy in this patient.

Given the evidence against vasculitis, the possibility of ischemia due to vascular insufficiency is concerning. Two ischemic complications of hemodialysis are known to cause distal multiple mononeuropathies. The first, ischemic monomelic neuropathy syndrome is seen almost exclusively in diabetics. It is characterized by the development of acute pain, weakness of the forearm and hand muscles, and sensory loss within minutes or hours of AV graft placement. Transient occlusion of the blood supply to the nerves of the forearm and hand induces nerve ischemia, but does not cause necrosis of other tissues. The nerve conduction findings in this patient are consistent with ischemic monomelic neuropathy syndrome. The delayed onset of her symptoms, however, makes this diagnosis unlikely.

The second ischemic complication of hemodialysis, and the likelier diagnosis, is vascular steal syndrome. This has a similar clinical and electrodiagnostic presentation to ischemic monomelic neuropathy syndrome, but has a latency period after surgery of days to months. Vascular steal occurs when a reversal of blood flow into the fistula steals flow from the palmar arch arteries and induces ischemia of the vasa nervorum. Vascular studies should be obtained urgently when this diagnosis is considered.

Evaluation of the arteriovenous graft and vascular surgery consultation were sought. Digital photoplethysmography revealed diminished waveforms in all fingers of the left hand. Arterial Doppler evaluation of the left upper extremity confirmed low‐velocity flow in the radial and ulnar arteries and failed to confirm flow in the brachial artery distal to the arteriovenous fistula. The patient underwent an angiogram of the left axillary and brachial arteries. There was normal flow until the level of the arteriovenous fistula but minimal flow distal to the fistula (Fig. 2).

Figure 2

The diminished waveforms on digital photoplethysmography are consistent with poor perfusion distally. The angiogram suggests that the multiple mononeuropathies are a consequence of ischemia from impaired blood flow.

Consulting the vascular surgeons in this setting is essential because restoring adequate blood flow to the affected nerves can prevent further loss of function. Prognosis is dependent on many factors, including the severity of the functional loss and the duration of the symptoms prior to the restoration of blood flow. The patient's severe weakness and substantial muscle atrophy, manifestations of axonal degeneration, imply a poorer prognosis for recovery of function.

Embolization of the arteriovenous fistula was performed by interventional radiology. Post embolization angiograms demonstrated improved peripheral arterial flow (Fig. 3). One day later, the patient's finger flexion and extension improved. She reported mildly decreased dysesthesias and on examination her fingers were warmer to the touch. One month after discharge, her strength continued to be impaired, though improved and she still experienced pain.

Figure 3

COMMENTARY

Hospitalists must be equipped to recognize urgent and potentially reversible causes of neuropathy. The hospitalist should maintain a high index of suspicion for ischemia (either due to vasculitis or vascular compromise), traumatic nerve injury, nerve compression or entrapment, lymphoma or metastatic infiltration, hepatitis C with cryoglobulinemia, Guillain‐Barre syndrome and toxic exposures. Table 1 highlights important causes of mononeuropathy multiplex and summarizes associated findings and indicated diagnostic tests for specific evaluation.

Another challenge for hospitalists is efficient evaluation of neuropathy. A systematic framework for creating a differential diagnosis and familiarity with available diagnostic tests is crucial. Hospitalists should be aware of three broad categories of neuropathy: mononeuropathy, polyneuropathy and mononeuropathy multiplex. Electrodiagnostic testing is essential to confirm the involved nerves and distinguishes axonal from demyelinating etiologies. Ultrasound, MR neurogram and, when indicated, nerve biopsy may be useful. Table 2 reviews these diagnostic tools as well as their indications and limitations.

Ischemic steal syndrome should be considered when neuropathy develops in a limb subsequent to arterio‐venous access procedures. Any vascular network, including the vertebral, carotid and coronary arteries, is at risk for steal. A feature common to all steal syndromes is the diversion of blood away from its original destination toward a lower pressure alternative. In some cases, this leads to a reversal of arterial flow and ischemia. Ischemic complications from AV access occur in 1‐9% of patients.1 Symptoms of steal can be mild, such as self‐limited dialysis induced pain, coldness and numbness, or severe, including severe pain, sensory and motor loss.2 If vascular compromise is sufficient, gangrene can ensue. Sensory deficits usually precede motor loss and the radial pulse is commonly absent or diminished. Other findings can include pallor of the fingers, muscle atrophy, resorption of the nail bed, and gangrene or ulcerations of the fingers. Risk factors for steal include atherosclerotic disease, female gender, age greater than 60 years, diabetes mellitus, previous surgery on the same arm, and use of the brachial artery as a donor.3 Symptoms of ischemic steal typically present within the first month after surgery, but can also be delayed; there is one report of a patient presenting one year postoperatively.4

Imaging studies such as doppler and angiography can be helpful in diagnosing ischemic steal syndrome. Fistulagrams may reveal a reversal of blood flow in the distal arm and hand, but these are reserved for cases with suspected proximal obstructive arterial disease.5 Vascular imaging studies can be misleading, however, as many patients will have physiologic but asymptomatic reversal of flow. Thus, a functional assessment such as digital plethysmography is recommended, especially in cases where clinical symptoms are vague. Digital pressures less than 60mmHg demonstrated 100% sensitivity and 87% specificity in one case control study of 40 patients.6 Treatment of ischemic steal syndrome is aimed at decreasing flow through the access shunt.

In conclusion, this case highlights the importance of timely and systematic evaluation of peripheral neuropathy in the hospital setting. Neuropathy with rapid progression and high potential for permanent damage necessitates early neurologic, or in this case, vascular consultation. Hospitalists should be facile in evaluating peripheral neuropathies and recognizing the appropriate indications for diagnostic tests and procedures.

A 60 year‐old woman with advanced kidney disease presented with one month of progressively worsening, sharp burning pain and decreased sensation in her left hand. Cold air exacerbated the pain. She noted decreasing ability to utilize her left fingers, a weakened grip and that the muscles in her hand looked smaller.

Localized sensory and motor symptoms in a discrete region of a single limb suggest neuropathy. The lack of symptoms in the face or ipsilateral lower extremity would dissuade a clinician from considering central etiologies; the presence of neuropathic pain is uncommon for cortical lesions. The involvement of motor and sensory nerves indicates peripheral nerve involvement.

The general approach to patients with peripheral neuropathy begins with identifying the neuropathy as a mononeuropathy (involving a single nerve), a polyneuropathy (symmetric involvement of multiple nerves) or a mononeuropathy multiplex (asymmetric involvement of multiple nerves). The patient, in this case, described subacute neuropathic pain, sensory loss, and weakness in her left hand in a distribution consistent with mononeuropathy or mononeuropathy multiplex.

This patient could have carpal tunnel syndrome given its prevalence in patients with advanced renal disease. The differential diagnosis is broad, however, and includes ulnar mononeuropathy, nerve ischemia due to vasculitis or vasculopathy, lower cervical radiculopathy (though the patient does not describe neck or radicular pain), lower brachial plexopathy, and complex regional pain syndrome.

The patient was diagnosed with advanced kidney disease one year ago when biopsy revealed focal segmental glomerulosclerosis secondary to lithium. Since her diagnosis, two grafts were placed in the left upper arm in anticipation of dialysis: the first, placed seven months prior to this admission, failed to mature; the second, placed one month prior to this admission, was complicated by bleeding at the fistula site and was not yet mature. Prior to this admission she had not required hemodialysis. Her past history included hypertension, dyslipidemia, hypothyroidism, secondary hyperparathyroidism, a remote history of cervical cancer (stage unknown, recent PAP smear negative), microcytosis and schizoaffective disorder. Her medications were furosemide, amlodipine, lisinopril, atenolol, atorvastatin, pantoprazole, olanzapine, levothyroxine, iron, darbepoetin, sevelamer, multivitamin and docusate.

Given the history of procedures in the left arm one should consider ischemic injury to the left median nerve. Other local complications could include compressive lesions such as an abscess or hematoma or direct nerve injury from the procedure. Carpal tunnel syndrome remains high on the differential due to its prevalence and because renal failure and hypothyroidism increase the risk of carpal tunnel syndrome.

A careful physical examination would localize the nerve or nerves involved. Examination findings that would be consistent with carpal tunnel syndrome include sensory loss in the distribution of the left median nerve, weakness of muscles innervated by the median nerve, including the abductor pollicus brevis and opponens muscles, and a Tinels and Phalens sign of the left wrist. A proximal median neuropathy resulting from ischemia or compression might also involve median‐innervated forearm muscle such as the pronator teres (forearm pronation) and flexor carpi radialis (hand flexion and abduction) muscles. Complex regional pain syndrome can be seen after a traumatic injury or surgery and is typified by severe neuropathic pain in a limb, often in combination with trophic changes in the affected extremity. A cervical radiculopathy would affect muscles supplied by the injured nerve root. For example, a C8 radiculopathy would affect all of the intrinsic hand muscles, the wrist and finger extensors, and the triceps brachii. A lower brachial plexopathy would present similarly to a lower cervical radiculopathy on clinical examination; electrodiagnostic evaluation would be necessary to distinguish these two disorders.

The patient appeared fatigued and her left hand was wrapped in blankets. Her vital signs were stable. There was no thyroid enlargement or lymphadenopathy. There was marked thenar, hypothenar and forearm atrophy on the left. Strength testing of her left hand demonstrated a grip strength of 2/5, finger extension and interosseous strength of 1/5, left wrist flexion and extension of 3/5; left biceps and triceps were 5/5. Sensation was mildly decreased to light touch, temperature, pain and proprioception throughout the left hand. Strength and sensation were intact in the right upper and bilateral lower extremities. Reflexes were 2+ throughout. The left radial pulse was diminished compared to the left ulnar that was 1+. The left hand was dry and cool. Laboratory evaluation revealed an elevated white blood cell count of 15,300/mm³, a hematocrit of 33 percent, mean corpuscular volume of 73 fL, and a normal platelet count. Electrolytes were consistent with advanced renal disease. The thyroid stimulating hormone level was normal.

The patient's examination reveals an injury to the sensory and motor components of the left ulnar, median, and distal radial nerves. The volar forearm wasting suggests proximal median motor nerve injury in the forearm. Because the triceps is spared, the weakness of finger and wrist extension implies distal radial motor nerve injury. The interosseous weakness is consistent with injury to the ulnar motor nerve. Weakness of grip and wrist flexion is less specific and it may be explained by injury to either the median or ulnar motor nerves. The diffuse sensory loss over the palmar and dorsal aspect of the left hand is consistent with neuropathic injury to the left median, ulnar, and radial sensory nerves. The preserved deep tendon reflexes are controlled by the musculocutaneous nerve (the biceps reflex) and branches arising from the proximal radial nerve (the triceps and brachioradialis reflexes), both of which are proximal to the apparent level of neuropathic insult.

Carpal tunnel syndrome is excluded since abnormalities extend beyond the median nerve distribution. The presentation is consistent with a mononeuropathy multiplex. Axonal etiologies of mononeuropathy multiplexincluding vasculitis, ischemia, neoplastic infiltration, and infectious etiologies such as Lyme diseaseare more common than demyelinating causes. Vasculitic neuropathy commonly involves the lower extremities and may have systemic symptoms, not present in this case. Neoplasm or other compressive lesions such as hematoma could explain these findings. Abscess must be considered given the leukocytosis. A lower brachial plexopathy, technically a mononeuropathy multiplex involving the proximal arm at the level of the brachial plexus, is also in the differential diagnosis. Another axonal disorder to consider would be neuralgic amyotrophy, an idiopathic form of acute brachial plexopathy associated with pain that is a complication of surgery that typically presents within a few hours to weeks of the procedure.

Demyelinating causes of mononeuropathy multiplex are less likely and include a variant of chronic inflammatory demyelinating polyneuropathy (Lewis‐Sumner syndrome) and hereditary neuropathy with liability to pressure palsies. Both processes are typically indolent and usually not painful, though the latter may present with fulminant numbness and weakness.

Nerve conduction and needle electromyography of the arm and cervical paraspinal muscles would differentiate axonal degeneration from demyelination. It would identify the affected nerves and any nerve root involvement. Given the concern for abscess or hematoma, MR neurography focused on the surgical site would also be important.

Electromyography and nerve conduction velocities demonstrated severe axonal loss of the left median, ulnar and distal radial sensory nerves consistent with acute denervation. A magnetic resonance neurogram following the course of these nerves revealed enlarged ulnar and median nerves with abnormal signal, but no compressive lesion (Fig. 1).

Figure 1

The electrodiagnostic testing is consistent with severe acute axonal injury to the left ulnar, median and radial nerves. This supports a diagnosis of mononeuropathy multiplex with axonal injury. Demyelinating causes are excluded at this point.

The MR neurogram demonstrates nonspecific nerve enlargement, which may be seen in ischemia, neoplastic processes (primary or metastatic), demyelinating disease, or, rarely, amyloidosis. Neoplastic involvement is unlikely in this case given the absence of a compressive mass lesion and the long segmental involvement of both the median and ulnar nerves. Compression from an abscess or hematoma is excluded. Neuralgic amyotrophy does not typically cause nerve enlargement.

Ischemia is the most likely diagnosis. Laboratory evaluation for vasculitis would be reasonable. Vasculitides that could present in this fashion include: polyarteritis nodosa, mixed connective tissue disease, Wegner's granulomatosis, Churg‐Strauss angiitis, Sjogren's, hepatitis C with serum cryoglobulinemia and possibly rheumatoid arthritis. Given the history of fistula placement in the affected limb, vascular sufficiency must be assessed.

Anti‐nuclear antibodies and anti‐neutrophilic cytoplasmic antibodies were negative, and a C‐ reactive protein was 5.9 mg/L (normal range, 0 to 10 mg/L). The erythrocyte sedimentation rate was 32mm/hr (normal range, 0 to 20) and serologies for hepatitis B and C were negative. There was no evidence of serum cryoglobulins.

A modestly elevated sedimentation rate and normal C‐reactive protein argue against a diagnosis of vasculitis. The negative ANA, ANCA, hepatitis serologies and cryoglobulin tests render unlikely the diagnoses of polyarteritis nodosa, Wegner's granulomatosis, Churg‐Strauss angiitis, or hepatitis related cryoglobulinemia. Eosinophilia (present in Churg‐Strauss), ENA (positive in mixed connective tissue disease), anti‐SSA and SSB (positive in Sjogren's) and a rheumatoid factor would round out this evaluation for vasculitis. A left radial sensory nerve biopsy could also be of value in diagnosing vasculitic neuropathy in this patient.

Given the evidence against vasculitis, the possibility of ischemia due to vascular insufficiency is concerning. Two ischemic complications of hemodialysis are known to cause distal multiple mononeuropathies. The first, ischemic monomelic neuropathy syndrome is seen almost exclusively in diabetics. It is characterized by the development of acute pain, weakness of the forearm and hand muscles, and sensory loss within minutes or hours of AV graft placement. Transient occlusion of the blood supply to the nerves of the forearm and hand induces nerve ischemia, but does not cause necrosis of other tissues. The nerve conduction findings in this patient are consistent with ischemic monomelic neuropathy syndrome. The delayed onset of her symptoms, however, makes this diagnosis unlikely.

The second ischemic complication of hemodialysis, and the likelier diagnosis, is vascular steal syndrome. This has a similar clinical and electrodiagnostic presentation to ischemic monomelic neuropathy syndrome, but has a latency period after surgery of days to months. Vascular steal occurs when a reversal of blood flow into the fistula steals flow from the palmar arch arteries and induces ischemia of the vasa nervorum. Vascular studies should be obtained urgently when this diagnosis is considered.

Evaluation of the arteriovenous graft and vascular surgery consultation were sought. Digital photoplethysmography revealed diminished waveforms in all fingers of the left hand. Arterial Doppler evaluation of the left upper extremity confirmed low‐velocity flow in the radial and ulnar arteries and failed to confirm flow in the brachial artery distal to the arteriovenous fistula. The patient underwent an angiogram of the left axillary and brachial arteries. There was normal flow until the level of the arteriovenous fistula but minimal flow distal to the fistula (Fig. 2).

Figure 2

The diminished waveforms on digital photoplethysmography are consistent with poor perfusion distally. The angiogram suggests that the multiple mononeuropathies are a consequence of ischemia from impaired blood flow.

Consulting the vascular surgeons in this setting is essential because restoring adequate blood flow to the affected nerves can prevent further loss of function. Prognosis is dependent on many factors, including the severity of the functional loss and the duration of the symptoms prior to the restoration of blood flow. The patient's severe weakness and substantial muscle atrophy, manifestations of axonal degeneration, imply a poorer prognosis for recovery of function.

Embolization of the arteriovenous fistula was performed by interventional radiology. Post embolization angiograms demonstrated improved peripheral arterial flow (Fig. 3). One day later, the patient's finger flexion and extension improved. She reported mildly decreased dysesthesias and on examination her fingers were warmer to the touch. One month after discharge, her strength continued to be impaired, though improved and she still experienced pain.

Figure 3

COMMENTARY

Hospitalists must be equipped to recognize urgent and potentially reversible causes of neuropathy. The hospitalist should maintain a high index of suspicion for ischemia (either due to vasculitis or vascular compromise), traumatic nerve injury, nerve compression or entrapment, lymphoma or metastatic infiltration, hepatitis C with cryoglobulinemia, Guillain‐Barre syndrome and toxic exposures. Table 1 highlights important causes of mononeuropathy multiplex and summarizes associated findings and indicated diagnostic tests for specific evaluation.

Another challenge for hospitalists is efficient evaluation of neuropathy. A systematic framework for creating a differential diagnosis and familiarity with available diagnostic tests is crucial. Hospitalists should be aware of three broad categories of neuropathy: mononeuropathy, polyneuropathy and mononeuropathy multiplex. Electrodiagnostic testing is essential to confirm the involved nerves and distinguishes axonal from demyelinating etiologies. Ultrasound, MR neurogram and, when indicated, nerve biopsy may be useful. Table 2 reviews these diagnostic tools as well as their indications and limitations.

Ischemic steal syndrome should be considered when neuropathy develops in a limb subsequent to arterio‐venous access procedures. Any vascular network, including the vertebral, carotid and coronary arteries, is at risk for steal. A feature common to all steal syndromes is the diversion of blood away from its original destination toward a lower pressure alternative. In some cases, this leads to a reversal of arterial flow and ischemia. Ischemic complications from AV access occur in 1‐9% of patients.1 Symptoms of steal can be mild, such as self‐limited dialysis induced pain, coldness and numbness, or severe, including severe pain, sensory and motor loss.2 If vascular compromise is sufficient, gangrene can ensue. Sensory deficits usually precede motor loss and the radial pulse is commonly absent or diminished. Other findings can include pallor of the fingers, muscle atrophy, resorption of the nail bed, and gangrene or ulcerations of the fingers. Risk factors for steal include atherosclerotic disease, female gender, age greater than 60 years, diabetes mellitus, previous surgery on the same arm, and use of the brachial artery as a donor.3 Symptoms of ischemic steal typically present within the first month after surgery, but can also be delayed; there is one report of a patient presenting one year postoperatively.4

Imaging studies such as doppler and angiography can be helpful in diagnosing ischemic steal syndrome. Fistulagrams may reveal a reversal of blood flow in the distal arm and hand, but these are reserved for cases with suspected proximal obstructive arterial disease.5 Vascular imaging studies can be misleading, however, as many patients will have physiologic but asymptomatic reversal of flow. Thus, a functional assessment such as digital plethysmography is recommended, especially in cases where clinical symptoms are vague. Digital pressures less than 60mmHg demonstrated 100% sensitivity and 87% specificity in one case control study of 40 patients.6 Treatment of ischemic steal syndrome is aimed at decreasing flow through the access shunt.

In conclusion, this case highlights the importance of timely and systematic evaluation of peripheral neuropathy in the hospital setting. Neuropathy with rapid progression and high potential for permanent damage necessitates early neurologic, or in this case, vascular consultation. Hospitalists should be facile in evaluating peripheral neuropathies and recognizing the appropriate indications for diagnostic tests and procedures.

To the seasoned intensivist, discussions with family members of critically ill patients in the intensive care unit (ICU) can be very predictable. However, this does not imply that these dialogues are straightforward or simple. Each day, we spend a significant amount of time meeting with family members at different stages of their loved one's ICU stay. Some family members are satisfied with these exchanges while others leave them distraught or in emotional shambles. At times, intensivists do not always effectively communicate with family members.13 Both the team and the families come to the ICU table with different sets of perspectives, expectations, conversational skill sets, life experiences, and tolerances for stress. These differences are magnified when the ICU course turns rocky.

We thought it useful to illustrate ICU dialogues with family members as we perceive them. We focus on the potential checkpoints where miscommunication and misunderstanding may occur throughout the roller coaster ICU experience. To this end, over the past few years, our Critical Care Medicine group has been collecting thought‐provoking comments from various family conferences. Herein, we present the dynamic phases of an ICU encounter contextually inserting relevant quotes.

The specter of a loved one lying helplessly in an ICU bed, attached to imposing machines, with tubes coming out on all sides can be quite numbing and frightening. No amount of schooling or training can really prepare a person for this emotionally taxing situation. Disbelief reigns! We frequently hear, How can a person go from being fine one day to being so sick the next? or He was shoveling snow just last week! or, She was just fine after surgery, talking, walking, and eating.

Not only is the ICU a strange and scary place, but oftentimes, in the midst of our first meeting with family members of newly admitted ICU patients, we quickly realize that they are not really sure who we are or what we do. It seems to us that Critical Care Medicine as a medical specialty suffers from a lack of brand recognition. Often the family members say, You're a what? An intensivist? We've never heard of an intensivist. Do you also work in the Emergency Room? I heard someone mention critical care, is that the same as intensive care? Or sometimes we get whacked, What about getting a real doctor, like a cardiologist or a pulmonologist!

Family members immediately find different ways to let us know how much the patient means to them. They try to impress upon us the vitality and unique nature of their loved ones in the hope that this will make us all work harder. He's a real fighter and never gives up. Or He's a young and healthy 90. Or You have to take extra care of her, she's very special, she's the mother of eight children. Sometimes political or social connections are used to further incentivize or push the ICU team. He's best friends with Mr. Z who is on the Board of Trustees, or She's friends with this or that politician.

But, Google has really altered the nature of our family discussions; everyone, it seems, can now be a doctor. We used to hear I'm not a doctor, but Now, the inevitable internet search leads to I've been doing some reading on the web about this new drug and I've heard that it's a wonder drug. Why isn't my mother getting it? Or What is the APACHE III score of my sister and how are you using this value?

Just as intensivists regularly look at reams of lab data and calculate all types of organ failure and prognostication scores, family members similarly reframe the ICU discussion to a numbers game. This approach to seemingly getting our arms around the complicated big picture is used in many aspects of our lives, whether tracking our retirement portfolios or determining the odds of next week's football game. Doctor, what are her chances for improvement ‐ 50/50, 30/70 or 80/20? Even one in a million? Over time, family members even become experts at looking at the bedside monitors and devices. I've been watching the numbers, I see that the heart rate is down and you were able to decrease the oxygen on the respirator to 80%, and the blood oxygen is still over 90%, so my father must be better, right?

As the days go by, some families become more desperate. They seek good news or even any news from every person they meet. And the ICU environment certainly offers family members a myriad of people with whom to converse. Unfortunately, this frantic search for information leads them to receive conflicting and unreliable data. Frustration results, Why do we keep getting mixed messages? Or, Doctor, we like the hospital, but why can't all of you get the story straight? As family members gather together with other patients' families in the waiting room day and night, they often share their ICU stories with each other. We'll overhear someone say, What about the new antibiotic the patient in Bed 8 is getting? Shouldn't my husband be getting that too? I see everyone is gowning up, I keep hearing about that bad Staph bug going around, my father better not catch it.

Occasionally we become concerned, even perplexed, when we cannot successfully convey our message to the family despite our best intentions and efforts. Some families are just in denial; the reality of no progress and/or likely poor outcome cannot be heard, much less accepted. Doctor, I have been badgered with the truth enough. I just don't want to believe it. Or, Doctor, don't you ever have any good news to report to us? The insatiable need for prolonged and repetitive family conferences may deflect time away from the care of other patients and meeting with other families.

Unfortunately, some patients get stuck in the ICU, either not improving, or just steadily deteriorating despite aggressive care. We broach treatment limitation or end of life care with the family as we realize that further ICU care is not going to be beneficial. Sometimes this news is greeted with stunned silence. The family often pleads for their loved one to be able to stay a little bit longer in the ICU, Let's just see how he does for a few more days or over the weekend. Or, We just need to buy some more time until everything turns around. Or other approaches are used to postpone the inevitable ICU transfer. Doctor, our 50th wedding anniversary is in two weeks, so let's continue to keep him in the ICU. Or, You can't discharge my mother, she averages 14.5 alarms per hour, how can the ward ever take care of her?

And then comes the quest for the miracle, Don't you believe in miracles? Haven't you ever seen someone in this condition get better? Are you giving up? We'll never give up! Or, ignoring the express wishes of the patient, Oh, Doctor, my father did have an advanced directive, but I'm not sure whether I want to give it to you. Or, the message now gets personal What would you do if this was your mother or father?

Such questions highlight the existential dichotomy of critical care. As intensivists, we sometimes have to reconcile the family members' unrealistic view of prognosis, overly hopeful expectations, and desire for endless futile ICU care with our own understandings of prognosis, goals of care, and appropriate use of ICU beds. Where, and when should we draw the line? How do we all let go?4

This collection of comments and thoughts reflects a synthesis of many different discussions conducted under diverse conditions. Thankfully, not all of the individual elements of the scenario described above occur with each patient. Family members and intensivists commonly have amicable discourse resulting in an acceptable degree of understanding and consensus regarding the prognosis and care plan. However, on occasion, things just don't go as well as hoped for, neither in clinical outcomes nor in our discussions. While effective ICU communication strategies have been designed and studied,512 even the best of these may not prevent conflict and disagreement. Nevertheless, our challenge as critical care practitioners is to ensure that our dialogues with family members are honest and direct and that we communicate in a timely, consistent and empathetic manner.

Well, onto the next family meeting!

To the seasoned intensivist, discussions with family members of critically ill patients in the intensive care unit (ICU) can be very predictable. However, this does not imply that these dialogues are straightforward or simple. Each day, we spend a significant amount of time meeting with family members at different stages of their loved one's ICU stay. Some family members are satisfied with these exchanges while others leave them distraught or in emotional shambles. At times, intensivists do not always effectively communicate with family members.13 Both the team and the families come to the ICU table with different sets of perspectives, expectations, conversational skill sets, life experiences, and tolerances for stress. These differences are magnified when the ICU course turns rocky.

We thought it useful to illustrate ICU dialogues with family members as we perceive them. We focus on the potential checkpoints where miscommunication and misunderstanding may occur throughout the roller coaster ICU experience. To this end, over the past few years, our Critical Care Medicine group has been collecting thought‐provoking comments from various family conferences. Herein, we present the dynamic phases of an ICU encounter contextually inserting relevant quotes.

The specter of a loved one lying helplessly in an ICU bed, attached to imposing machines, with tubes coming out on all sides can be quite numbing and frightening. No amount of schooling or training can really prepare a person for this emotionally taxing situation. Disbelief reigns! We frequently hear, How can a person go from being fine one day to being so sick the next? or He was shoveling snow just last week! or, She was just fine after surgery, talking, walking, and eating.

Not only is the ICU a strange and scary place, but oftentimes, in the midst of our first meeting with family members of newly admitted ICU patients, we quickly realize that they are not really sure who we are or what we do. It seems to us that Critical Care Medicine as a medical specialty suffers from a lack of brand recognition. Often the family members say, You're a what? An intensivist? We've never heard of an intensivist. Do you also work in the Emergency Room? I heard someone mention critical care, is that the same as intensive care? Or sometimes we get whacked, What about getting a real doctor, like a cardiologist or a pulmonologist!

Family members immediately find different ways to let us know how much the patient means to them. They try to impress upon us the vitality and unique nature of their loved ones in the hope that this will make us all work harder. He's a real fighter and never gives up. Or He's a young and healthy 90. Or You have to take extra care of her, she's very special, she's the mother of eight children. Sometimes political or social connections are used to further incentivize or push the ICU team. He's best friends with Mr. Z who is on the Board of Trustees, or She's friends with this or that politician.

But, Google has really altered the nature of our family discussions; everyone, it seems, can now be a doctor. We used to hear I'm not a doctor, but Now, the inevitable internet search leads to I've been doing some reading on the web about this new drug and I've heard that it's a wonder drug. Why isn't my mother getting it? Or What is the APACHE III score of my sister and how are you using this value?

Just as intensivists regularly look at reams of lab data and calculate all types of organ failure and prognostication scores, family members similarly reframe the ICU discussion to a numbers game. This approach to seemingly getting our arms around the complicated big picture is used in many aspects of our lives, whether tracking our retirement portfolios or determining the odds of next week's football game. Doctor, what are her chances for improvement ‐ 50/50, 30/70 or 80/20? Even one in a million? Over time, family members even become experts at looking at the bedside monitors and devices. I've been watching the numbers, I see that the heart rate is down and you were able to decrease the oxygen on the respirator to 80%, and the blood oxygen is still over 90%, so my father must be better, right?

As the days go by, some families become more desperate. They seek good news or even any news from every person they meet. And the ICU environment certainly offers family members a myriad of people with whom to converse. Unfortunately, this frantic search for information leads them to receive conflicting and unreliable data. Frustration results, Why do we keep getting mixed messages? Or, Doctor, we like the hospital, but why can't all of you get the story straight? As family members gather together with other patients' families in the waiting room day and night, they often share their ICU stories with each other. We'll overhear someone say, What about the new antibiotic the patient in Bed 8 is getting? Shouldn't my husband be getting that too? I see everyone is gowning up, I keep hearing about that bad Staph bug going around, my father better not catch it.

Occasionally we become concerned, even perplexed, when we cannot successfully convey our message to the family despite our best intentions and efforts. Some families are just in denial; the reality of no progress and/or likely poor outcome cannot be heard, much less accepted. Doctor, I have been badgered with the truth enough. I just don't want to believe it. Or, Doctor, don't you ever have any good news to report to us? The insatiable need for prolonged and repetitive family conferences may deflect time away from the care of other patients and meeting with other families.

Unfortunately, some patients get stuck in the ICU, either not improving, or just steadily deteriorating despite aggressive care. We broach treatment limitation or end of life care with the family as we realize that further ICU care is not going to be beneficial. Sometimes this news is greeted with stunned silence. The family often pleads for their loved one to be able to stay a little bit longer in the ICU, Let's just see how he does for a few more days or over the weekend. Or, We just need to buy some more time until everything turns around. Or other approaches are used to postpone the inevitable ICU transfer. Doctor, our 50th wedding anniversary is in two weeks, so let's continue to keep him in the ICU. Or, You can't discharge my mother, she averages 14.5 alarms per hour, how can the ward ever take care of her?

And then comes the quest for the miracle, Don't you believe in miracles? Haven't you ever seen someone in this condition get better? Are you giving up? We'll never give up! Or, ignoring the express wishes of the patient, Oh, Doctor, my father did have an advanced directive, but I'm not sure whether I want to give it to you. Or, the message now gets personal What would you do if this was your mother or father?

Such questions highlight the existential dichotomy of critical care. As intensivists, we sometimes have to reconcile the family members' unrealistic view of prognosis, overly hopeful expectations, and desire for endless futile ICU care with our own understandings of prognosis, goals of care, and appropriate use of ICU beds. Where, and when should we draw the line? How do we all let go?4

This collection of comments and thoughts reflects a synthesis of many different discussions conducted under diverse conditions. Thankfully, not all of the individual elements of the scenario described above occur with each patient. Family members and intensivists commonly have amicable discourse resulting in an acceptable degree of understanding and consensus regarding the prognosis and care plan. However, on occasion, things just don't go as well as hoped for, neither in clinical outcomes nor in our discussions. While effective ICU communication strategies have been designed and studied,512 even the best of these may not prevent conflict and disagreement. Nevertheless, our challenge as critical care practitioners is to ensure that our dialogues with family members are honest and direct and that we communicate in a timely, consistent and empathetic manner.

Well, onto the next family meeting!

To the seasoned intensivist, discussions with family members of critically ill patients in the intensive care unit (ICU) can be very predictable. However, this does not imply that these dialogues are straightforward or simple. Each day, we spend a significant amount of time meeting with family members at different stages of their loved one's ICU stay. Some family members are satisfied with these exchanges while others leave them distraught or in emotional shambles. At times, intensivists do not always effectively communicate with family members.13 Both the team and the families come to the ICU table with different sets of perspectives, expectations, conversational skill sets, life experiences, and tolerances for stress. These differences are magnified when the ICU course turns rocky.

We thought it useful to illustrate ICU dialogues with family members as we perceive them. We focus on the potential checkpoints where miscommunication and misunderstanding may occur throughout the roller coaster ICU experience. To this end, over the past few years, our Critical Care Medicine group has been collecting thought‐provoking comments from various family conferences. Herein, we present the dynamic phases of an ICU encounter contextually inserting relevant quotes.

The specter of a loved one lying helplessly in an ICU bed, attached to imposing machines, with tubes coming out on all sides can be quite numbing and frightening. No amount of schooling or training can really prepare a person for this emotionally taxing situation. Disbelief reigns! We frequently hear, How can a person go from being fine one day to being so sick the next? or He was shoveling snow just last week! or, She was just fine after surgery, talking, walking, and eating.

Not only is the ICU a strange and scary place, but oftentimes, in the midst of our first meeting with family members of newly admitted ICU patients, we quickly realize that they are not really sure who we are or what we do. It seems to us that Critical Care Medicine as a medical specialty suffers from a lack of brand recognition. Often the family members say, You're a what? An intensivist? We've never heard of an intensivist. Do you also work in the Emergency Room? I heard someone mention critical care, is that the same as intensive care? Or sometimes we get whacked, What about getting a real doctor, like a cardiologist or a pulmonologist!

Family members immediately find different ways to let us know how much the patient means to them. They try to impress upon us the vitality and unique nature of their loved ones in the hope that this will make us all work harder. He's a real fighter and never gives up. Or He's a young and healthy 90. Or You have to take extra care of her, she's very special, she's the mother of eight children. Sometimes political or social connections are used to further incentivize or push the ICU team. He's best friends with Mr. Z who is on the Board of Trustees, or She's friends with this or that politician.

But, Google has really altered the nature of our family discussions; everyone, it seems, can now be a doctor. We used to hear I'm not a doctor, but Now, the inevitable internet search leads to I've been doing some reading on the web about this new drug and I've heard that it's a wonder drug. Why isn't my mother getting it? Or What is the APACHE III score of my sister and how are you using this value?

Just as intensivists regularly look at reams of lab data and calculate all types of organ failure and prognostication scores, family members similarly reframe the ICU discussion to a numbers game. This approach to seemingly getting our arms around the complicated big picture is used in many aspects of our lives, whether tracking our retirement portfolios or determining the odds of next week's football game. Doctor, what are her chances for improvement ‐ 50/50, 30/70 or 80/20? Even one in a million? Over time, family members even become experts at looking at the bedside monitors and devices. I've been watching the numbers, I see that the heart rate is down and you were able to decrease the oxygen on the respirator to 80%, and the blood oxygen is still over 90%, so my father must be better, right?

As the days go by, some families become more desperate. They seek good news or even any news from every person they meet. And the ICU environment certainly offers family members a myriad of people with whom to converse. Unfortunately, this frantic search for information leads them to receive conflicting and unreliable data. Frustration results, Why do we keep getting mixed messages? Or, Doctor, we like the hospital, but why can't all of you get the story straight? As family members gather together with other patients' families in the waiting room day and night, they often share their ICU stories with each other. We'll overhear someone say, What about the new antibiotic the patient in Bed 8 is getting? Shouldn't my husband be getting that too? I see everyone is gowning up, I keep hearing about that bad Staph bug going around, my father better not catch it.

Occasionally we become concerned, even perplexed, when we cannot successfully convey our message to the family despite our best intentions and efforts. Some families are just in denial; the reality of no progress and/or likely poor outcome cannot be heard, much less accepted. Doctor, I have been badgered with the truth enough. I just don't want to believe it. Or, Doctor, don't you ever have any good news to report to us? The insatiable need for prolonged and repetitive family conferences may deflect time away from the care of other patients and meeting with other families.

Unfortunately, some patients get stuck in the ICU, either not improving, or just steadily deteriorating despite aggressive care. We broach treatment limitation or end of life care with the family as we realize that further ICU care is not going to be beneficial. Sometimes this news is greeted with stunned silence. The family often pleads for their loved one to be able to stay a little bit longer in the ICU, Let's just see how he does for a few more days or over the weekend. Or, We just need to buy some more time until everything turns around. Or other approaches are used to postpone the inevitable ICU transfer. Doctor, our 50th wedding anniversary is in two weeks, so let's continue to keep him in the ICU. Or, You can't discharge my mother, she averages 14.5 alarms per hour, how can the ward ever take care of her?

And then comes the quest for the miracle, Don't you believe in miracles? Haven't you ever seen someone in this condition get better? Are you giving up? We'll never give up! Or, ignoring the express wishes of the patient, Oh, Doctor, my father did have an advanced directive, but I'm not sure whether I want to give it to you. Or, the message now gets personal What would you do if this was your mother or father?

Such questions highlight the existential dichotomy of critical care. As intensivists, we sometimes have to reconcile the family members' unrealistic view of prognosis, overly hopeful expectations, and desire for endless futile ICU care with our own understandings of prognosis, goals of care, and appropriate use of ICU beds. Where, and when should we draw the line? How do we all let go?4

This collection of comments and thoughts reflects a synthesis of many different discussions conducted under diverse conditions. Thankfully, not all of the individual elements of the scenario described above occur with each patient. Family members and intensivists commonly have amicable discourse resulting in an acceptable degree of understanding and consensus regarding the prognosis and care plan. However, on occasion, things just don't go as well as hoped for, neither in clinical outcomes nor in our discussions. While effective ICU communication strategies have been designed and studied,512 even the best of these may not prevent conflict and disagreement. Nevertheless, our challenge as critical care practitioners is to ensure that our dialogues with family members are honest and direct and that we communicate in a timely, consistent and empathetic manner.

Well, onto the next family meeting!

The patient is an 86‐year‐old woman with a history of mild dementia, major depression with psychotic features, congestive heart failure, hypertension, hyperlipidemia, osteoporosis, and hypothyroidism. She presented to her primary care physician (PCP) complaining of 4 days of bilateral lower extremity edema and dyspnea on exertion. She was admitted to the hospitalist service for exacerbation of congestive heart failure.

MEDICATIONS

Donepezil, olanzapine, mirtazapine, sertraline, spironolactone, triamterene/hydrochlorothizide, simvastatin, alendronate, levothyroxine, multivitamin.

SOCIAL HISTORY

She lived alone in an independent‐living retirement apartment that provided meals but not medical care, and she was able to function independently in her activities of daily living. Her pharmacy delivered her medications via courier service, whereas visiting home nurses filled her medication box and checked on her status weekly.

HOSPITAL COURSE

Admission vitals were: heart rate, 83; blood pressure, 158/84; respiratory rate, 20; temperature, 36.4, and saturation, 95% on room air. Echocardiogram revealed intact ejection fraction, left ventricular hypertrophy, and impaired relaxation. A TSH of 6.6 demonstrated undertreated hypothyroidism. Telemetry monitoring was significant for frequent short bursts of narrow‐complex tachycardia without clear atrial activity. The etiology of her heart failure exacerbation was presumed to be paroxysmal atrial fibrillation in the setting of diastolic dysfunction. Given her mild hyperkalemia (5.1), her diuretics were changed to monotherapy with furosemide. Low‐dose beta blockade and antithrombotic therapy were started as well as increased supplementation of levothyroxine. After several days of diuresis, her potassium had normalized, she tolerated initiation of a new ACE inhibitor, and her dyspnea had resolved. On the last hospital day, her dentures were accidentally discarded with her breakfast tray, causing her great distress.

On discharge she was on 4 new medications, 2 old medications had been stopped, and 1 prior medication's dose had been increased. During medication reconciliation, the patient reported that she had not been taking olanzapine for weeks, and thus this was omitted from her home health medication orders, with instructions to discuss with her PCP on first follow‐up within the week. The patient was provided with congestive heart failure instructions and a complete medication list. Unfortunately, the day of discharge was the first day of a holiday weekend.

Case management was unavailable on the weekend; her out‐of‐state family member was unable to be reached by phone, and her usual pharmacy courier service was closed. As she did not have a friend or family member to pick up her prescriptions from an alternate pharmacy, her prescriptions were provided as handwritten scripts, called in to her pharmacy's voice mail, and written on the home health orders. The patient was discharged to her home with communication to her PCP via telephone, e‐mail, and electronic discharge summary.

POSTDISCHARGE

Medications were not delivered to the patient until the third postdischarge day. Three days after discharge, the daughter from out of state left a message for the PCP expressing concern that the patient was failingnot eating or taking any of her medications. An expedited home nursing visit was arranged. Five days after discharge, the pharmacist called the PCP stating he had not received a prescription for the beta‐blocker. Her PCP saw the patient in clinic 6 days after discharge and reconciled the medication list, restarting the olanzapine that the patient had stopped a few weeks before and the mirtazapine, which had not been restarted despite its presence on the discharge orders and patient instructions. She continued to have poor appetite and mood and was taking her medications only with great effort from her visiting nurse and staff at the retirement community. A major cause of this decline was the significant worsening of her depression brought on by hospitalization, lapses in her psychiatric medications, and emotional distress induced by the loss of her dentures during her hospital stay. She was readmitted 10 days after her discharge because she was unable to care for herself.

DISCUSSION

This case demonstrates numerous pitfalls in the transition process. Despite communication between the hospitalists and the PCP and a common electronic medical record, this patient failed the transition from the acute care hospital to the ambulatory setting. On the holiday weekend, ancillary support services were unavailable, including case management to contact her home care agency and her pharmacy to fill and deliver her new prescriptions. Despite efforts by the discharging physician, the out‐of‐town family could not be contacted. Thus, an elderly woman with cognitive impairment was left to process a new diagnosis and 7 medication changes with an unreliable mechanism to obtain her new medications.

With the rise of hospital medicine, it has increasingly been recognized that transitions represent a point of vulnerability in the care of geriatric patients. A change in physical location of care and handoffs between caregivers create the potential for error and loss of information. Prior research has demonstrated frequent quantitative and qualitative deficiencies in the information conveyed between inpatient and outpatient physicians, with direct communication occurring less than 20% of the time.1 In this case, communication occurred between the hospitalists and the outpatient physician, demonstrating that communication is just one element of successful transitions.

Components of effective care transitions have been described in the literature, including: preparation of the patient and caregiver for the transition, medication reconciliation, instructions to patient and caregiver about symptoms and signs of worsening, and an explicit follow‐up plan for tests and appointments.2 Optimally, there is interactive discussion between the hospitalist and the receiving clinician with a summary of events including an updated medication, allergy, and problem list, current advance directives, and a common plan of care.2 This case illustrates that these elements are necessary but may not be sufficient.

Some interventions have been found to be effective. A nurse‐led multidisciplinary approach to the discharge of elderly patients with congestive heart failure led to decreased readmission rates after 90 days and was found to be a cost‐saving measure.3 Similar results have been seen in geriatric patients with a variety of diagnoses in trials using advanced‐practice nurses to bridge the vulnerable period of discharge or by interventions to improve the ability of family caregivers to handle the challenges of the transition.46 Individualized attention to the unique needs of each patient and members of their social support structure, and investment in resources to do so, has the ability to decrease readmissions.

Medication errors, medication omissions, or the inability to fill medications on discharge represent a patient safety challenge. There has been increasing emphasis on medication reconciliation at admission and discharge, but in some cases the gold standard medication list is hard to determine. Electronic medical records would seem to be a natural solution to this problem, but as this case illustrates, the electronic record may not reflect the reality of patient adherence. As in this case, clarification may require another visit with the primary provider, leaving a period of time with an uncertain medication list and therefore a vulnerable patient. Access to medications after discharge was also a problem in this case, but this is not rare. A 2001 study found that 2 days after discharge from a general medicine hospital service, 1 in 5 patients had been unable to obtain all discharge medications.7 A pharmacist‐led medication reconciliation intervention in nursing home patients led to decreases in length of stay and discrepancy‐related adverse drug events. Furthermore, a follow‐up call allowed for clarification of medication questions in 25% of cases.7, 8

Patient characteristics such as depression and cognitive dysfunction have been found to affect readmission rates and are important to assess in addressing risk for poor outcomes after discharge. A 2000 study comparing readmission rates after discharge from a geriatric rehabilitation hospital found that patients with depression had an odds ratio of 3.5 for readmission compared with those without depression.9 Inadequate health literacy is also associated with decreased ability to self‐manage chronic disease and is associated with increased risk of mortality in community‐dwelling elderly such as the patient in this case.10 Asking patients or their proxies to explain their own understanding of the discharge plan can unmask comprehension issues that otherwise may go undetected.

Discharge on a weekend presents a period of critical vulnerability. Early recognition that a transition is susceptible to failure allows the events necessary for success to occur during the week when services are available. An example in the present case might include having had the pharmacy fill the prescriptions prior to the day of discharge. This does introduce a new opportunity for error in cases in which the plan of care changes but would have solved the inability to have prescriptions filled once the holiday weekend had begun.

In cases in which the usual mechanisms break down, increased effort on the part of the hospitalist can usually create a unique solution to the problem. Examples of creative solutions that did not occur in this case might include contacting the manager of the patient's retirement apartment to determine if this individual might be willing to fill prescriptions at an alternate pharmacy. A better alternative would be to change the system such that the solution is readily accessible and time efficient. Weekend availability of case management would be one such step. A means for the hospital's inpatient pharmacy to provide 2‐3 days of bridging medications would prevent weekend prescription access from affecting timely discharges of multiple patients over the course of a year. The hospitalist is in a unique position to take a leadership role in effecting system change to address these issues.

Ultimately, it is the duty of the hospitalist to take responsibility for the safety and well‐being of the patient, and if no solution can be found, it may be necessary to hold discharge or find an alternate disposition until logistical hurdles have been overcome. Indeed, for patients admitted for myocardial infarction, discharges were less likely to occur on weekends, presumably because of lack of ancillary services.11 With foresight, creative problem solving, and systems improvement, this should rarely be necessary.

Transitions continue to be a difficult time for the most vulnerable patients. Intense efforts have improved outcomes in selected populations but have not been broadly applied. Identification of patients at the highest risk, such as those with depression, poor social support, and cognitive limitations, would allow anticipation of difficult transitions and potential utilization of proven interventions, such as advanced‐practice nurses or follow‐up pharmacy contact. Processes such as these might have prevented some of the problems in this patient's discharge. Appreciation of the weekend discharge as a time of particular challenges allows barriers to be identified and solutions created during the week, when resources are still available. Attention to all elements of effective transitions should become part of the growing culture of patient safety.

The patient is an 86‐year‐old woman with a history of mild dementia, major depression with psychotic features, congestive heart failure, hypertension, hyperlipidemia, osteoporosis, and hypothyroidism. She presented to her primary care physician (PCP) complaining of 4 days of bilateral lower extremity edema and dyspnea on exertion. She was admitted to the hospitalist service for exacerbation of congestive heart failure.

MEDICATIONS

Donepezil, olanzapine, mirtazapine, sertraline, spironolactone, triamterene/hydrochlorothizide, simvastatin, alendronate, levothyroxine, multivitamin.

SOCIAL HISTORY

She lived alone in an independent‐living retirement apartment that provided meals but not medical care, and she was able to function independently in her activities of daily living. Her pharmacy delivered her medications via courier service, whereas visiting home nurses filled her medication box and checked on her status weekly.

HOSPITAL COURSE

Admission vitals were: heart rate, 83; blood pressure, 158/84; respiratory rate, 20; temperature, 36.4, and saturation, 95% on room air. Echocardiogram revealed intact ejection fraction, left ventricular hypertrophy, and impaired relaxation. A TSH of 6.6 demonstrated undertreated hypothyroidism. Telemetry monitoring was significant for frequent short bursts of narrow‐complex tachycardia without clear atrial activity. The etiology of her heart failure exacerbation was presumed to be paroxysmal atrial fibrillation in the setting of diastolic dysfunction. Given her mild hyperkalemia (5.1), her diuretics were changed to monotherapy with furosemide. Low‐dose beta blockade and antithrombotic therapy were started as well as increased supplementation of levothyroxine. After several days of diuresis, her potassium had normalized, she tolerated initiation of a new ACE inhibitor, and her dyspnea had resolved. On the last hospital day, her dentures were accidentally discarded with her breakfast tray, causing her great distress.

On discharge she was on 4 new medications, 2 old medications had been stopped, and 1 prior medication's dose had been increased. During medication reconciliation, the patient reported that she had not been taking olanzapine for weeks, and thus this was omitted from her home health medication orders, with instructions to discuss with her PCP on first follow‐up within the week. The patient was provided with congestive heart failure instructions and a complete medication list. Unfortunately, the day of discharge was the first day of a holiday weekend.

Case management was unavailable on the weekend; her out‐of‐state family member was unable to be reached by phone, and her usual pharmacy courier service was closed. As she did not have a friend or family member to pick up her prescriptions from an alternate pharmacy, her prescriptions were provided as handwritten scripts, called in to her pharmacy's voice mail, and written on the home health orders. The patient was discharged to her home with communication to her PCP via telephone, e‐mail, and electronic discharge summary.

POSTDISCHARGE

Medications were not delivered to the patient until the third postdischarge day. Three days after discharge, the daughter from out of state left a message for the PCP expressing concern that the patient was failingnot eating or taking any of her medications. An expedited home nursing visit was arranged. Five days after discharge, the pharmacist called the PCP stating he had not received a prescription for the beta‐blocker. Her PCP saw the patient in clinic 6 days after discharge and reconciled the medication list, restarting the olanzapine that the patient had stopped a few weeks before and the mirtazapine, which had not been restarted despite its presence on the discharge orders and patient instructions. She continued to have poor appetite and mood and was taking her medications only with great effort from her visiting nurse and staff at the retirement community. A major cause of this decline was the significant worsening of her depression brought on by hospitalization, lapses in her psychiatric medications, and emotional distress induced by the loss of her dentures during her hospital stay. She was readmitted 10 days after her discharge because she was unable to care for herself.

DISCUSSION

This case demonstrates numerous pitfalls in the transition process. Despite communication between the hospitalists and the PCP and a common electronic medical record, this patient failed the transition from the acute care hospital to the ambulatory setting. On the holiday weekend, ancillary support services were unavailable, including case management to contact her home care agency and her pharmacy to fill and deliver her new prescriptions. Despite efforts by the discharging physician, the out‐of‐town family could not be contacted. Thus, an elderly woman with cognitive impairment was left to process a new diagnosis and 7 medication changes with an unreliable mechanism to obtain her new medications.

With the rise of hospital medicine, it has increasingly been recognized that transitions represent a point of vulnerability in the care of geriatric patients. A change in physical location of care and handoffs between caregivers create the potential for error and loss of information. Prior research has demonstrated frequent quantitative and qualitative deficiencies in the information conveyed between inpatient and outpatient physicians, with direct communication occurring less than 20% of the time.1 In this case, communication occurred between the hospitalists and the outpatient physician, demonstrating that communication is just one element of successful transitions.

Components of effective care transitions have been described in the literature, including: preparation of the patient and caregiver for the transition, medication reconciliation, instructions to patient and caregiver about symptoms and signs of worsening, and an explicit follow‐up plan for tests and appointments.2 Optimally, there is interactive discussion between the hospitalist and the receiving clinician with a summary of events including an updated medication, allergy, and problem list, current advance directives, and a common plan of care.2 This case illustrates that these elements are necessary but may not be sufficient.

Some interventions have been found to be effective. A nurse‐led multidisciplinary approach to the discharge of elderly patients with congestive heart failure led to decreased readmission rates after 90 days and was found to be a cost‐saving measure.3 Similar results have been seen in geriatric patients with a variety of diagnoses in trials using advanced‐practice nurses to bridge the vulnerable period of discharge or by interventions to improve the ability of family caregivers to handle the challenges of the transition.46 Individualized attention to the unique needs of each patient and members of their social support structure, and investment in resources to do so, has the ability to decrease readmissions.

Medication errors, medication omissions, or the inability to fill medications on discharge represent a patient safety challenge. There has been increasing emphasis on medication reconciliation at admission and discharge, but in some cases the gold standard medication list is hard to determine. Electronic medical records would seem to be a natural solution to this problem, but as this case illustrates, the electronic record may not reflect the reality of patient adherence. As in this case, clarification may require another visit with the primary provider, leaving a period of time with an uncertain medication list and therefore a vulnerable patient. Access to medications after discharge was also a problem in this case, but this is not rare. A 2001 study found that 2 days after discharge from a general medicine hospital service, 1 in 5 patients had been unable to obtain all discharge medications.7 A pharmacist‐led medication reconciliation intervention in nursing home patients led to decreases in length of stay and discrepancy‐related adverse drug events. Furthermore, a follow‐up call allowed for clarification of medication questions in 25% of cases.7, 8

Patient characteristics such as depression and cognitive dysfunction have been found to affect readmission rates and are important to assess in addressing risk for poor outcomes after discharge. A 2000 study comparing readmission rates after discharge from a geriatric rehabilitation hospital found that patients with depression had an odds ratio of 3.5 for readmission compared with those without depression.9 Inadequate health literacy is also associated with decreased ability to self‐manage chronic disease and is associated with increased risk of mortality in community‐dwelling elderly such as the patient in this case.10 Asking patients or their proxies to explain their own understanding of the discharge plan can unmask comprehension issues that otherwise may go undetected.

Discharge on a weekend presents a period of critical vulnerability. Early recognition that a transition is susceptible to failure allows the events necessary for success to occur during the week when services are available. An example in the present case might include having had the pharmacy fill the prescriptions prior to the day of discharge. This does introduce a new opportunity for error in cases in which the plan of care changes but would have solved the inability to have prescriptions filled once the holiday weekend had begun.

In cases in which the usual mechanisms break down, increased effort on the part of the hospitalist can usually create a unique solution to the problem. Examples of creative solutions that did not occur in this case might include contacting the manager of the patient's retirement apartment to determine if this individual might be willing to fill prescriptions at an alternate pharmacy. A better alternative would be to change the system such that the solution is readily accessible and time efficient. Weekend availability of case management would be one such step. A means for the hospital's inpatient pharmacy to provide 2‐3 days of bridging medications would prevent weekend prescription access from affecting timely discharges of multiple patients over the course of a year. The hospitalist is in a unique position to take a leadership role in effecting system change to address these issues.

Ultimately, it is the duty of the hospitalist to take responsibility for the safety and well‐being of the patient, and if no solution can be found, it may be necessary to hold discharge or find an alternate disposition until logistical hurdles have been overcome. Indeed, for patients admitted for myocardial infarction, discharges were less likely to occur on weekends, presumably because of lack of ancillary services.11 With foresight, creative problem solving, and systems improvement, this should rarely be necessary.

Transitions continue to be a difficult time for the most vulnerable patients. Intense efforts have improved outcomes in selected populations but have not been broadly applied. Identification of patients at the highest risk, such as those with depression, poor social support, and cognitive limitations, would allow anticipation of difficult transitions and potential utilization of proven interventions, such as advanced‐practice nurses or follow‐up pharmacy contact. Processes such as these might have prevented some of the problems in this patient's discharge. Appreciation of the weekend discharge as a time of particular challenges allows barriers to be identified and solutions created during the week, when resources are still available. Attention to all elements of effective transitions should become part of the growing culture of patient safety.

The patient is an 86‐year‐old woman with a history of mild dementia, major depression with psychotic features, congestive heart failure, hypertension, hyperlipidemia, osteoporosis, and hypothyroidism. She presented to her primary care physician (PCP) complaining of 4 days of bilateral lower extremity edema and dyspnea on exertion. She was admitted to the hospitalist service for exacerbation of congestive heart failure.

MEDICATIONS

Donepezil, olanzapine, mirtazapine, sertraline, spironolactone, triamterene/hydrochlorothizide, simvastatin, alendronate, levothyroxine, multivitamin.

SOCIAL HISTORY

She lived alone in an independent‐living retirement apartment that provided meals but not medical care, and she was able to function independently in her activities of daily living. Her pharmacy delivered her medications via courier service, whereas visiting home nurses filled her medication box and checked on her status weekly.

HOSPITAL COURSE

Admission vitals were: heart rate, 83; blood pressure, 158/84; respiratory rate, 20; temperature, 36.4, and saturation, 95% on room air. Echocardiogram revealed intact ejection fraction, left ventricular hypertrophy, and impaired relaxation. A TSH of 6.6 demonstrated undertreated hypothyroidism. Telemetry monitoring was significant for frequent short bursts of narrow‐complex tachycardia without clear atrial activity. The etiology of her heart failure exacerbation was presumed to be paroxysmal atrial fibrillation in the setting of diastolic dysfunction. Given her mild hyperkalemia (5.1), her diuretics were changed to monotherapy with furosemide. Low‐dose beta blockade and antithrombotic therapy were started as well as increased supplementation of levothyroxine. After several days of diuresis, her potassium had normalized, she tolerated initiation of a new ACE inhibitor, and her dyspnea had resolved. On the last hospital day, her dentures were accidentally discarded with her breakfast tray, causing her great distress.

On discharge she was on 4 new medications, 2 old medications had been stopped, and 1 prior medication's dose had been increased. During medication reconciliation, the patient reported that she had not been taking olanzapine for weeks, and thus this was omitted from her home health medication orders, with instructions to discuss with her PCP on first follow‐up within the week. The patient was provided with congestive heart failure instructions and a complete medication list. Unfortunately, the day of discharge was the first day of a holiday weekend.

Case management was unavailable on the weekend; her out‐of‐state family member was unable to be reached by phone, and her usual pharmacy courier service was closed. As she did not have a friend or family member to pick up her prescriptions from an alternate pharmacy, her prescriptions were provided as handwritten scripts, called in to her pharmacy's voice mail, and written on the home health orders. The patient was discharged to her home with communication to her PCP via telephone, e‐mail, and electronic discharge summary.

POSTDISCHARGE

Medications were not delivered to the patient until the third postdischarge day. Three days after discharge, the daughter from out of state left a message for the PCP expressing concern that the patient was failingnot eating or taking any of her medications. An expedited home nursing visit was arranged. Five days after discharge, the pharmacist called the PCP stating he had not received a prescription for the beta‐blocker. Her PCP saw the patient in clinic 6 days after discharge and reconciled the medication list, restarting the olanzapine that the patient had stopped a few weeks before and the mirtazapine, which had not been restarted despite its presence on the discharge orders and patient instructions. She continued to have poor appetite and mood and was taking her medications only with great effort from her visiting nurse and staff at the retirement community. A major cause of this decline was the significant worsening of her depression brought on by hospitalization, lapses in her psychiatric medications, and emotional distress induced by the loss of her dentures during her hospital stay. She was readmitted 10 days after her discharge because she was unable to care for herself.

DISCUSSION

This case demonstrates numerous pitfalls in the transition process. Despite communication between the hospitalists and the PCP and a common electronic medical record, this patient failed the transition from the acute care hospital to the ambulatory setting. On the holiday weekend, ancillary support services were unavailable, including case management to contact her home care agency and her pharmacy to fill and deliver her new prescriptions. Despite efforts by the discharging physician, the out‐of‐town family could not be contacted. Thus, an elderly woman with cognitive impairment was left to process a new diagnosis and 7 medication changes with an unreliable mechanism to obtain her new medications.

With the rise of hospital medicine, it has increasingly been recognized that transitions represent a point of vulnerability in the care of geriatric patients. A change in physical location of care and handoffs between caregivers create the potential for error and loss of information. Prior research has demonstrated frequent quantitative and qualitative deficiencies in the information conveyed between inpatient and outpatient physicians, with direct communication occurring less than 20% of the time.1 In this case, communication occurred between the hospitalists and the outpatient physician, demonstrating that communication is just one element of successful transitions.

Components of effective care transitions have been described in the literature, including: preparation of the patient and caregiver for the transition, medication reconciliation, instructions to patient and caregiver about symptoms and signs of worsening, and an explicit follow‐up plan for tests and appointments.2 Optimally, there is interactive discussion between the hospitalist and the receiving clinician with a summary of events including an updated medication, allergy, and problem list, current advance directives, and a common plan of care.2 This case illustrates that these elements are necessary but may not be sufficient.

Some interventions have been found to be effective. A nurse‐led multidisciplinary approach to the discharge of elderly patients with congestive heart failure led to decreased readmission rates after 90 days and was found to be a cost‐saving measure.3 Similar results have been seen in geriatric patients with a variety of diagnoses in trials using advanced‐practice nurses to bridge the vulnerable period of discharge or by interventions to improve the ability of family caregivers to handle the challenges of the transition.46 Individualized attention to the unique needs of each patient and members of their social support structure, and investment in resources to do so, has the ability to decrease readmissions.

Medication errors, medication omissions, or the inability to fill medications on discharge represent a patient safety challenge. There has been increasing emphasis on medication reconciliation at admission and discharge, but in some cases the gold standard medication list is hard to determine. Electronic medical records would seem to be a natural solution to this problem, but as this case illustrates, the electronic record may not reflect the reality of patient adherence. As in this case, clarification may require another visit with the primary provider, leaving a period of time with an uncertain medication list and therefore a vulnerable patient. Access to medications after discharge was also a problem in this case, but this is not rare. A 2001 study found that 2 days after discharge from a general medicine hospital service, 1 in 5 patients had been unable to obtain all discharge medications.7 A pharmacist‐led medication reconciliation intervention in nursing home patients led to decreases in length of stay and discrepancy‐related adverse drug events. Furthermore, a follow‐up call allowed for clarification of medication questions in 25% of cases.7, 8

Patient characteristics such as depression and cognitive dysfunction have been found to affect readmission rates and are important to assess in addressing risk for poor outcomes after discharge. A 2000 study comparing readmission rates after discharge from a geriatric rehabilitation hospital found that patients with depression had an odds ratio of 3.5 for readmission compared with those without depression.9 Inadequate health literacy is also associated with decreased ability to self‐manage chronic disease and is associated with increased risk of mortality in community‐dwelling elderly such as the patient in this case.10 Asking patients or their proxies to explain their own understanding of the discharge plan can unmask comprehension issues that otherwise may go undetected.

Discharge on a weekend presents a period of critical vulnerability. Early recognition that a transition is susceptible to failure allows the events necessary for success to occur during the week when services are available. An example in the present case might include having had the pharmacy fill the prescriptions prior to the day of discharge. This does introduce a new opportunity for error in cases in which the plan of care changes but would have solved the inability to have prescriptions filled once the holiday weekend had begun.

In cases in which the usual mechanisms break down, increased effort on the part of the hospitalist can usually create a unique solution to the problem. Examples of creative solutions that did not occur in this case might include contacting the manager of the patient's retirement apartment to determine if this individual might be willing to fill prescriptions at an alternate pharmacy. A better alternative would be to change the system such that the solution is readily accessible and time efficient. Weekend availability of case management would be one such step. A means for the hospital's inpatient pharmacy to provide 2‐3 days of bridging medications would prevent weekend prescription access from affecting timely discharges of multiple patients over the course of a year. The hospitalist is in a unique position to take a leadership role in effecting system change to address these issues.

Ultimately, it is the duty of the hospitalist to take responsibility for the safety and well‐being of the patient, and if no solution can be found, it may be necessary to hold discharge or find an alternate disposition until logistical hurdles have been overcome. Indeed, for patients admitted for myocardial infarction, discharges were less likely to occur on weekends, presumably because of lack of ancillary services.11 With foresight, creative problem solving, and systems improvement, this should rarely be necessary.

Transitions continue to be a difficult time for the most vulnerable patients. Intense efforts have improved outcomes in selected populations but have not been broadly applied. Identification of patients at the highest risk, such as those with depression, poor social support, and cognitive limitations, would allow anticipation of difficult transitions and potential utilization of proven interventions, such as advanced‐practice nurses or follow‐up pharmacy contact. Processes such as these might have prevented some of the problems in this patient's discharge. Appreciation of the weekend discharge as a time of particular challenges allows barriers to be identified and solutions created during the week, when resources are still available. Attention to all elements of effective transitions should become part of the growing culture of patient safety.

Dramatic changes in the organization, financing, and delivery of hospital care that began a decade ago continue to accelerate. One of the most important changes has been the emergence of hospitalists as providers of inpatient care.1 Hospitalists are physicians, usually general internists, whose clinical focus is the hospitalized patient. As patient illnesses have become more severe and complex, physicians have found it difficult to balance inpatient and outpatient care and have focused on one of the two.25 It is estimated that there are currently 15,000 practicing hospitalists nationally, and projections suggest that this number may exceed 30,000 by 2010, which is equal to the number of cardiologists currently practicing in the United States.6 A 2003 survey from the American Hospital Association showed that more than 30% of the nation's 4900 community hospitals have hospital medicine groups.7 Furthermore, more than 70% of the nation's largest hospitals (>500 beds) and 66% of major teaching hospitals use hospitalists.7

The transition to a hospitalist model generates multiple new research questions about the best approach to caring for the hospitalized patient. Additionally, hospitalists may spawn new areas of clinical research by tackling clinical issues that formerly lacked a large number of specialist investigators. Examples include implementation‐based studies,8, 9 inpatient safety practices,1012 quasi‐experimental studies focusing on common inpatient issues,13, 14 and the evaluation of new methods for reducing resource utilization within various inpatient care delivery structures.15, 16

Similarly, if future clinical trials are to be carried out in real‐world settings, by necessity these will require the participation of hospitalists. Clinical research performed by hospitalists and hospital medicine programs, however, remains underdeveloped. Although this has been attributed to several variables, including the youth of the field, a paucity of fellowship‐trained hospitalist researchers, and a lack of a hospitalist‐oriented national funding source, we also believe that additional barriers exist which could be overcome if hospitalists actively partnered with specialists to perform hospital‐based clinical and translational research.

Hospitalists lack clinical expertise in many clinical issues. In both academic and nonacademic settings, the diagnostic approach, individual treatment decisions, and follow‐up of complex patients occur with frequent consultation of specialists. Specialists often provide a deeper understanding of both the pathophysiologic concepts and scientific principles underlying important clinical questions and are more likely to have had fellowship training that included clinical research experience. Specialists also have more access to extramural funding for disease‐based investigation, and thus their involvement in hospital‐based clinical research would likely enhance funding opportunities, improve project feasibility, and increase dissemination of the results. A successful clinical research program will therefore be one that combines specialists and hospitalists working collaboratively to determine the best way to care for inpatients. With that in mind, we created the University of Michigan SpecialistHospitalist Allied Research Program (SHARP).

METHODS

DISCUSSION AND NEXT STEPS

SHARP is a novel clinical research program partnering hospitalists with specialists. Its current focus targets single‐institution studies that generate pilot data leading to larger projects. The ultimate goal is to develop the ability to do larger multicenter investigator‐initiated projects. The SHARP program will also have the ability to perform observational studies to identify predictors and risk factors and the ability to carry out implementation studies that show how best to translate results from published articles to direct patient care.

A specialisthospitalist collaboration overcomes barriers that we feel may impede hospital medicine research at an academic medical center. For a similar program to succeed at other institutions, key components from our program will have to be replicated. First, senior, fellowship‐trained researchers are required to mentor junior investigators (who may or may not have additional fellowship training), help guide project selection, oversee grant and manuscript submissions, and troubleshoot problems that arise in the course of any clinical research project. In our institution, this comes from within our hospitalist program and from our specialist collaborators. In institutions lacking hospitalists with research experience, this guidance could come from within a division of general medicine, internal medicine specialty divisions, internal medicine department leadership, or even noninternal medicine departments (eg, emergency medicine, neurology, and surgery) that have traditionally been involved in clinical research programs.

A second key component that must be considered is funding. An initial investment is necessary to fund key personnel dedicated to getting projects started on the right track, collecting pilot data, and ensuring project completion and dissemination of the results. The positive margin generated by our hospitalist program facilitated the initial investment. In the absence of a positive margin, resources could come directly from the hospital, the medical school, the department of internal medicine, or perhaps a foundation. The case would need to be made that an initial short‐term investment would enhance the academic standing of the institution, enhance the careers of young investigators, and over time lead to a self‐sustaining program through investigator‐initiated grants and extramural funding. In addition to experienced leadership and funding, we created an oversight committee, but we feel that this is not a critical component. A potential concern with a program that partners with specialists might be that research topics become too disease‐specific or specialty‐oriented. We specifically created the oversight committee to protect against this possibility, and other institutions might need similar safeguards.

Our next step includes leveraging existing hospitalist collaboratives that reach beyond academic medical centers to expand further the reach of SHARP. Ultimately, any new therapy, clinical tool, diagnostic paradigm, or implementation strategy that is developed or evaluated bythe SHARP program would need to be tested in a real‐world setting to assess external validity. With support from the Blue Cross Blue Shield of Michigan Foundation, we have created a multihospital patient safety consortium, the Hospitalists as Emerging Leaders in Patient Safety Consortium, which includes academic, government, urban, rural, teaching, and nonteaching hospitals.23 Although the initial focus is patient safety, our goal for the consortium is to develop it into a multihospital clinical research program that could take pilot projects developed by SHARP and test them in real‐world settings. We believe that full‐scale multihospital studies based on SHARP pilot data will be very attractive to external funding agencies and will help SHARP become financially self‐sufficient after the initial 3‐year start‐up.

Hospital medicine research is desperately needed.24, 25 Unfortunately, the clinical research capabilities of most hospital medicine programs are quite underdeveloped. We believe that partnering hospitalists with specialists can facilitate collaborative research to identify the best way to care for inpatients. If successful, we believe that variations of this model can be replicated at other institutions and will be a critical factor in jumpstarting hospital medicine clinical research.

Dramatic changes in the organization, financing, and delivery of hospital care that began a decade ago continue to accelerate. One of the most important changes has been the emergence of hospitalists as providers of inpatient care.1 Hospitalists are physicians, usually general internists, whose clinical focus is the hospitalized patient. As patient illnesses have become more severe and complex, physicians have found it difficult to balance inpatient and outpatient care and have focused on one of the two.25 It is estimated that there are currently 15,000 practicing hospitalists nationally, and projections suggest that this number may exceed 30,000 by 2010, which is equal to the number of cardiologists currently practicing in the United States.6 A 2003 survey from the American Hospital Association showed that more than 30% of the nation's 4900 community hospitals have hospital medicine groups.7 Furthermore, more than 70% of the nation's largest hospitals (>500 beds) and 66% of major teaching hospitals use hospitalists.7

The transition to a hospitalist model generates multiple new research questions about the best approach to caring for the hospitalized patient. Additionally, hospitalists may spawn new areas of clinical research by tackling clinical issues that formerly lacked a large number of specialist investigators. Examples include implementation‐based studies,8, 9 inpatient safety practices,1012 quasi‐experimental studies focusing on common inpatient issues,13, 14 and the evaluation of new methods for reducing resource utilization within various inpatient care delivery structures.15, 16

Similarly, if future clinical trials are to be carried out in real‐world settings, by necessity these will require the participation of hospitalists. Clinical research performed by hospitalists and hospital medicine programs, however, remains underdeveloped. Although this has been attributed to several variables, including the youth of the field, a paucity of fellowship‐trained hospitalist researchers, and a lack of a hospitalist‐oriented national funding source, we also believe that additional barriers exist which could be overcome if hospitalists actively partnered with specialists to perform hospital‐based clinical and translational research.

Hospitalists lack clinical expertise in many clinical issues. In both academic and nonacademic settings, the diagnostic approach, individual treatment decisions, and follow‐up of complex patients occur with frequent consultation of specialists. Specialists often provide a deeper understanding of both the pathophysiologic concepts and scientific principles underlying important clinical questions and are more likely to have had fellowship training that included clinical research experience. Specialists also have more access to extramural funding for disease‐based investigation, and thus their involvement in hospital‐based clinical research would likely enhance funding opportunities, improve project feasibility, and increase dissemination of the results. A successful clinical research program will therefore be one that combines specialists and hospitalists working collaboratively to determine the best way to care for inpatients. With that in mind, we created the University of Michigan SpecialistHospitalist Allied Research Program (SHARP).

METHODS