Literature at a Glance

A guide to this month’s studies.

• Survival of in-hospital cardiac arrest decreases during nights and weekends.
• C-reactive protein levels predict severity and complications in community-acquired pneumonia.
• Adherence to current guidelines improves outcome in treatment of prosthetic joint infection.
• Sodium phosphate bowel prep use is associated with a decline in GFR.
• Minimally interrupted cardiac resuscitation improves survival.
• Lower aPTT increases risk for future VTE.
• Warfarin combined with antiplatelet therapy increases hemorrhage rate.
• Creatinine rise during MI hospitalization is associated with long-term risk of death and ESRD.
• Communication improves patient adherence to beta-blockers after MI.

Does Time of Day Affect Survival of Cardiac Arrest Patients?

Background: In-hospital cardiac arrest is a major public health problem. Small studies have demonstrated survival after cardiac arrest is worse at night as compared with all other times. Multiple hypothesis are proffered for this decreased survival, including less effective detection and treatment of the warning signs of impending arrest during the night hours.

Study design: Prospective registry.

Setting: 507 hospitals participating through the National Registry of Cardiopulmonary Resuscitation.

Synopsis: 86,748 consecutive, inpatient cardiac arrests were reported from Jan. 1, 2000, through Feb. 1, 2007, including 58,593 cases during day/evening hours and 28,155 cases during night hours.

Rates of survival to discharge (14.7 % vs. 19.8%), survival at 24 hours (28.9% vs. 35.4%), and favorable neurological outcomes (11.0% vs. 15.2%) were substantially lower during the night compared with day/evening (all p values < 0.001). The first documented rhythm at night was more frequently asystole as opposed to ventricular fibrillation during the day/evening. There also was a higher survival rate with cardiac arrests during day/evening hours occurring on weekdays compared with weekends (odds ratio [OR] 1.15). There was no difference in survival rates between weekdays or weekends among cardiac arrests occurring during the night hours.

Bottom line: Survival rates for in-hospital cardiac arrest are lower during nights and weekends, which may relate to differential physician and hospital staffing patterns during these hours.

Citation: Peberdy MA, Ornato JP, Larkin GL et al. Survival from in-hospital cardiac arrest during nights and weekends. JAMA 2008;299:785-792.

Can CRP Identify Risk in CAP Patients?

Background: Small initial studies suggest an elevated C-reactive protein (CRP) is relatively nonspecific but may have a role in predicting disease severity in community-acquired pneumonia (CAP).

Study design: Prospective study.

Setting: Large academic center in the United Kingdom.

Synopsis: In this study of 570 patients over a two-year time period, all patients presenting to the hospital with a diagnosis of CAP and the absence of exclusion criteria were evaluated. CRP was measured on admission and repeated on day four of hospitalization.

Low CRP levels (less than 100 mg/L) were independently associated with a reduced risk of 30-day mortality (OR 0.18; p=0.03), need for invasive ventilation and/or inotropic support (OR 0.21; p=0.002), and complicated pneumonia (OR 0.05; p=0.003). In addition, the failure of CRP to fall by 50% or more at day four of hospitalization was associated with an increased risk of 30-day mortality (OR 24.5; p<0.0001), need for mechanical ventilation and/or inotropic support (OR 7.1; p<0.0001), and complicated pneumonia (OR 15.4; p<0.0001).

Patients with chronic lung disease, immunosuppression, active malignancy or hospital-acquired pneumonia were excluded from the study and the conclusions cannot be extrapolated to these higher risk populations.

Bottom line: C-reactive protein is an independent marker of severity in CAP, and low levels can be used as an adjunct to clinical judgment to help identify patients who may be safely discharged from the hospital.

Citation: Chalmers JD, Singanayagam A, Hill AT. C-reactive protein is an independent predictor of severity in community-acquired pneumonia. Am J of Med. 2008;121:219-225.

Do Aggressive Surgical Intervention and Antimicrobial Treatment Improve Outcomes in Patients Suffering PJI?

Background: Prosthetic joint infection (PJI) is a severe complication, causing significant morbidity and healthcare costs. A recent article put forth up to date guidelines for the management of PJI. The purpose of this current study was to evaluate the external clinical validity of these treatment recommendations.

Study design: Retrospective cohort analysis.

Setting: 1,000-bed tertiary care center in Switzerland.

Synopsis: 68 consecutive episodes of PJI from January 1995 through December 2004 were reviewed. Patients with polymicrobial infections and with treatment failures prior to referral to this center were included.

The success rate for treatment of PJI was highest (67%) when the surgical strategy met current recommendations and antimicrobial treatment was adequate or partially adequate. The preferred surgical strategy was a two-stage exchange. The risk of treatment failure was higher for PJI treated with a surgical strategy other than that recommended (hazard ratio [HR] 2.34, p=0.01) and for PJIs treated with antibiotics not corresponding to recommendations (HR 3.45, p=0.002).

This study was limited by its small sample size and retrospective nature. Patients were not randomized, and cure rates for PJI were significantly lower than in prior published studies.

Bottom line: Treatment of PJI in higher risk populations in accordance with currently recommended surgical and antimicrobial treatment recommendations is associated with better outcomes and cure rates.

Citation: Betsch BY, Eggli S, Siebenrock KA, Tauber MG, Muhlemann K. Treatment of joint prosthesis infection in accordance with current recommendations improves outcome. Clin Infect Dis. 2008;46:1221-1226.

Do Oral Sodium Phosphate Drugs in Large Bowel Prep for Endoscopy Worsen Renal Function?

Background: Proper bowel preparation is essential for adequate colonoscopy and flexible sigmoidoscopy. Oral agents that are most commonly used for bowel preparation are sodium phosphate drugs, polyethylglycol and magnesium citrate. Sodium phosphate drugs are often preferred because of the decreased amount of fluid necessary for bowel preparation.

Study design: Retrospective study.

Setting: Scott and White Clinic, Temple, Texas.

Synopsis: Researchers compared 286 patients receiving an oral sodium phosphate solution for colonoscopy bowel preparation with 125 patients with similar comorbidities who received a non-sodium phosphate solution for bowel preparation. All patients had normal baseline creatinine levels.

The baseline, six-month, and one-year glomerular filtration rates (GFR) were compared between the two groups. GFR declined from 79 to 73 to 71 ml/min/1.73 m2 in the study group vs. 76 to 74 to 74 ml/min/1.73 m2 in the control group for the baseline, six-month, and one-year time periods, respectively.

This is an observational study and thus limited by its non-randomized nature. Sodium phosphate has a black-box warning in stage four and five chronic kidney disease because of its deleterious effect on renal function and the potential for inducing electrolyte abnormalities. Given the findings of this study and the availability of other effective preps alternative regimens should be considered for colonic preparation.

Bottom line: Oral sodium phosphate drugs may cause an acute and chronic decline in renal function as measured by the GFR.

Citation: Khurana A, McLean L, Atkinson S, Foulks C. The effect of oral sodium phosphate drug products on renal function in adults is undergoing bowel endoscopy. Arch Intern Med. 2008; 168(6):593-597.

Does MICR Improve Survival-to-hospital Discharge vs. Traditional CPR and ACLS in Cardiac Arrest Outside the Hospital?

Background: Minimally interrupted cardiac resuscitation (MICR), also known as cardiocerebral resuscitation, is hypothesized to increase survival compared with traditional CPR and advanced cardiac life support (ACLS) in out-of-hospital cardiac arrest. In MICR, 200 “pre-shock” chest compressions (100 compressions/minute) are initially given. The rhythm is then analyzed, with a single shock given if indicated.

Study design: Prospective study.

Setting: Two cities in Arizona.

Synopsis: Using data in the Save Heart in Arizona Registry and Education (SHARE) program, outcomes of 218 individuals with cardiac arrest receiving traditional CPR/ACLS were compared to 668 individuals after MICR training was instituted in the same two metropolitan cities. Survival-to-hospital discharge increased from 1.8% (4/218) before MICR training to 5.4% (36/668) after MICR training. The authors then compared the outcomes of 1,799 individuals with cardiac arrest resuscitated by emergency medical services (EMS) who did not receive training in MICR to 661 individuals who received MICR training over the same period. Survival-to-hospital discharge was 9.1% (60/601) in the patients cared from by EMS that received MICR training versus 3.8% (69/1730) in their non-MICR trained colleagues.

This study is limited by its observational nature and lack of randomization. Surprisingly, more individuals were intubated in the MICR groups. For hospitalists, the results could have a dramatic affect on cardiac arrest survival and lead to future changes to CPR/ACLS protocols.

Bottom line: MICR has a significant impact on survival in out-of-hospital cardiac arrest as compared with traditional CPR and ACLS.

Citation: Bobrow B, Clark L, Ewy G, et al. Minimally interrupted cardiac resuscitation by emergency medical services for out-of-hospital cardiac arrest. JAMA. 2008;299(10):1158-1165.

Does Lower aPTT Increase Future VTE Independent of Other Pro-coagulant Factors?

Background: Certain factors, such as obesity, D-dimer levels, and factor V Leiden gene mutations, increase the risk of future venous thromboembolism (VTE) events. This study sought to determine whether lower baseline levels of activated partial thromboplastin time aPTT also increase this risk.

Study design: Prospective multicenter cohort study.

Setting: Longitudinal Investigation of Thromboembolism Etiology research study (Atherosclerosis Risk in Communities portion) in four U.S. communities.

Synopsis: 13,880 individuals with baseline aPTT measurements were followed for 13 years for future VTE events. Of those, 260 developed a VTE of which 111 were described as idiopathic. Individuals in the lowest two quartiles of aPTT compared with the highest fourth quartile had a 2.4-fold and 1.9-fold increase in the risk of VTE, respectively. A lower aPTT further increased the risk of VTE when associated with obesity, elevated D-dimer level, and particularly factor V Leiden.

This study was limited by the relatively small number of VTE events. It also did not clarify whether aPTT measurements in high-risk groups such as those with positive family history of VTE were useful for predicting risk of future VTE. For hospitalists, patients with a lower initial aPTT may warrant more aggressive inpatient DVT prophylaxis.

Bottom line: aPTT below the median level increases the risk of future VTE events, especially if associated with obesity, elevated D-dimer levels, and/or factor V Leiden.

Citation: Zakai NA, Ohira T, White R, Folsom A, Cushman M. Activated partial thromboplastin time and risk of future venous thromboembolism. Am J of Med. 2008;121:231-238.

What Bleeding Outcomes are Associated with Using Warfarin with Antiplatelet Agents?

Background: Despite a high prevalence of combining antiplatelet and warfarin therapy, the timing, safety, and efficacy of this strategy remain controversial.

Study design: Retrospective cohort study.

Setting: Kaiser Permanente Colorado.

Synopsis: Using a pharmacy database, the authors identified 2,560 patients receiving warfarin alone (monotherapy cohort) and 1,623 patients receiving warfarin combined with antiplatelet agents (combination therapy cohort).

In the combination therapy cohort, aspirin was the most common antiplatelet agent (37%) followed by clopidogrel (13%) and dipyridamole (2%). During a six-month period, the combination therapy cohort had a 4.2% risk of hemorrhage and a 2.0% risk of major hemorrhage. Warfarin monotherapy was associated with a 2% risk of hemorrhage and 0.9% risk of major hemorrhage.

At baseline, the combination therapy patients were twice as likely to have diabetes or congestive heart failure and four times as likely to have coronary artery disease. In both cohorts, the most common reason for warfarin therapy was atrial fibrillation.

Since this was a retrospective investigation, hospitalists should be careful about drawing conclusions from this study alone, but are reminded to discuss risks carefully and engage in shared decision-making with patients when using combined warfarin and antiplatelet therapy.

Bottom line: Warfarin use in combination with antiplatelet therapy is associated with more than double the risk of bleeding compared with warfarin monotherapy.

Citation: Johnson SG, Rogers K, Delate T, Witt DM. Outcomes associated with combined antiplatelet and anticoagulant therapy. Chest. 2008;133:948-954.

Does a Rise in Serum Creatinine Affect Post-hospitalization Mortality and ESRD in Elderly MI Patients?

Background: Previous studies found an association between small changes in serum creatinine during hospitalization and short-term mortality. Data has shown patients experiencing a rise in creatinine at the time of CABG have increased in-hospital and long-term follow-up mortality.

Study design: Retrospective cohort study

Setting: Nationwide Medicare database of acute MI hospitalizations.

Synopsis: The authors reviewed outcomes data for 87,094 patients hospitalized for acute myocardial infarction (MI) from 1994-1995 with follow-up data through 2004. Patients were classified into groups with no rise in creatinine during hospitalization and those with rises of 0.1 mg/dL, 0.2 mg/dL, 0.3-0.5 mg/dL, and 0.6-3 mg/dL.

Compared with patients with no rise in creatinine, a rise of 0.1 mg/dL was associated with an adjusted hazard ratio of 1.45 for end-stage renal disease (ESRD) and 1.14 for post-hospitalization death during long-term follow-up. An incremental increase in poor outcomes was seen with more dramatic increases in creatinine, with patients in the group with a 0.6-3 mg/dL rise in creatinine having an adjusted hazard ratio of 3.26 for ESRD and 1.39 for post-hospitalization death. Among patients with a creatinine rise, the absolute risk of mortality (15% annually) was greater than that of ESRD (0.3% annually).

Hospitalists should note limitations of this retrospective study, including its restriction to hospitalized elderly patients.

Bottom line: Even small rises in serum creatinine during acute hospitalization for MI are associated with long-term risk for death and ESRD in elderly patients.

Citation: Newsome BB, Warnock DG, McClellan WM, et al. Long-term risk of mortality and end-stage renal disease among the elderly after small increases in serum creatinine level during hospitalization for acute myocardial infarction. Arch Intern Med. 2008;168(6):609-616.

Does Direct-to-patient Communication Improve Adherence to Beta-blocker Therapy Following an MI?

Background: The joint American Heart Association and American College of Cardiology guidelines have specific treatment recommendations regarding care of a patient post-myocardial infarction (MI). A key component of this regimen is beta-blocker therapy. Beta-blockers routinely are prescribed at hospital discharge following MI; however, patient adherence has been shown to decline substantially over time.

Study design: Cluster randomized control trial.

Setting: Four health maintenance organizations in Boston, Minneapolis, Atlanta, and Portland, Ore.

Synopsis: 836 post-MI patients were given a beta-blocker prescription upon discharge from the hospital. The intervention group received two mailed communications. The first was a personalized, simply worded letter from a health plan physician-administrator, followed two months later by a similar letter with a brochure. Mailers were low cost and easily replicable; they addressed the importance of these medications, the risks of non-adherence, and adverse effects.

The primary outcome measure was beta-blocker adherence. Medication adherence was analyzed as a continuous measure and as a monthly proportion of days covered (PDC) of 80% or greater. Across all months of follow-up, a mean of 64.8% of intervention patients had a PDC of more than 80% compared with 58.5% of control group patients (number needed to treat=16). The intervention group was 17% more likely to have a PDC of 80% or greater over the entire post-intervention period.

These interventions were studied in a prepaid integrated care delivery system—limiting generalization to other insurance types. Nevertheless, finding ways to improve patient compliance and decrease recurrent cardiac events is liking to result in cost saving to any healthcare plan.

Bottom line: A low-cost direct-to-patient communication effort can have a positive effect on beta-blocker adherence following MI.

Citation: Smith DH, Kramer JM, Perrin N, et al. A randomized trial of direct-to-patient communication to enhance adherence to beta-blocker therapy following myocardial infarction. Arch Intern Med. 2008;168(5):477-483. TH

Literature at a Glance

A guide to this month’s studies.

• Survival of in-hospital cardiac arrest decreases during nights and weekends.
• C-reactive protein levels predict severity and complications in community-acquired pneumonia.
• Adherence to current guidelines improves outcome in treatment of prosthetic joint infection.
• Sodium phosphate bowel prep use is associated with a decline in GFR.
• Minimally interrupted cardiac resuscitation improves survival.
• Lower aPTT increases risk for future VTE.
• Warfarin combined with antiplatelet therapy increases hemorrhage rate.
• Creatinine rise during MI hospitalization is associated with long-term risk of death and ESRD.
• Communication improves patient adherence to beta-blockers after MI.

Does Time of Day Affect Survival of Cardiac Arrest Patients?

Background: In-hospital cardiac arrest is a major public health problem. Small studies have demonstrated survival after cardiac arrest is worse at night as compared with all other times. Multiple hypothesis are proffered for this decreased survival, including less effective detection and treatment of the warning signs of impending arrest during the night hours.

Study design: Prospective registry.

Setting: 507 hospitals participating through the National Registry of Cardiopulmonary Resuscitation.

Synopsis: 86,748 consecutive, inpatient cardiac arrests were reported from Jan. 1, 2000, through Feb. 1, 2007, including 58,593 cases during day/evening hours and 28,155 cases during night hours.

Rates of survival to discharge (14.7 % vs. 19.8%), survival at 24 hours (28.9% vs. 35.4%), and favorable neurological outcomes (11.0% vs. 15.2%) were substantially lower during the night compared with day/evening (all p values < 0.001). The first documented rhythm at night was more frequently asystole as opposed to ventricular fibrillation during the day/evening. There also was a higher survival rate with cardiac arrests during day/evening hours occurring on weekdays compared with weekends (odds ratio [OR] 1.15). There was no difference in survival rates between weekdays or weekends among cardiac arrests occurring during the night hours.

Bottom line: Survival rates for in-hospital cardiac arrest are lower during nights and weekends, which may relate to differential physician and hospital staffing patterns during these hours.

Citation: Peberdy MA, Ornato JP, Larkin GL et al. Survival from in-hospital cardiac arrest during nights and weekends. JAMA 2008;299:785-792.

Can CRP Identify Risk in CAP Patients?

Background: Small initial studies suggest an elevated C-reactive protein (CRP) is relatively nonspecific but may have a role in predicting disease severity in community-acquired pneumonia (CAP).

Study design: Prospective study.

Setting: Large academic center in the United Kingdom.

Synopsis: In this study of 570 patients over a two-year time period, all patients presenting to the hospital with a diagnosis of CAP and the absence of exclusion criteria were evaluated. CRP was measured on admission and repeated on day four of hospitalization.

Low CRP levels (less than 100 mg/L) were independently associated with a reduced risk of 30-day mortality (OR 0.18; p=0.03), need for invasive ventilation and/or inotropic support (OR 0.21; p=0.002), and complicated pneumonia (OR 0.05; p=0.003). In addition, the failure of CRP to fall by 50% or more at day four of hospitalization was associated with an increased risk of 30-day mortality (OR 24.5; p<0.0001), need for mechanical ventilation and/or inotropic support (OR 7.1; p<0.0001), and complicated pneumonia (OR 15.4; p<0.0001).

Patients with chronic lung disease, immunosuppression, active malignancy or hospital-acquired pneumonia were excluded from the study and the conclusions cannot be extrapolated to these higher risk populations.

Bottom line: C-reactive protein is an independent marker of severity in CAP, and low levels can be used as an adjunct to clinical judgment to help identify patients who may be safely discharged from the hospital.

Citation: Chalmers JD, Singanayagam A, Hill AT. C-reactive protein is an independent predictor of severity in community-acquired pneumonia. Am J of Med. 2008;121:219-225.

Do Aggressive Surgical Intervention and Antimicrobial Treatment Improve Outcomes in Patients Suffering PJI?

Background: Prosthetic joint infection (PJI) is a severe complication, causing significant morbidity and healthcare costs. A recent article put forth up to date guidelines for the management of PJI. The purpose of this current study was to evaluate the external clinical validity of these treatment recommendations.

Study design: Retrospective cohort analysis.

Setting: 1,000-bed tertiary care center in Switzerland.

Synopsis: 68 consecutive episodes of PJI from January 1995 through December 2004 were reviewed. Patients with polymicrobial infections and with treatment failures prior to referral to this center were included.

The success rate for treatment of PJI was highest (67%) when the surgical strategy met current recommendations and antimicrobial treatment was adequate or partially adequate. The preferred surgical strategy was a two-stage exchange. The risk of treatment failure was higher for PJI treated with a surgical strategy other than that recommended (hazard ratio [HR] 2.34, p=0.01) and for PJIs treated with antibiotics not corresponding to recommendations (HR 3.45, p=0.002).

This study was limited by its small sample size and retrospective nature. Patients were not randomized, and cure rates for PJI were significantly lower than in prior published studies.

Bottom line: Treatment of PJI in higher risk populations in accordance with currently recommended surgical and antimicrobial treatment recommendations is associated with better outcomes and cure rates.

Citation: Betsch BY, Eggli S, Siebenrock KA, Tauber MG, Muhlemann K. Treatment of joint prosthesis infection in accordance with current recommendations improves outcome. Clin Infect Dis. 2008;46:1221-1226.

Do Oral Sodium Phosphate Drugs in Large Bowel Prep for Endoscopy Worsen Renal Function?

Background: Proper bowel preparation is essential for adequate colonoscopy and flexible sigmoidoscopy. Oral agents that are most commonly used for bowel preparation are sodium phosphate drugs, polyethylglycol and magnesium citrate. Sodium phosphate drugs are often preferred because of the decreased amount of fluid necessary for bowel preparation.

Study design: Retrospective study.

Setting: Scott and White Clinic, Temple, Texas.

Synopsis: Researchers compared 286 patients receiving an oral sodium phosphate solution for colonoscopy bowel preparation with 125 patients with similar comorbidities who received a non-sodium phosphate solution for bowel preparation. All patients had normal baseline creatinine levels.

The baseline, six-month, and one-year glomerular filtration rates (GFR) were compared between the two groups. GFR declined from 79 to 73 to 71 ml/min/1.73 m2 in the study group vs. 76 to 74 to 74 ml/min/1.73 m2 in the control group for the baseline, six-month, and one-year time periods, respectively.

This is an observational study and thus limited by its non-randomized nature. Sodium phosphate has a black-box warning in stage four and five chronic kidney disease because of its deleterious effect on renal function and the potential for inducing electrolyte abnormalities. Given the findings of this study and the availability of other effective preps alternative regimens should be considered for colonic preparation.

Bottom line: Oral sodium phosphate drugs may cause an acute and chronic decline in renal function as measured by the GFR.

Citation: Khurana A, McLean L, Atkinson S, Foulks C. The effect of oral sodium phosphate drug products on renal function in adults is undergoing bowel endoscopy. Arch Intern Med. 2008; 168(6):593-597.

Does MICR Improve Survival-to-hospital Discharge vs. Traditional CPR and ACLS in Cardiac Arrest Outside the Hospital?

Background: Minimally interrupted cardiac resuscitation (MICR), also known as cardiocerebral resuscitation, is hypothesized to increase survival compared with traditional CPR and advanced cardiac life support (ACLS) in out-of-hospital cardiac arrest. In MICR, 200 “pre-shock” chest compressions (100 compressions/minute) are initially given. The rhythm is then analyzed, with a single shock given if indicated.

Study design: Prospective study.

Setting: Two cities in Arizona.

Synopsis: Using data in the Save Heart in Arizona Registry and Education (SHARE) program, outcomes of 218 individuals with cardiac arrest receiving traditional CPR/ACLS were compared to 668 individuals after MICR training was instituted in the same two metropolitan cities. Survival-to-hospital discharge increased from 1.8% (4/218) before MICR training to 5.4% (36/668) after MICR training. The authors then compared the outcomes of 1,799 individuals with cardiac arrest resuscitated by emergency medical services (EMS) who did not receive training in MICR to 661 individuals who received MICR training over the same period. Survival-to-hospital discharge was 9.1% (60/601) in the patients cared from by EMS that received MICR training versus 3.8% (69/1730) in their non-MICR trained colleagues.

This study is limited by its observational nature and lack of randomization. Surprisingly, more individuals were intubated in the MICR groups. For hospitalists, the results could have a dramatic affect on cardiac arrest survival and lead to future changes to CPR/ACLS protocols.

Bottom line: MICR has a significant impact on survival in out-of-hospital cardiac arrest as compared with traditional CPR and ACLS.

Citation: Bobrow B, Clark L, Ewy G, et al. Minimally interrupted cardiac resuscitation by emergency medical services for out-of-hospital cardiac arrest. JAMA. 2008;299(10):1158-1165.

Does Lower aPTT Increase Future VTE Independent of Other Pro-coagulant Factors?

Background: Certain factors, such as obesity, D-dimer levels, and factor V Leiden gene mutations, increase the risk of future venous thromboembolism (VTE) events. This study sought to determine whether lower baseline levels of activated partial thromboplastin time aPTT also increase this risk.

Study design: Prospective multicenter cohort study.

Setting: Longitudinal Investigation of Thromboembolism Etiology research study (Atherosclerosis Risk in Communities portion) in four U.S. communities.

Synopsis: 13,880 individuals with baseline aPTT measurements were followed for 13 years for future VTE events. Of those, 260 developed a VTE of which 111 were described as idiopathic. Individuals in the lowest two quartiles of aPTT compared with the highest fourth quartile had a 2.4-fold and 1.9-fold increase in the risk of VTE, respectively. A lower aPTT further increased the risk of VTE when associated with obesity, elevated D-dimer level, and particularly factor V Leiden.

This study was limited by the relatively small number of VTE events. It also did not clarify whether aPTT measurements in high-risk groups such as those with positive family history of VTE were useful for predicting risk of future VTE. For hospitalists, patients with a lower initial aPTT may warrant more aggressive inpatient DVT prophylaxis.

Bottom line: aPTT below the median level increases the risk of future VTE events, especially if associated with obesity, elevated D-dimer levels, and/or factor V Leiden.

Citation: Zakai NA, Ohira T, White R, Folsom A, Cushman M. Activated partial thromboplastin time and risk of future venous thromboembolism. Am J of Med. 2008;121:231-238.

What Bleeding Outcomes are Associated with Using Warfarin with Antiplatelet Agents?

Background: Despite a high prevalence of combining antiplatelet and warfarin therapy, the timing, safety, and efficacy of this strategy remain controversial.

Study design: Retrospective cohort study.

Setting: Kaiser Permanente Colorado.

Synopsis: Using a pharmacy database, the authors identified 2,560 patients receiving warfarin alone (monotherapy cohort) and 1,623 patients receiving warfarin combined with antiplatelet agents (combination therapy cohort).

In the combination therapy cohort, aspirin was the most common antiplatelet agent (37%) followed by clopidogrel (13%) and dipyridamole (2%). During a six-month period, the combination therapy cohort had a 4.2% risk of hemorrhage and a 2.0% risk of major hemorrhage. Warfarin monotherapy was associated with a 2% risk of hemorrhage and 0.9% risk of major hemorrhage.

At baseline, the combination therapy patients were twice as likely to have diabetes or congestive heart failure and four times as likely to have coronary artery disease. In both cohorts, the most common reason for warfarin therapy was atrial fibrillation.

Since this was a retrospective investigation, hospitalists should be careful about drawing conclusions from this study alone, but are reminded to discuss risks carefully and engage in shared decision-making with patients when using combined warfarin and antiplatelet therapy.

Bottom line: Warfarin use in combination with antiplatelet therapy is associated with more than double the risk of bleeding compared with warfarin monotherapy.

Citation: Johnson SG, Rogers K, Delate T, Witt DM. Outcomes associated with combined antiplatelet and anticoagulant therapy. Chest. 2008;133:948-954.

Does a Rise in Serum Creatinine Affect Post-hospitalization Mortality and ESRD in Elderly MI Patients?

Background: Previous studies found an association between small changes in serum creatinine during hospitalization and short-term mortality. Data has shown patients experiencing a rise in creatinine at the time of CABG have increased in-hospital and long-term follow-up mortality.

Study design: Retrospective cohort study

Setting: Nationwide Medicare database of acute MI hospitalizations.

Synopsis: The authors reviewed outcomes data for 87,094 patients hospitalized for acute myocardial infarction (MI) from 1994-1995 with follow-up data through 2004. Patients were classified into groups with no rise in creatinine during hospitalization and those with rises of 0.1 mg/dL, 0.2 mg/dL, 0.3-0.5 mg/dL, and 0.6-3 mg/dL.

Compared with patients with no rise in creatinine, a rise of 0.1 mg/dL was associated with an adjusted hazard ratio of 1.45 for end-stage renal disease (ESRD) and 1.14 for post-hospitalization death during long-term follow-up. An incremental increase in poor outcomes was seen with more dramatic increases in creatinine, with patients in the group with a 0.6-3 mg/dL rise in creatinine having an adjusted hazard ratio of 3.26 for ESRD and 1.39 for post-hospitalization death. Among patients with a creatinine rise, the absolute risk of mortality (15% annually) was greater than that of ESRD (0.3% annually).

Hospitalists should note limitations of this retrospective study, including its restriction to hospitalized elderly patients.

Bottom line: Even small rises in serum creatinine during acute hospitalization for MI are associated with long-term risk for death and ESRD in elderly patients.

Citation: Newsome BB, Warnock DG, McClellan WM, et al. Long-term risk of mortality and end-stage renal disease among the elderly after small increases in serum creatinine level during hospitalization for acute myocardial infarction. Arch Intern Med. 2008;168(6):609-616.

Does Direct-to-patient Communication Improve Adherence to Beta-blocker Therapy Following an MI?

Background: The joint American Heart Association and American College of Cardiology guidelines have specific treatment recommendations regarding care of a patient post-myocardial infarction (MI). A key component of this regimen is beta-blocker therapy. Beta-blockers routinely are prescribed at hospital discharge following MI; however, patient adherence has been shown to decline substantially over time.

Study design: Cluster randomized control trial.

Setting: Four health maintenance organizations in Boston, Minneapolis, Atlanta, and Portland, Ore.

Synopsis: 836 post-MI patients were given a beta-blocker prescription upon discharge from the hospital. The intervention group received two mailed communications. The first was a personalized, simply worded letter from a health plan physician-administrator, followed two months later by a similar letter with a brochure. Mailers were low cost and easily replicable; they addressed the importance of these medications, the risks of non-adherence, and adverse effects.

The primary outcome measure was beta-blocker adherence. Medication adherence was analyzed as a continuous measure and as a monthly proportion of days covered (PDC) of 80% or greater. Across all months of follow-up, a mean of 64.8% of intervention patients had a PDC of more than 80% compared with 58.5% of control group patients (number needed to treat=16). The intervention group was 17% more likely to have a PDC of 80% or greater over the entire post-intervention period.

These interventions were studied in a prepaid integrated care delivery system—limiting generalization to other insurance types. Nevertheless, finding ways to improve patient compliance and decrease recurrent cardiac events is liking to result in cost saving to any healthcare plan.

Bottom line: A low-cost direct-to-patient communication effort can have a positive effect on beta-blocker adherence following MI.

Citation: Smith DH, Kramer JM, Perrin N, et al. A randomized trial of direct-to-patient communication to enhance adherence to beta-blocker therapy following myocardial infarction. Arch Intern Med. 2008;168(5):477-483. TH

Literature at a Glance

A guide to this month’s studies.

• Survival of in-hospital cardiac arrest decreases during nights and weekends.
• C-reactive protein levels predict severity and complications in community-acquired pneumonia.
• Adherence to current guidelines improves outcome in treatment of prosthetic joint infection.
• Sodium phosphate bowel prep use is associated with a decline in GFR.
• Minimally interrupted cardiac resuscitation improves survival.
• Lower aPTT increases risk for future VTE.
• Warfarin combined with antiplatelet therapy increases hemorrhage rate.
• Creatinine rise during MI hospitalization is associated with long-term risk of death and ESRD.
• Communication improves patient adherence to beta-blockers after MI.

Does Time of Day Affect Survival of Cardiac Arrest Patients?

Background: In-hospital cardiac arrest is a major public health problem. Small studies have demonstrated survival after cardiac arrest is worse at night as compared with all other times. Multiple hypothesis are proffered for this decreased survival, including less effective detection and treatment of the warning signs of impending arrest during the night hours.

Study design: Prospective registry.

Setting: 507 hospitals participating through the National Registry of Cardiopulmonary Resuscitation.

Synopsis: 86,748 consecutive, inpatient cardiac arrests were reported from Jan. 1, 2000, through Feb. 1, 2007, including 58,593 cases during day/evening hours and 28,155 cases during night hours.

Rates of survival to discharge (14.7 % vs. 19.8%), survival at 24 hours (28.9% vs. 35.4%), and favorable neurological outcomes (11.0% vs. 15.2%) were substantially lower during the night compared with day/evening (all p values < 0.001). The first documented rhythm at night was more frequently asystole as opposed to ventricular fibrillation during the day/evening. There also was a higher survival rate with cardiac arrests during day/evening hours occurring on weekdays compared with weekends (odds ratio [OR] 1.15). There was no difference in survival rates between weekdays or weekends among cardiac arrests occurring during the night hours.

Bottom line: Survival rates for in-hospital cardiac arrest are lower during nights and weekends, which may relate to differential physician and hospital staffing patterns during these hours.

Citation: Peberdy MA, Ornato JP, Larkin GL et al. Survival from in-hospital cardiac arrest during nights and weekends. JAMA 2008;299:785-792.

Can CRP Identify Risk in CAP Patients?

Background: Small initial studies suggest an elevated C-reactive protein (CRP) is relatively nonspecific but may have a role in predicting disease severity in community-acquired pneumonia (CAP).

Study design: Prospective study.

Setting: Large academic center in the United Kingdom.

Synopsis: In this study of 570 patients over a two-year time period, all patients presenting to the hospital with a diagnosis of CAP and the absence of exclusion criteria were evaluated. CRP was measured on admission and repeated on day four of hospitalization.

Low CRP levels (less than 100 mg/L) were independently associated with a reduced risk of 30-day mortality (OR 0.18; p=0.03), need for invasive ventilation and/or inotropic support (OR 0.21; p=0.002), and complicated pneumonia (OR 0.05; p=0.003). In addition, the failure of CRP to fall by 50% or more at day four of hospitalization was associated with an increased risk of 30-day mortality (OR 24.5; p<0.0001), need for mechanical ventilation and/or inotropic support (OR 7.1; p<0.0001), and complicated pneumonia (OR 15.4; p<0.0001).

Patients with chronic lung disease, immunosuppression, active malignancy or hospital-acquired pneumonia were excluded from the study and the conclusions cannot be extrapolated to these higher risk populations.

Bottom line: C-reactive protein is an independent marker of severity in CAP, and low levels can be used as an adjunct to clinical judgment to help identify patients who may be safely discharged from the hospital.

Citation: Chalmers JD, Singanayagam A, Hill AT. C-reactive protein is an independent predictor of severity in community-acquired pneumonia. Am J of Med. 2008;121:219-225.

Do Aggressive Surgical Intervention and Antimicrobial Treatment Improve Outcomes in Patients Suffering PJI?

Background: Prosthetic joint infection (PJI) is a severe complication, causing significant morbidity and healthcare costs. A recent article put forth up to date guidelines for the management of PJI. The purpose of this current study was to evaluate the external clinical validity of these treatment recommendations.

Study design: Retrospective cohort analysis.

Setting: 1,000-bed tertiary care center in Switzerland.

Synopsis: 68 consecutive episodes of PJI from January 1995 through December 2004 were reviewed. Patients with polymicrobial infections and with treatment failures prior to referral to this center were included.

The success rate for treatment of PJI was highest (67%) when the surgical strategy met current recommendations and antimicrobial treatment was adequate or partially adequate. The preferred surgical strategy was a two-stage exchange. The risk of treatment failure was higher for PJI treated with a surgical strategy other than that recommended (hazard ratio [HR] 2.34, p=0.01) and for PJIs treated with antibiotics not corresponding to recommendations (HR 3.45, p=0.002).

This study was limited by its small sample size and retrospective nature. Patients were not randomized, and cure rates for PJI were significantly lower than in prior published studies.

Bottom line: Treatment of PJI in higher risk populations in accordance with currently recommended surgical and antimicrobial treatment recommendations is associated with better outcomes and cure rates.

Citation: Betsch BY, Eggli S, Siebenrock KA, Tauber MG, Muhlemann K. Treatment of joint prosthesis infection in accordance with current recommendations improves outcome. Clin Infect Dis. 2008;46:1221-1226.

Do Oral Sodium Phosphate Drugs in Large Bowel Prep for Endoscopy Worsen Renal Function?

Background: Proper bowel preparation is essential for adequate colonoscopy and flexible sigmoidoscopy. Oral agents that are most commonly used for bowel preparation are sodium phosphate drugs, polyethylglycol and magnesium citrate. Sodium phosphate drugs are often preferred because of the decreased amount of fluid necessary for bowel preparation.

Study design: Retrospective study.

Setting: Scott and White Clinic, Temple, Texas.

Synopsis: Researchers compared 286 patients receiving an oral sodium phosphate solution for colonoscopy bowel preparation with 125 patients with similar comorbidities who received a non-sodium phosphate solution for bowel preparation. All patients had normal baseline creatinine levels.

The baseline, six-month, and one-year glomerular filtration rates (GFR) were compared between the two groups. GFR declined from 79 to 73 to 71 ml/min/1.73 m2 in the study group vs. 76 to 74 to 74 ml/min/1.73 m2 in the control group for the baseline, six-month, and one-year time periods, respectively.

This is an observational study and thus limited by its non-randomized nature. Sodium phosphate has a black-box warning in stage four and five chronic kidney disease because of its deleterious effect on renal function and the potential for inducing electrolyte abnormalities. Given the findings of this study and the availability of other effective preps alternative regimens should be considered for colonic preparation.

Bottom line: Oral sodium phosphate drugs may cause an acute and chronic decline in renal function as measured by the GFR.

Citation: Khurana A, McLean L, Atkinson S, Foulks C. The effect of oral sodium phosphate drug products on renal function in adults is undergoing bowel endoscopy. Arch Intern Med. 2008; 168(6):593-597.

Does MICR Improve Survival-to-hospital Discharge vs. Traditional CPR and ACLS in Cardiac Arrest Outside the Hospital?

Background: Minimally interrupted cardiac resuscitation (MICR), also known as cardiocerebral resuscitation, is hypothesized to increase survival compared with traditional CPR and advanced cardiac life support (ACLS) in out-of-hospital cardiac arrest. In MICR, 200 “pre-shock” chest compressions (100 compressions/minute) are initially given. The rhythm is then analyzed, with a single shock given if indicated.

Study design: Prospective study.

Setting: Two cities in Arizona.

Synopsis: Using data in the Save Heart in Arizona Registry and Education (SHARE) program, outcomes of 218 individuals with cardiac arrest receiving traditional CPR/ACLS were compared to 668 individuals after MICR training was instituted in the same two metropolitan cities. Survival-to-hospital discharge increased from 1.8% (4/218) before MICR training to 5.4% (36/668) after MICR training. The authors then compared the outcomes of 1,799 individuals with cardiac arrest resuscitated by emergency medical services (EMS) who did not receive training in MICR to 661 individuals who received MICR training over the same period. Survival-to-hospital discharge was 9.1% (60/601) in the patients cared from by EMS that received MICR training versus 3.8% (69/1730) in their non-MICR trained colleagues.

This study is limited by its observational nature and lack of randomization. Surprisingly, more individuals were intubated in the MICR groups. For hospitalists, the results could have a dramatic affect on cardiac arrest survival and lead to future changes to CPR/ACLS protocols.

Bottom line: MICR has a significant impact on survival in out-of-hospital cardiac arrest as compared with traditional CPR and ACLS.

Citation: Bobrow B, Clark L, Ewy G, et al. Minimally interrupted cardiac resuscitation by emergency medical services for out-of-hospital cardiac arrest. JAMA. 2008;299(10):1158-1165.

Does Lower aPTT Increase Future VTE Independent of Other Pro-coagulant Factors?

Background: Certain factors, such as obesity, D-dimer levels, and factor V Leiden gene mutations, increase the risk of future venous thromboembolism (VTE) events. This study sought to determine whether lower baseline levels of activated partial thromboplastin time aPTT also increase this risk.

Study design: Prospective multicenter cohort study.

Setting: Longitudinal Investigation of Thromboembolism Etiology research study (Atherosclerosis Risk in Communities portion) in four U.S. communities.

Synopsis: 13,880 individuals with baseline aPTT measurements were followed for 13 years for future VTE events. Of those, 260 developed a VTE of which 111 were described as idiopathic. Individuals in the lowest two quartiles of aPTT compared with the highest fourth quartile had a 2.4-fold and 1.9-fold increase in the risk of VTE, respectively. A lower aPTT further increased the risk of VTE when associated with obesity, elevated D-dimer level, and particularly factor V Leiden.

This study was limited by the relatively small number of VTE events. It also did not clarify whether aPTT measurements in high-risk groups such as those with positive family history of VTE were useful for predicting risk of future VTE. For hospitalists, patients with a lower initial aPTT may warrant more aggressive inpatient DVT prophylaxis.

Bottom line: aPTT below the median level increases the risk of future VTE events, especially if associated with obesity, elevated D-dimer levels, and/or factor V Leiden.

Citation: Zakai NA, Ohira T, White R, Folsom A, Cushman M. Activated partial thromboplastin time and risk of future venous thromboembolism. Am J of Med. 2008;121:231-238.

What Bleeding Outcomes are Associated with Using Warfarin with Antiplatelet Agents?

Background: Despite a high prevalence of combining antiplatelet and warfarin therapy, the timing, safety, and efficacy of this strategy remain controversial.

Study design: Retrospective cohort study.

Setting: Kaiser Permanente Colorado.

Synopsis: Using a pharmacy database, the authors identified 2,560 patients receiving warfarin alone (monotherapy cohort) and 1,623 patients receiving warfarin combined with antiplatelet agents (combination therapy cohort).

In the combination therapy cohort, aspirin was the most common antiplatelet agent (37%) followed by clopidogrel (13%) and dipyridamole (2%). During a six-month period, the combination therapy cohort had a 4.2% risk of hemorrhage and a 2.0% risk of major hemorrhage. Warfarin monotherapy was associated with a 2% risk of hemorrhage and 0.9% risk of major hemorrhage.

At baseline, the combination therapy patients were twice as likely to have diabetes or congestive heart failure and four times as likely to have coronary artery disease. In both cohorts, the most common reason for warfarin therapy was atrial fibrillation.

Since this was a retrospective investigation, hospitalists should be careful about drawing conclusions from this study alone, but are reminded to discuss risks carefully and engage in shared decision-making with patients when using combined warfarin and antiplatelet therapy.

Bottom line: Warfarin use in combination with antiplatelet therapy is associated with more than double the risk of bleeding compared with warfarin monotherapy.

Citation: Johnson SG, Rogers K, Delate T, Witt DM. Outcomes associated with combined antiplatelet and anticoagulant therapy. Chest. 2008;133:948-954.

Does a Rise in Serum Creatinine Affect Post-hospitalization Mortality and ESRD in Elderly MI Patients?

Background: Previous studies found an association between small changes in serum creatinine during hospitalization and short-term mortality. Data has shown patients experiencing a rise in creatinine at the time of CABG have increased in-hospital and long-term follow-up mortality.

Study design: Retrospective cohort study

Setting: Nationwide Medicare database of acute MI hospitalizations.

Synopsis: The authors reviewed outcomes data for 87,094 patients hospitalized for acute myocardial infarction (MI) from 1994-1995 with follow-up data through 2004. Patients were classified into groups with no rise in creatinine during hospitalization and those with rises of 0.1 mg/dL, 0.2 mg/dL, 0.3-0.5 mg/dL, and 0.6-3 mg/dL.

Compared with patients with no rise in creatinine, a rise of 0.1 mg/dL was associated with an adjusted hazard ratio of 1.45 for end-stage renal disease (ESRD) and 1.14 for post-hospitalization death during long-term follow-up. An incremental increase in poor outcomes was seen with more dramatic increases in creatinine, with patients in the group with a 0.6-3 mg/dL rise in creatinine having an adjusted hazard ratio of 3.26 for ESRD and 1.39 for post-hospitalization death. Among patients with a creatinine rise, the absolute risk of mortality (15% annually) was greater than that of ESRD (0.3% annually).

Hospitalists should note limitations of this retrospective study, including its restriction to hospitalized elderly patients.

Bottom line: Even small rises in serum creatinine during acute hospitalization for MI are associated with long-term risk for death and ESRD in elderly patients.

Citation: Newsome BB, Warnock DG, McClellan WM, et al. Long-term risk of mortality and end-stage renal disease among the elderly after small increases in serum creatinine level during hospitalization for acute myocardial infarction. Arch Intern Med. 2008;168(6):609-616.

Does Direct-to-patient Communication Improve Adherence to Beta-blocker Therapy Following an MI?

Background: The joint American Heart Association and American College of Cardiology guidelines have specific treatment recommendations regarding care of a patient post-myocardial infarction (MI). A key component of this regimen is beta-blocker therapy. Beta-blockers routinely are prescribed at hospital discharge following MI; however, patient adherence has been shown to decline substantially over time.

Study design: Cluster randomized control trial.

Setting: Four health maintenance organizations in Boston, Minneapolis, Atlanta, and Portland, Ore.

Synopsis: 836 post-MI patients were given a beta-blocker prescription upon discharge from the hospital. The intervention group received two mailed communications. The first was a personalized, simply worded letter from a health plan physician-administrator, followed two months later by a similar letter with a brochure. Mailers were low cost and easily replicable; they addressed the importance of these medications, the risks of non-adherence, and adverse effects.

The primary outcome measure was beta-blocker adherence. Medication adherence was analyzed as a continuous measure and as a monthly proportion of days covered (PDC) of 80% or greater. Across all months of follow-up, a mean of 64.8% of intervention patients had a PDC of more than 80% compared with 58.5% of control group patients (number needed to treat=16). The intervention group was 17% more likely to have a PDC of 80% or greater over the entire post-intervention period.

These interventions were studied in a prepaid integrated care delivery system—limiting generalization to other insurance types. Nevertheless, finding ways to improve patient compliance and decrease recurrent cardiac events is liking to result in cost saving to any healthcare plan.

Bottom line: A low-cost direct-to-patient communication effort can have a positive effect on beta-blocker adherence following MI.

Citation: Smith DH, Kramer JM, Perrin N, et al. A randomized trial of direct-to-patient communication to enhance adherence to beta-blocker therapy following myocardial infarction. Arch Intern Med. 2008;168(5):477-483. TH

Dear John Q. Hospitalist,

Recently, a pair of college students in our office presented an impressive summary of Web 2.0, including Facebook.com and LinkedIn.com, to the rest of the SHM staff.

As I listened to their presentation and heard the energy in their voices, I couldn’t help but think about my initial experience and excitement with the World Wide Web. Instead of doing homework, I spent many late nights searching the Internet looking for more information to help me create my first Web page. After countless hours of coding and debugging, as well as throwing the keyboard a time or two, I published my Web page and became a part of the Internet. I was hooked.

After listening to these students I was inspired to check out LinkedIn.com and create my own LinkedIn profile. While I did not stay up until the very early morning sending invites or completing every part of my profile, I found connections to old colleagues, college friends, high school buddies, and family members. Today, I eagerly await the flood of e-mail from people accepting me as a friend in their network, some of them members of SHM. I am hooked again.

Seeing SHM members on LinkedIn got me thinking about how SHM might use social networking technology. I think there is an opportunity here to create an interactive resource that will empower hospitalists to find other hospitalists, make connections, and build their own networks. I’m interested in getting your perspective. Do you think our members will use this type of an online resource?

Many social networking sites on the Internet grew out of individuals in an academic setting trying to find ways to connect with each other. I would imagine many of our student and resident members already are using social networking sites. Do you think this is the case? If so, what features and functions of a social networking tool do you think are most important? Is that different from a third-year resident, or a hospitalist who has been practicing hospital medicine for a number of years?

Johnson

I know I have thrown a bunch of questions at you, so let me share with you some ideas and maybe we can begin a dialogue that will help SHM find ways in which we can leverage social networking and other Web 2.0 tools.

One of the tasks in creating a LinkedIn account is selecting the college or institution you attended and the years in which you attended. Immediately after setting up my account I was able to see the names of other alumni who attended my university during my four years and invite old friends to join my network. I can envision a scenario where an SHM member indicates which medical school he or she attended and is able to see a list of other colleagues who attended at the same time.

For the general member, someone who hasn’t attended a meeting, participated in a committee, or been more actively engaged in SHM, an online network might be a first step to increased involvement with SHM. Members could use this site to connect with other hospitalists in their area and share their interests and experience with others.

Along the way, they might learn about an SHM initiative they are interested in and connect with another hospitalist who working on this project and begin to have a dialogue. Throughout time, this person builds their network and establishes new connections. When it’s time to register for next year’s SHM Annual Meeting in Chicago, they already know a few faces in the crowd—and maybe a couple of them have become friends.

These are just a couple of ways I think SHM and our members might benefit from social networking. I am confident there are many, many more ways this technology can help our members and the hospital medicine community. What do you think? I would love to hear your thoughts and ideas. E-mail me at [email protected]. TH

Dear John Q. Hospitalist,

Recently, a pair of college students in our office presented an impressive summary of Web 2.0, including Facebook.com and LinkedIn.com, to the rest of the SHM staff.

As I listened to their presentation and heard the energy in their voices, I couldn’t help but think about my initial experience and excitement with the World Wide Web. Instead of doing homework, I spent many late nights searching the Internet looking for more information to help me create my first Web page. After countless hours of coding and debugging, as well as throwing the keyboard a time or two, I published my Web page and became a part of the Internet. I was hooked.

After listening to these students I was inspired to check out LinkedIn.com and create my own LinkedIn profile. While I did not stay up until the very early morning sending invites or completing every part of my profile, I found connections to old colleagues, college friends, high school buddies, and family members. Today, I eagerly await the flood of e-mail from people accepting me as a friend in their network, some of them members of SHM. I am hooked again.

Seeing SHM members on LinkedIn got me thinking about how SHM might use social networking technology. I think there is an opportunity here to create an interactive resource that will empower hospitalists to find other hospitalists, make connections, and build their own networks. I’m interested in getting your perspective. Do you think our members will use this type of an online resource?

Many social networking sites on the Internet grew out of individuals in an academic setting trying to find ways to connect with each other. I would imagine many of our student and resident members already are using social networking sites. Do you think this is the case? If so, what features and functions of a social networking tool do you think are most important? Is that different from a third-year resident, or a hospitalist who has been practicing hospital medicine for a number of years?

Johnson

I know I have thrown a bunch of questions at you, so let me share with you some ideas and maybe we can begin a dialogue that will help SHM find ways in which we can leverage social networking and other Web 2.0 tools.

One of the tasks in creating a LinkedIn account is selecting the college or institution you attended and the years in which you attended. Immediately after setting up my account I was able to see the names of other alumni who attended my university during my four years and invite old friends to join my network. I can envision a scenario where an SHM member indicates which medical school he or she attended and is able to see a list of other colleagues who attended at the same time.

For the general member, someone who hasn’t attended a meeting, participated in a committee, or been more actively engaged in SHM, an online network might be a first step to increased involvement with SHM. Members could use this site to connect with other hospitalists in their area and share their interests and experience with others.

Along the way, they might learn about an SHM initiative they are interested in and connect with another hospitalist who working on this project and begin to have a dialogue. Throughout time, this person builds their network and establishes new connections. When it’s time to register for next year’s SHM Annual Meeting in Chicago, they already know a few faces in the crowd—and maybe a couple of them have become friends.

These are just a couple of ways I think SHM and our members might benefit from social networking. I am confident there are many, many more ways this technology can help our members and the hospital medicine community. What do you think? I would love to hear your thoughts and ideas. E-mail me at [email protected]. TH

Dear John Q. Hospitalist,

Recently, a pair of college students in our office presented an impressive summary of Web 2.0, including Facebook.com and LinkedIn.com, to the rest of the SHM staff.

As I listened to their presentation and heard the energy in their voices, I couldn’t help but think about my initial experience and excitement with the World Wide Web. Instead of doing homework, I spent many late nights searching the Internet looking for more information to help me create my first Web page. After countless hours of coding and debugging, as well as throwing the keyboard a time or two, I published my Web page and became a part of the Internet. I was hooked.

After listening to these students I was inspired to check out LinkedIn.com and create my own LinkedIn profile. While I did not stay up until the very early morning sending invites or completing every part of my profile, I found connections to old colleagues, college friends, high school buddies, and family members. Today, I eagerly await the flood of e-mail from people accepting me as a friend in their network, some of them members of SHM. I am hooked again.

Seeing SHM members on LinkedIn got me thinking about how SHM might use social networking technology. I think there is an opportunity here to create an interactive resource that will empower hospitalists to find other hospitalists, make connections, and build their own networks. I’m interested in getting your perspective. Do you think our members will use this type of an online resource?

Many social networking sites on the Internet grew out of individuals in an academic setting trying to find ways to connect with each other. I would imagine many of our student and resident members already are using social networking sites. Do you think this is the case? If so, what features and functions of a social networking tool do you think are most important? Is that different from a third-year resident, or a hospitalist who has been practicing hospital medicine for a number of years?

Johnson

I know I have thrown a bunch of questions at you, so let me share with you some ideas and maybe we can begin a dialogue that will help SHM find ways in which we can leverage social networking and other Web 2.0 tools.

One of the tasks in creating a LinkedIn account is selecting the college or institution you attended and the years in which you attended. Immediately after setting up my account I was able to see the names of other alumni who attended my university during my four years and invite old friends to join my network. I can envision a scenario where an SHM member indicates which medical school he or she attended and is able to see a list of other colleagues who attended at the same time.

For the general member, someone who hasn’t attended a meeting, participated in a committee, or been more actively engaged in SHM, an online network might be a first step to increased involvement with SHM. Members could use this site to connect with other hospitalists in their area and share their interests and experience with others.

Along the way, they might learn about an SHM initiative they are interested in and connect with another hospitalist who working on this project and begin to have a dialogue. Throughout time, this person builds their network and establishes new connections. When it’s time to register for next year’s SHM Annual Meeting in Chicago, they already know a few faces in the crowd—and maybe a couple of them have become friends.

These are just a couple of ways I think SHM and our members might benefit from social networking. I am confident there are many, many more ways this technology can help our members and the hospital medicine community. What do you think? I would love to hear your thoughts and ideas. E-mail me at [email protected]. TH

One of the best things about working in the membership department is seeing the passion and commitment our members have for the success of our specialty. You also get to hear about what is needed to elevate our field to the next level.

Throughout the past 10 years, hospitalists have been asking for a unique designation that clearly identifies those physicians whose career is committed to hospital medicine. SHM wants you to know that we have heard you and have responded to this need by creating a Fellowship in Hospital Medicine (FHM) program.

Hospitalists who earn their FHM designation will have a demonstrated commitment to hospital medicine, systems change, and quality improvement principals. There will be three designations available, each with increasingly more challenging selection criteria:

• Fellow in Hospital Medicine (FHM);
• Senior Fellow in Hospital Medicine (SFHM); and
• Master in Hospital Medicine (MHM).

Application requirements include:

• Five years as a practicing hospitalist;
• Demonstration of competencies tied to the SHM Core Competencies in Hospital Medicine, including systems improvement, quality initiatives, and clinical excellence; and
• Endorsement by two active SHM members.

Applications for Fellow in Hospital Medicine will be available this fall, and the first class will be inducted at Hospital Medicine 2009 in Chicago.

One of the best things about working in the membership department is seeing the passion and commitment our members have for the success of our specialty. You also get to hear about what is needed to elevate our field to the next level.

Throughout the past 10 years, hospitalists have been asking for a unique designation that clearly identifies those physicians whose career is committed to hospital medicine. SHM wants you to know that we have heard you and have responded to this need by creating a Fellowship in Hospital Medicine (FHM) program.

Hospitalists who earn their FHM designation will have a demonstrated commitment to hospital medicine, systems change, and quality improvement principals. There will be three designations available, each with increasingly more challenging selection criteria:

• Fellow in Hospital Medicine (FHM);
• Senior Fellow in Hospital Medicine (SFHM); and
• Master in Hospital Medicine (MHM).

Application requirements include:

• Five years as a practicing hospitalist;
• Demonstration of competencies tied to the SHM Core Competencies in Hospital Medicine, including systems improvement, quality initiatives, and clinical excellence; and
• Endorsement by two active SHM members.

Applications for Fellow in Hospital Medicine will be available this fall, and the first class will be inducted at Hospital Medicine 2009 in Chicago.

One of the best things about working in the membership department is seeing the passion and commitment our members have for the success of our specialty. You also get to hear about what is needed to elevate our field to the next level.

Throughout the past 10 years, hospitalists have been asking for a unique designation that clearly identifies those physicians whose career is committed to hospital medicine. SHM wants you to know that we have heard you and have responded to this need by creating a Fellowship in Hospital Medicine (FHM) program.

Hospitalists who earn their FHM designation will have a demonstrated commitment to hospital medicine, systems change, and quality improvement principals. There will be three designations available, each with increasingly more challenging selection criteria:

• Fellow in Hospital Medicine (FHM);
• Senior Fellow in Hospital Medicine (SFHM); and
• Master in Hospital Medicine (MHM).

Application requirements include:

• Five years as a practicing hospitalist;
• Demonstration of competencies tied to the SHM Core Competencies in Hospital Medicine, including systems improvement, quality initiatives, and clinical excellence; and
• Endorsement by two active SHM members.

Applications for Fellow in Hospital Medicine will be available this fall, and the first class will be inducted at Hospital Medicine 2009 in Chicago.

Supplement Editor:
Brian F. Mandell, MD, PhD

Contents

Despite its treatability, gout remains a problem: Confessions of a goutophile
Brian F. Mandell, MD, PhD

The pathogenesis of gout
H. Ralph Schumacher, Jr, MD

Clinical manifestations of hyperuricemia and gout
Brian F. Mandell, MD, PhD

Epidemiology of gout
Arthur L. Weaver, MD, MS

The role of hyperuricemia and gout in kidney and cardiovascular disease
N. Lawrence Edwards, MD

The gout diagnosis
Robin K. Dore, MD

The practical management of gout
H. Ralph Schumacher, Jr, MD, and Lan X. Chen, MD, PhD

Supplement Editor:
Brian F. Mandell, MD, PhD

Contents

Despite its treatability, gout remains a problem: Confessions of a goutophile
Brian F. Mandell, MD, PhD

The pathogenesis of gout
H. Ralph Schumacher, Jr, MD

Clinical manifestations of hyperuricemia and gout
Brian F. Mandell, MD, PhD

Epidemiology of gout
Arthur L. Weaver, MD, MS

The role of hyperuricemia and gout in kidney and cardiovascular disease
N. Lawrence Edwards, MD

The gout diagnosis
Robin K. Dore, MD

The practical management of gout
H. Ralph Schumacher, Jr, MD, and Lan X. Chen, MD, PhD

Supplement Editor:
Brian F. Mandell, MD, PhD

Contents

Despite its treatability, gout remains a problem: Confessions of a goutophile
Brian F. Mandell, MD, PhD

The pathogenesis of gout
H. Ralph Schumacher, Jr, MD

Clinical manifestations of hyperuricemia and gout
Brian F. Mandell, MD, PhD

Epidemiology of gout
Arthur L. Weaver, MD, MS

The role of hyperuricemia and gout in kidney and cardiovascular disease
N. Lawrence Edwards, MD

The gout diagnosis
Robin K. Dore, MD

The practical management of gout
H. Ralph Schumacher, Jr, MD, and Lan X. Chen, MD, PhD

Budnitz

Somebody oughta fix that. I’m sure you’ve heard that phrase from friends or relatives lamenting their recent visit to a hospital.

My grandmother will use just about any opportunity to inform me that hospitals make people sicker. I’ve tried to explain that her perspective is skewed. After all, a lot of her friends were pretty sick before they entered the hospital. But “it’s that place” she vows. “They oughta change it.” Fortunately for me, my grandmother still hasn’t figured out what I do professionally, so I’m not considered part of “they.” I let her rant to my husband, since he has the letters MD after his name.

The truth, as you know, is many hospital medicine physicians and their teams are working their tails off trying to improve inpatient care. Much like my grandmother, hospital administrators haven’t identified who the “they” (change agents) are or what exactly the “it” (practices and systems that lead to suboptimal care) is that needs to be changed. It often is unclear how quality improvement initiatives affect the bottom line or which initiatives will ultimately improve outcomes. Today, a considerable amount of improvement efforts depend on the good will and perseverance of a few champions working with minimal institutional support.

The Hospital Quality and Patient Safety Committee (HQPSC) and SHM leadership recently convened a summit to define a vision for the optimal hospital stay and determine how to best train and support hospitalists as leaders and change agents.

The HQPSC and summit participants concluded SHM is, and should be, a national leader in quality improvement efforts including aspects of education, clinical care, and political advocacy for the hospital setting. To that end, the following strategy recently was submitted by the HQPSC and approved by the SHM Board of Directors to promote development of local, regional, and national infrastructures that support quality and patient safety:

Advance a national quality agenda for hospitals and hospitalists.

• Create a task force reporting to the HQPSC that partners with stakeholders to define the “ideal hospital stay” and promote quality improvement;
• Inform federal accrediting and policy-making groups about the effect of current quality measures and changes required to better support the “ideal hospital stay”;
• Advocate for the alignment of reimbursement practices that reward providers and institutions that demonstrate value and translate these practices into improved quality and patient safety;
• Establish an Acute Care Collaborative (ACC) comprising national organizations representing nurses, pharmacists, case managers, social workers, and other allied medical professionals. The ACC might be expanded to include other key physician groups (e.g., emergency physicians, geriatricians, intensivists); and
• Determine what other key national organizations are doing in quality improvement (QI) and look for opportunities for SHM to partner in these efforts.

Develop educational programs and technical support tools for all practicing hospitalists (entry level to QI leaders) engaged in quality improvement efforts.

• Delineate entry-level and advanced quality improvement offerings. Develop offerings specifically for advanced level participants;
• Expand mentored implementation programs to accommodate more participants and assess the need for other types of programs that provide longitudinal support or coaching;
• Expand current offerings, including resource rooms, mentored implementation, and expert training sessions, to other disease states, system processes, and special populations with attention to coordinating this with SHM’s The Core Competencies in Hospital Medicine: A Framework for Curriculum Development;
• Assess the need for new instructional modalities to reach a broader audience (e.g., Web based self-study modules); and
• Promote QI training in medical school, residency, and fellowship programs. Promote systems-based practice and QI throughout the continuum of education. This would include programs that engage medical students, residents, and fellows as well as the development of performance improvement modules (PIMs) for the American Board of Internal Medicine.

Improves the perceived value of implementing and sustaining QI efforts, and hospitalist leadership of those efforts.

• Advocate directly to the C-suites of hospitals to facilitate alignment of incentives that support hospitalists leading quality initiatives;
• Conduct a survey to quantify resources needed for hospitalists to successfully lead quality initiatives and develop safety programs. Develop a white paper based on survey results and distribute it to the C-suite;
• Encourage QI research that creates evidence and outcomes that can influence C-suites to commit adequate resources to QI activities;
• Explore opportunities to use existing local and national infrastructures to promote a more proactive and evidence-based approach to quality and safety rather than reactive and compliance-oriented quality projects; and
• Create a monthly column in The Hospitalist spotlighting QI efforts and assign staff to recruit submissions of “improvement stories” for the Web site.

Evaluate effectiveness of SHM’s current QI resources, educational offerings. SHM needs to assess its current offerings to understand and improve their effectiveness. Process and impact data are needed to obtain external money to create or sustain QI offerings.

• Create robust evaluations and collect better data to assess use and impact of resource rooms, quality precourses, and other SHM offerings.

Promote and support hospitalists as quality improvement experts. Contribute to the “new science” of quality improvement.

• Partner with the Research Committee to define and publish key areas in need of future research related to quality improvement. Advocate to granting agencies to put out RFPs that will help define the ideal hospital stay and support SHM’s research agenda;
• Partner with federal agencies to assess the value of current performance measures and facilitate development of more reliable and meaningful measures;
• Develop trainings for hospitalists on the methods and science of quality improvement research;
• Partner with the Research Committee to develop a research network; and
• Seek money to support demonstration projects that support our quality agenda.

Another goal, to promote development and adoption of health information technology and decision support tools that advance quality and patient safety, recently was discussed by the HQPSC and will be integrated into the next stage of planning.

Next Steps

SHM has an impressive history of working with its members to develop and implement quality initiatives. Our programs have helped reduce rates of venous thromboembolisms, improve glycemic control, and improve the discharge process. Our highly praised online resource rooms provide free tutorials in QI and implementation guides for specific interventions.

More than 250 healthcare professionals have completed our QI pre-course and more than one thousand hospitalists have completed our leadership programs. This momentum, combined with the acceleration of national interest in quality and patient safety, brings an unprecedented opportunity for SHM to advance hospital medicine, promote the highest quality care for our patients, and position hospitalists to be leaders in transforming hospital care.

During the next year, HQPSC will be translating this strategy into specific activities. SHM staff, HQPSC, and other members are developing additional training programs and technical tools.

If you are interested in becoming more involved in SHM’s quality initiatives, please contact me at [email protected].

If you have a QI success story to share, please consider submission to the Improvement Stories section of the online Resource Rooms or The Hospitalist.

I thank all of you who are part of the “they” who are working tirelessly with SHM to fix “it.” Together we can move mountains, or something more impervious like healthcare systems and performance measures.

Budnitz

Somebody oughta fix that. I’m sure you’ve heard that phrase from friends or relatives lamenting their recent visit to a hospital.

My grandmother will use just about any opportunity to inform me that hospitals make people sicker. I’ve tried to explain that her perspective is skewed. After all, a lot of her friends were pretty sick before they entered the hospital. But “it’s that place” she vows. “They oughta change it.” Fortunately for me, my grandmother still hasn’t figured out what I do professionally, so I’m not considered part of “they.” I let her rant to my husband, since he has the letters MD after his name.

The truth, as you know, is many hospital medicine physicians and their teams are working their tails off trying to improve inpatient care. Much like my grandmother, hospital administrators haven’t identified who the “they” (change agents) are or what exactly the “it” (practices and systems that lead to suboptimal care) is that needs to be changed. It often is unclear how quality improvement initiatives affect the bottom line or which initiatives will ultimately improve outcomes. Today, a considerable amount of improvement efforts depend on the good will and perseverance of a few champions working with minimal institutional support.

The Hospital Quality and Patient Safety Committee (HQPSC) and SHM leadership recently convened a summit to define a vision for the optimal hospital stay and determine how to best train and support hospitalists as leaders and change agents.

The HQPSC and summit participants concluded SHM is, and should be, a national leader in quality improvement efforts including aspects of education, clinical care, and political advocacy for the hospital setting. To that end, the following strategy recently was submitted by the HQPSC and approved by the SHM Board of Directors to promote development of local, regional, and national infrastructures that support quality and patient safety:

Advance a national quality agenda for hospitals and hospitalists.

• Create a task force reporting to the HQPSC that partners with stakeholders to define the “ideal hospital stay” and promote quality improvement;
• Inform federal accrediting and policy-making groups about the effect of current quality measures and changes required to better support the “ideal hospital stay”;
• Advocate for the alignment of reimbursement practices that reward providers and institutions that demonstrate value and translate these practices into improved quality and patient safety;
• Establish an Acute Care Collaborative (ACC) comprising national organizations representing nurses, pharmacists, case managers, social workers, and other allied medical professionals. The ACC might be expanded to include other key physician groups (e.g., emergency physicians, geriatricians, intensivists); and
• Determine what other key national organizations are doing in quality improvement (QI) and look for opportunities for SHM to partner in these efforts.

Develop educational programs and technical support tools for all practicing hospitalists (entry level to QI leaders) engaged in quality improvement efforts.

• Delineate entry-level and advanced quality improvement offerings. Develop offerings specifically for advanced level participants;
• Expand mentored implementation programs to accommodate more participants and assess the need for other types of programs that provide longitudinal support or coaching;
• Expand current offerings, including resource rooms, mentored implementation, and expert training sessions, to other disease states, system processes, and special populations with attention to coordinating this with SHM’s The Core Competencies in Hospital Medicine: A Framework for Curriculum Development;
• Assess the need for new instructional modalities to reach a broader audience (e.g., Web based self-study modules); and
• Promote QI training in medical school, residency, and fellowship programs. Promote systems-based practice and QI throughout the continuum of education. This would include programs that engage medical students, residents, and fellows as well as the development of performance improvement modules (PIMs) for the American Board of Internal Medicine.

Improves the perceived value of implementing and sustaining QI efforts, and hospitalist leadership of those efforts.

• Advocate directly to the C-suites of hospitals to facilitate alignment of incentives that support hospitalists leading quality initiatives;
• Conduct a survey to quantify resources needed for hospitalists to successfully lead quality initiatives and develop safety programs. Develop a white paper based on survey results and distribute it to the C-suite;
• Encourage QI research that creates evidence and outcomes that can influence C-suites to commit adequate resources to QI activities;
• Explore opportunities to use existing local and national infrastructures to promote a more proactive and evidence-based approach to quality and safety rather than reactive and compliance-oriented quality projects; and
• Create a monthly column in The Hospitalist spotlighting QI efforts and assign staff to recruit submissions of “improvement stories” for the Web site.

Evaluate effectiveness of SHM’s current QI resources, educational offerings. SHM needs to assess its current offerings to understand and improve their effectiveness. Process and impact data are needed to obtain external money to create or sustain QI offerings.

• Create robust evaluations and collect better data to assess use and impact of resource rooms, quality precourses, and other SHM offerings.

Promote and support hospitalists as quality improvement experts. Contribute to the “new science” of quality improvement.

• Partner with the Research Committee to define and publish key areas in need of future research related to quality improvement. Advocate to granting agencies to put out RFPs that will help define the ideal hospital stay and support SHM’s research agenda;
• Partner with federal agencies to assess the value of current performance measures and facilitate development of more reliable and meaningful measures;
• Develop trainings for hospitalists on the methods and science of quality improvement research;
• Partner with the Research Committee to develop a research network; and
• Seek money to support demonstration projects that support our quality agenda.

Another goal, to promote development and adoption of health information technology and decision support tools that advance quality and patient safety, recently was discussed by the HQPSC and will be integrated into the next stage of planning.

Next Steps

SHM has an impressive history of working with its members to develop and implement quality initiatives. Our programs have helped reduce rates of venous thromboembolisms, improve glycemic control, and improve the discharge process. Our highly praised online resource rooms provide free tutorials in QI and implementation guides for specific interventions.

More than 250 healthcare professionals have completed our QI pre-course and more than one thousand hospitalists have completed our leadership programs. This momentum, combined with the acceleration of national interest in quality and patient safety, brings an unprecedented opportunity for SHM to advance hospital medicine, promote the highest quality care for our patients, and position hospitalists to be leaders in transforming hospital care.

During the next year, HQPSC will be translating this strategy into specific activities. SHM staff, HQPSC, and other members are developing additional training programs and technical tools.

If you are interested in becoming more involved in SHM’s quality initiatives, please contact me at [email protected].

If you have a QI success story to share, please consider submission to the Improvement Stories section of the online Resource Rooms or The Hospitalist.

I thank all of you who are part of the “they” who are working tirelessly with SHM to fix “it.” Together we can move mountains, or something more impervious like healthcare systems and performance measures.

Budnitz

Somebody oughta fix that. I’m sure you’ve heard that phrase from friends or relatives lamenting their recent visit to a hospital.

My grandmother will use just about any opportunity to inform me that hospitals make people sicker. I’ve tried to explain that her perspective is skewed. After all, a lot of her friends were pretty sick before they entered the hospital. But “it’s that place” she vows. “They oughta change it.” Fortunately for me, my grandmother still hasn’t figured out what I do professionally, so I’m not considered part of “they.” I let her rant to my husband, since he has the letters MD after his name.

The truth, as you know, is many hospital medicine physicians and their teams are working their tails off trying to improve inpatient care. Much like my grandmother, hospital administrators haven’t identified who the “they” (change agents) are or what exactly the “it” (practices and systems that lead to suboptimal care) is that needs to be changed. It often is unclear how quality improvement initiatives affect the bottom line or which initiatives will ultimately improve outcomes. Today, a considerable amount of improvement efforts depend on the good will and perseverance of a few champions working with minimal institutional support.

The Hospital Quality and Patient Safety Committee (HQPSC) and SHM leadership recently convened a summit to define a vision for the optimal hospital stay and determine how to best train and support hospitalists as leaders and change agents.

The HQPSC and summit participants concluded SHM is, and should be, a national leader in quality improvement efforts including aspects of education, clinical care, and political advocacy for the hospital setting. To that end, the following strategy recently was submitted by the HQPSC and approved by the SHM Board of Directors to promote development of local, regional, and national infrastructures that support quality and patient safety:

Advance a national quality agenda for hospitals and hospitalists.

• Create a task force reporting to the HQPSC that partners with stakeholders to define the “ideal hospital stay” and promote quality improvement;
• Inform federal accrediting and policy-making groups about the effect of current quality measures and changes required to better support the “ideal hospital stay”;
• Advocate for the alignment of reimbursement practices that reward providers and institutions that demonstrate value and translate these practices into improved quality and patient safety;
• Establish an Acute Care Collaborative (ACC) comprising national organizations representing nurses, pharmacists, case managers, social workers, and other allied medical professionals. The ACC might be expanded to include other key physician groups (e.g., emergency physicians, geriatricians, intensivists); and
• Determine what other key national organizations are doing in quality improvement (QI) and look for opportunities for SHM to partner in these efforts.

Develop educational programs and technical support tools for all practicing hospitalists (entry level to QI leaders) engaged in quality improvement efforts.

• Delineate entry-level and advanced quality improvement offerings. Develop offerings specifically for advanced level participants;
• Expand mentored implementation programs to accommodate more participants and assess the need for other types of programs that provide longitudinal support or coaching;
• Expand current offerings, including resource rooms, mentored implementation, and expert training sessions, to other disease states, system processes, and special populations with attention to coordinating this with SHM’s The Core Competencies in Hospital Medicine: A Framework for Curriculum Development;
• Assess the need for new instructional modalities to reach a broader audience (e.g., Web based self-study modules); and
• Promote QI training in medical school, residency, and fellowship programs. Promote systems-based practice and QI throughout the continuum of education. This would include programs that engage medical students, residents, and fellows as well as the development of performance improvement modules (PIMs) for the American Board of Internal Medicine.

Improves the perceived value of implementing and sustaining QI efforts, and hospitalist leadership of those efforts.

• Advocate directly to the C-suites of hospitals to facilitate alignment of incentives that support hospitalists leading quality initiatives;
• Conduct a survey to quantify resources needed for hospitalists to successfully lead quality initiatives and develop safety programs. Develop a white paper based on survey results and distribute it to the C-suite;
• Encourage QI research that creates evidence and outcomes that can influence C-suites to commit adequate resources to QI activities;
• Explore opportunities to use existing local and national infrastructures to promote a more proactive and evidence-based approach to quality and safety rather than reactive and compliance-oriented quality projects; and
• Create a monthly column in The Hospitalist spotlighting QI efforts and assign staff to recruit submissions of “improvement stories” for the Web site.

Evaluate effectiveness of SHM’s current QI resources, educational offerings. SHM needs to assess its current offerings to understand and improve their effectiveness. Process and impact data are needed to obtain external money to create or sustain QI offerings.

• Create robust evaluations and collect better data to assess use and impact of resource rooms, quality precourses, and other SHM offerings.

Promote and support hospitalists as quality improvement experts. Contribute to the “new science” of quality improvement.

• Partner with the Research Committee to define and publish key areas in need of future research related to quality improvement. Advocate to granting agencies to put out RFPs that will help define the ideal hospital stay and support SHM’s research agenda;
• Partner with federal agencies to assess the value of current performance measures and facilitate development of more reliable and meaningful measures;
• Develop trainings for hospitalists on the methods and science of quality improvement research;
• Partner with the Research Committee to develop a research network; and
• Seek money to support demonstration projects that support our quality agenda.

Another goal, to promote development and adoption of health information technology and decision support tools that advance quality and patient safety, recently was discussed by the HQPSC and will be integrated into the next stage of planning.

Next Steps

SHM has an impressive history of working with its members to develop and implement quality initiatives. Our programs have helped reduce rates of venous thromboembolisms, improve glycemic control, and improve the discharge process. Our highly praised online resource rooms provide free tutorials in QI and implementation guides for specific interventions.

More than 250 healthcare professionals have completed our QI pre-course and more than one thousand hospitalists have completed our leadership programs. This momentum, combined with the acceleration of national interest in quality and patient safety, brings an unprecedented opportunity for SHM to advance hospital medicine, promote the highest quality care for our patients, and position hospitalists to be leaders in transforming hospital care.

During the next year, HQPSC will be translating this strategy into specific activities. SHM staff, HQPSC, and other members are developing additional training programs and technical tools.

If you are interested in becoming more involved in SHM’s quality initiatives, please contact me at [email protected].

If you have a QI success story to share, please consider submission to the Improvement Stories section of the online Resource Rooms or The Hospitalist.

I thank all of you who are part of the “they” who are working tirelessly with SHM to fix “it.” Together we can move mountains, or something more impervious like healthcare systems and performance measures.

The rapidly expanding field of hospital medicine has spurred a growing number of textbooks devoted to the specialty. Textbooks by some of the specialty’s leading voices are available to those keen on honing their knowledge.

Ranging in scope from practice management issues to clinical synopses, titles include:

• “Hospitalists: A Guide to Building and Sustaining a Successful Program” by SHM founders John Nelson, MD, and Win Whitcomb, MD, and Joe Miller, SHM’s executive adviser to the CEO. (Health Administration Press, 2007, $72);
• “Comprehensive Hospital Medicine,” by Mark Williams, MD, chief, division of hospital medicine, Feinberg School of Medicine, Chicago (Elsevier, 2007, $109);
• “Hospital Medicine Secrets,” by The Hospitalist physician editor Jeff Glasheen, MD (Mosby/Elsevier, 2007, $39);
• “Understanding Patient Safety” by Robert Wachter, MD, chief of the Division of Hospital Medicine, and chief of the Medical Service at the University of California, at San Francisco Medical Center, and author of “Wachter’s World,” a blog featured on The Hospitalist Web site (McGraw-Hill, 2007, $35);
• “Hospital Medicine: Just the Facts,” by Sylvia McKean, MD, director, hospitalist service, Brigham and Women’s Hospital, Boston (McGraw-Hill, 2008, $50);
• “First Exposure. Internal Medicine: Hospital Medicine” by Charles Griffith, MD, inpatient internal medicine clerkship director, and Andrew R. Hoellein, MD, outpatient internal medicine clerkship director, Department of Internal Medicine, University of Kentucky, Lexington (McGraw-Hill, 2007, $34); and
• “Tools and Strategies for an Effective Hospitalist Program” by Jeffrey R. Dichter, MD, and Kenneth G. Simone, MD (HCPro, 2008, $299).

SHM’s book

SHM’s offering in the arena reinforces the ideas of “the critical need for leadership of HMGs and the need to create an ownership mentality for hospitalists within an HMG,” Miller says. “The book is filled with examples, tools, and checklists” and has sold approximately 500 copies so far.

The newest text, just off the press in May, is Dr. McKean’s. “This book provides concise, templated information designed to save the clinician valuable time,” she says. It also has a variety of uses, including exam review, clinical reference, point-of-care lookup, [and] quick updates in hospital medicine for those attending on the wards. It covers vital information on issues in administration and management.”

Dr. Wachter wrote his text “because I didn’t see any book for those seeking to learn the key clinical, organizational, and systems issues in patient safety,” he says. “I tried to write it in a lively and accessible style and fill it with illustrative cases and analyses, as well as up-to-date tables, graphics, references, and tools. My goal was to introduce the patient safety field to physicians—particularly hospitalists—nurses, pharmacists, and hospital administrators, as well as to trainees in these fields. [I hope it’s a] go-to book for experienced clinicians and nonclinicians alike.”

Already in its second printing, Dr. Wachter estimates it has sold between 7,500 and 10,000 copies. He plans to update the book every two years and is working on producing some Web-based learning modules. TH

The rapidly expanding field of hospital medicine has spurred a growing number of textbooks devoted to the specialty. Textbooks by some of the specialty’s leading voices are available to those keen on honing their knowledge.

Ranging in scope from practice management issues to clinical synopses, titles include:

• “Hospitalists: A Guide to Building and Sustaining a Successful Program” by SHM founders John Nelson, MD, and Win Whitcomb, MD, and Joe Miller, SHM’s executive adviser to the CEO. (Health Administration Press, 2007, $72);
• “Comprehensive Hospital Medicine,” by Mark Williams, MD, chief, division of hospital medicine, Feinberg School of Medicine, Chicago (Elsevier, 2007, $109);
• “Hospital Medicine Secrets,” by The Hospitalist physician editor Jeff Glasheen, MD (Mosby/Elsevier, 2007, $39);
• “Understanding Patient Safety” by Robert Wachter, MD, chief of the Division of Hospital Medicine, and chief of the Medical Service at the University of California, at San Francisco Medical Center, and author of “Wachter’s World,” a blog featured on The Hospitalist Web site (McGraw-Hill, 2007, $35);
• “Hospital Medicine: Just the Facts,” by Sylvia McKean, MD, director, hospitalist service, Brigham and Women’s Hospital, Boston (McGraw-Hill, 2008, $50);
• “First Exposure. Internal Medicine: Hospital Medicine” by Charles Griffith, MD, inpatient internal medicine clerkship director, and Andrew R. Hoellein, MD, outpatient internal medicine clerkship director, Department of Internal Medicine, University of Kentucky, Lexington (McGraw-Hill, 2007, $34); and
• “Tools and Strategies for an Effective Hospitalist Program” by Jeffrey R. Dichter, MD, and Kenneth G. Simone, MD (HCPro, 2008, $299).

SHM’s book

SHM’s offering in the arena reinforces the ideas of “the critical need for leadership of HMGs and the need to create an ownership mentality for hospitalists within an HMG,” Miller says. “The book is filled with examples, tools, and checklists” and has sold approximately 500 copies so far.

The newest text, just off the press in May, is Dr. McKean’s. “This book provides concise, templated information designed to save the clinician valuable time,” she says. It also has a variety of uses, including exam review, clinical reference, point-of-care lookup, [and] quick updates in hospital medicine for those attending on the wards. It covers vital information on issues in administration and management.”

Dr. Wachter wrote his text “because I didn’t see any book for those seeking to learn the key clinical, organizational, and systems issues in patient safety,” he says. “I tried to write it in a lively and accessible style and fill it with illustrative cases and analyses, as well as up-to-date tables, graphics, references, and tools. My goal was to introduce the patient safety field to physicians—particularly hospitalists—nurses, pharmacists, and hospital administrators, as well as to trainees in these fields. [I hope it’s a] go-to book for experienced clinicians and nonclinicians alike.”

Already in its second printing, Dr. Wachter estimates it has sold between 7,500 and 10,000 copies. He plans to update the book every two years and is working on producing some Web-based learning modules. TH

The rapidly expanding field of hospital medicine has spurred a growing number of textbooks devoted to the specialty. Textbooks by some of the specialty’s leading voices are available to those keen on honing their knowledge.

Ranging in scope from practice management issues to clinical synopses, titles include:

• “Hospitalists: A Guide to Building and Sustaining a Successful Program” by SHM founders John Nelson, MD, and Win Whitcomb, MD, and Joe Miller, SHM’s executive adviser to the CEO. (Health Administration Press, 2007, $72);
• “Comprehensive Hospital Medicine,” by Mark Williams, MD, chief, division of hospital medicine, Feinberg School of Medicine, Chicago (Elsevier, 2007, $109);
• “Hospital Medicine Secrets,” by The Hospitalist physician editor Jeff Glasheen, MD (Mosby/Elsevier, 2007, $39);
• “Understanding Patient Safety” by Robert Wachter, MD, chief of the Division of Hospital Medicine, and chief of the Medical Service at the University of California, at San Francisco Medical Center, and author of “Wachter’s World,” a blog featured on The Hospitalist Web site (McGraw-Hill, 2007, $35);
• “Hospital Medicine: Just the Facts,” by Sylvia McKean, MD, director, hospitalist service, Brigham and Women’s Hospital, Boston (McGraw-Hill, 2008, $50);
• “First Exposure. Internal Medicine: Hospital Medicine” by Charles Griffith, MD, inpatient internal medicine clerkship director, and Andrew R. Hoellein, MD, outpatient internal medicine clerkship director, Department of Internal Medicine, University of Kentucky, Lexington (McGraw-Hill, 2007, $34); and
• “Tools and Strategies for an Effective Hospitalist Program” by Jeffrey R. Dichter, MD, and Kenneth G. Simone, MD (HCPro, 2008, $299).

SHM’s book

SHM’s offering in the arena reinforces the ideas of “the critical need for leadership of HMGs and the need to create an ownership mentality for hospitalists within an HMG,” Miller says. “The book is filled with examples, tools, and checklists” and has sold approximately 500 copies so far.

The newest text, just off the press in May, is Dr. McKean’s. “This book provides concise, templated information designed to save the clinician valuable time,” she says. It also has a variety of uses, including exam review, clinical reference, point-of-care lookup, [and] quick updates in hospital medicine for those attending on the wards. It covers vital information on issues in administration and management.”

Dr. Wachter wrote his text “because I didn’t see any book for those seeking to learn the key clinical, organizational, and systems issues in patient safety,” he says. “I tried to write it in a lively and accessible style and fill it with illustrative cases and analyses, as well as up-to-date tables, graphics, references, and tools. My goal was to introduce the patient safety field to physicians—particularly hospitalists—nurses, pharmacists, and hospital administrators, as well as to trainees in these fields. [I hope it’s a] go-to book for experienced clinicians and nonclinicians alike.”

Already in its second printing, Dr. Wachter estimates it has sold between 7,500 and 10,000 copies. He plans to update the book every two years and is working on producing some Web-based learning modules. TH

It used to be so simple. The relationship between doctors and hospitals was a straightforward quid pro quo.

Hospitals granted privileges to physicians to admit and treat their patients, and the physicians returned the favor by assuming unpaid responsibilities like taking call, providing care to uninsured or emergency patients, and serving on administrative committees.

The hospital was like a friendly club whose members exchanged benefits for duties—a win-win situation. No more.

“You used to be part of a fraternity,” explains Win Whitcomb, MD, director of performance improvement at Mercy Medical Center in Springfield, Mass., and a co-founder of SHM. “There were social rewards. There was opportunity for collegial interchange.”

Economic pressure has taken that all away. “The pace of care has greatly intensified, and the financial reward system has deteriorated significantly,” Dr. Whitcomb continues. “We treat larger numbers of uninsured patients with chronic unmanaged illnesses that require intervention. The reward system for physicians to take call and fulfill their obligation to the hospital no longer matches the responsibility.”

To illustrate the change, Dr. Whitcomb offers an example: “We have some days of the month where the call roster for general surgeries has vacancies. A month ago we had to send a patient to another hospital for an appendectomy.”

It is not an isolated instance. “Every hospital is struggling with the fact that many physicians don’t view unassigned call as a part of membership on the staff; they want to be paid for it,” says SHM President Patrick Cawley, MD, executive medical director of the Medical University of South Carolina (MUSC). And extra “pay” for services that used to be rendered gratis is one thing today’s strapped hospitals can little afford.

Committee staffing is another area undergoing change. Attending physicians are simply declining the duty. Neal Axon, MD, a hospitalist and assistant professor of medicine and pediatrics at MUSC, has seen the transformation firsthand. At one hospital his service covered, he saw the following: “At the first staff meeting there were 50 people; there was food, liquor. It was social and attendance was mandatory. You had to make three or four meetings a year to be on medical staff at this hospital.” But then, he says, attendance waned, and in the last year “dropped off precipitously.”

The old ways don’t work so what will replace them? “The point is that both physicians and hospitals need to put something on the table to collaborate,” Dr. Cawley says. “Many are saying that the hospital-physician relationship needs to change, but everyone is still feeling their way through it. What does it mean?”

SHM’s CEO Larry Wellikson, MD, sees a layered structure ahead. “Clearly the system is evolving into three kinds of physicians who use the hospital,” he says. “We are not advocating for it—just saying what it is. This is what is evolving, and hospital staffs need to see this is coming.” His three kinds of physicians are categorized by their relationship to the hospital.

The home team: “The first group is those physicians who work only at the hospital,” he says. “Their professional life is with the hospital as an institution: hospitalists, ER doctors, critical care physicians, and sometimes the anesthesiologists and radiologists. The hospital is the location of their work and provides the tools to do their job. If the hospital works well, they can do their job well. If hospital is dysfunctional, they can’t work well.”

He describes their relationship to the hospital with an anecdote: “When I was regular physician who came to the hospital just to see my patient, if they couldn’t find the chart I would scream and yell about that one patient.” Every physician faced with a missing chart thinks of it as an individual problem. “But now as a hospitalist, I try to fix the system, because all my patients are affected,” he says. “Hospitalists are on the inside trying to make it work.”

Important visitors: The second group, whom Dr. Wellikson calls important visitors, has a totally different relationship with the hospital. These are the cardiologists, orthopedic surgeons, and other medical specialties. “They are very important,” he says. “But they use the hospital intermittently and are not as tightly connected to it.”

Even so, they desire a high-quality hospital for their patients and will be willing to help set performance standards to achieve it. But their interest may not extend to patients who are not their own. “If the hospital says you have to also take care of free patients, they may choose not to,” Dr. Wellikson notes. “In fact, sometimes they have their own outpatient centers,” making them direct competitors as their competing practices sap revenue from money-making patients and procedures—all the while sending the sickest and costliest patients to the hospital.

Office-based physicians: The last of Dr. Wellikson’s groups is office-based physicians. These are the doctors who once made daily morning and evening rounds of their hospitalized patients but are now infrequently found at the bedside. “They are the physicians who don’t come to the hospital anymore: primary care physicians, endocrinologists, rheumatologists, neurologists, physicians who do all their surgery as outpatient procedures,” Dr. Wellikson explains.

This upheaval is due to tectonic shifts in both medical economics and lifestyle preferences. “Because the reimbursement for care has gone down, physicians have to see more patients to make the same amount of money,” explains Dr. Wellikson. Turning their hospitalized patients over to hospitalists allows office-based physicians to maximize their income and optimize their time.

“It increases satisfaction, limits the hours you spend in the hospital, and puts some boundaries on your work day,” Dr. Cawley says.

“Doctors want more a predictable lifestyle,” Dr. Whitcomb says. In fact, their absence is already a fait accompli in many community hospitals.

As Dr. Axon succinctly puts it: “The primary care doc has left the building.”

Dr. Cawley believes the new system is a relief to many office-based physicians. “Some do miss going to the hospital and seeing other physicians to network with them. Some miss taking care of their acute in-care patients. But I think most are relieved to not have to go to hospital. They say, ‘No, things are better this way.’”

With so many other physicians withdrawing from hospitals to their offices and clinics, Dr. Wellikson believes hospitalists will become increasingly crucial to the institution’s operation and governance. “Now the home team is going to be more active; how you staff, how you make the hospital more efficient,” he says. “The inside physicians will be much more interactive. That’s why hospital medicine has grown so rapidly.”

The explosive expansion of hospital medicine as a specialty is a direct result of the need to increase efficiency and quality standards in this new hospital atmosphere.

In addition, good home teams create a milieu in which other physicians—the important visitors (cardiologists, surgeons, orthopedists)—will want to work. “My job (as a hospitalist) is to create an environment where you can come in and do your surgery,” Dr. Wellikson points out.

The home team offers something else too: medical expertise. Providing post-operative care is not cost-effective for many surgeons. “The surgical specialists are not paid to manage medical issues,” Dr. Cawley says. “It takes time and if somebody else can manage it, that’s great.” That somebody is often a hospitalist. “There is a quality-control aspect as well,” he adds. “With hospitalists focusing on medical issues, the result is better patient care.”

Melding these groups of physicians with disparate interests and responsibilities is the next challenge for hospital leadership. It is a challenge fraught with potential pitfalls. As Dr. Wellikson explains, “The biggest obstacle is that physicians don’t do change very well.”

Administrators will turn to their institution’s hospitalists (both hospital-employed and contracted) to effect these changes and ensure overall standards and efficiency.

“I think hospitalists are in a position to bridge the gap between administrators and medical staff,” says David Yu, MD, medical director of hospitalist services at Decatur Memorial Hospital in Illinois. “I think that’s why there will be more and more hospitalists in leadership positions. That’s why hospitalists are unique: they have their feet in both worlds.”

Dr. Wellikson believes the home team will step up to the plate and take over many of the leadership duties of the new hospital.

Kenneth Patrick, MD, the ICU director of Chestnut Hill Hospital in Philadelphia, sounds a more cautionary note. Dr. Patrick, a trained hospitalist and intensivist, believes the demise of the old “hospital privilege” model is dissolving ties between physicians and their workplace. “I think younger physicians will be much more transient and more concerned with their position, work hours, and pay,” he says.

He sees a young workforce—whether hospital or office-based—as more disengaged than physicians used to be. “They will meet hospital standards, but not be actively involved in developing them,” he believes. That will be left to a small group of hospital-based physicians “who will voluntarily come forward because it is their civic responsibility. It would be nice if more physicians would work on committees, but they look at them like jury duty and they don’t want to serve.”

Dr. Patrick

“The question everyone asks is ‘What’s in it for me?’” Dr. Yu says. He notes a common sticking point: the requirement for increased documentation, which often means more work for doctors. “I think administrators are going to be in shock if they think practitioners are going to line up and say, ‘Well that’s great for the hospital.’”

The key to cooperation, says Dr. Yu, is the linking of changes to mutual benefit and patient welfare: “The administrators have to communicate that in the long run everyone will gain and it will ultimately lead to better patient care. You have to share your vision, inspire, motivate, and develop a culture of providing quality care. It’s easier said than done, but it’s the essence of medical care.”

What about patients? How do they react when a group of strangers takes over their hospital care rather than the primary care physician they often have gotten to know and trust for years? “Wanting your doctor present is counterbalanced by not having your doctor in the house,” Dr. Axon says. “Now you can see a physician anytime during the day.” And most patients are glad for the tradeoff. Dr. Yu has found the same dynamic with his patients at Decatur Memorial Hospital. “I can just count on one hand patients who were not happy the primary care physician wasn’t there,” he says. “Patients are more concerned with having their problems solved than with who is solving them.” And he makes sure his hospitalist staff never undermines the office based physicians. “We always say we are not better physicians, we are just more available.”

While they may have left the hospital, office-based physicians still will be a large presence in it by advocating for their patients. “If my whole currency is, ‘Do I have hospital privileges?’ then all my decisions are based on that,” Dr. Wellikson says.

Armed with the power of their patient referrals, office-based physicians will be able to demand that hospitals show proof of performance—thus becoming their patients’ ombudsmen. “I’m your shopper for the best healthcare, so the hospital has to step up to the plate and make sure it gets the business,” Dr. Wellikson explains. “They want standards because their patients need the best treatment, and they will have a choice of which hospital to put their patients into. If I now have a choice of three hospitals, I am looking to see that you are the Lexus of healthcare for my patients.”

Looking out for their patients’ interests is not the only way office-based physicians will continue to affect hospitals. As in-patient revenue declines, hospitals must look to the outpatient side to make up the difference. “The hospital is lucky if they break even on the inpatient side; they get the vast majority of money on the outpatient side: testing and procedures that private attendings are sending to the hospital,” Dr. Yu says.

He cautions against alienating those private practitioners by forcing change that is not mutually beneficial. “If you alienate them, you might lose money because they can send their patients to a different institution,” he warns. “These are the same doctors that never admit patients but do order the outpatient ultrasounds, blood tests, and therapies that are all money makers for the hospital. Why would you want to alienate these physicians?”

Dr. Patrick agrees: office-based physicians and hospitalists need each other. “I have to work with the primaries,” he says. “They are my source of referrals.”

There is another group that hospitals must learn to court, according to Dr. Axon: its own hospitalists. “I think you will see more innovative solutions to problems of recruiting hospital-based physicians to perform these functions,” he says. “For that to happen, the doctors will need to get more out of it. Many hospitalist groups are in a quandary; they are expected to do all these extra things, but pay is closely liked to clinical production and the number of patients they see. Those incentives will have to be aligned.”

All of which increases the reliance on—and importance of—those physicians who do work in the hospital—the home team. As Dr. Yu puts it: “I think the hospitalist model, whether you like or hate it, is the wave of the future.” TH

Carol Berczuk is a journalist based in New York.

It used to be so simple. The relationship between doctors and hospitals was a straightforward quid pro quo.

Hospitals granted privileges to physicians to admit and treat their patients, and the physicians returned the favor by assuming unpaid responsibilities like taking call, providing care to uninsured or emergency patients, and serving on administrative committees.

The hospital was like a friendly club whose members exchanged benefits for duties—a win-win situation. No more.

“You used to be part of a fraternity,” explains Win Whitcomb, MD, director of performance improvement at Mercy Medical Center in Springfield, Mass., and a co-founder of SHM. “There were social rewards. There was opportunity for collegial interchange.”

Economic pressure has taken that all away. “The pace of care has greatly intensified, and the financial reward system has deteriorated significantly,” Dr. Whitcomb continues. “We treat larger numbers of uninsured patients with chronic unmanaged illnesses that require intervention. The reward system for physicians to take call and fulfill their obligation to the hospital no longer matches the responsibility.”

To illustrate the change, Dr. Whitcomb offers an example: “We have some days of the month where the call roster for general surgeries has vacancies. A month ago we had to send a patient to another hospital for an appendectomy.”

It is not an isolated instance. “Every hospital is struggling with the fact that many physicians don’t view unassigned call as a part of membership on the staff; they want to be paid for it,” says SHM President Patrick Cawley, MD, executive medical director of the Medical University of South Carolina (MUSC). And extra “pay” for services that used to be rendered gratis is one thing today’s strapped hospitals can little afford.

Committee staffing is another area undergoing change. Attending physicians are simply declining the duty. Neal Axon, MD, a hospitalist and assistant professor of medicine and pediatrics at MUSC, has seen the transformation firsthand. At one hospital his service covered, he saw the following: “At the first staff meeting there were 50 people; there was food, liquor. It was social and attendance was mandatory. You had to make three or four meetings a year to be on medical staff at this hospital.” But then, he says, attendance waned, and in the last year “dropped off precipitously.”

The old ways don’t work so what will replace them? “The point is that both physicians and hospitals need to put something on the table to collaborate,” Dr. Cawley says. “Many are saying that the hospital-physician relationship needs to change, but everyone is still feeling their way through it. What does it mean?”

SHM’s CEO Larry Wellikson, MD, sees a layered structure ahead. “Clearly the system is evolving into three kinds of physicians who use the hospital,” he says. “We are not advocating for it—just saying what it is. This is what is evolving, and hospital staffs need to see this is coming.” His three kinds of physicians are categorized by their relationship to the hospital.

The home team: “The first group is those physicians who work only at the hospital,” he says. “Their professional life is with the hospital as an institution: hospitalists, ER doctors, critical care physicians, and sometimes the anesthesiologists and radiologists. The hospital is the location of their work and provides the tools to do their job. If the hospital works well, they can do their job well. If hospital is dysfunctional, they can’t work well.”

He describes their relationship to the hospital with an anecdote: “When I was regular physician who came to the hospital just to see my patient, if they couldn’t find the chart I would scream and yell about that one patient.” Every physician faced with a missing chart thinks of it as an individual problem. “But now as a hospitalist, I try to fix the system, because all my patients are affected,” he says. “Hospitalists are on the inside trying to make it work.”

Important visitors: The second group, whom Dr. Wellikson calls important visitors, has a totally different relationship with the hospital. These are the cardiologists, orthopedic surgeons, and other medical specialties. “They are very important,” he says. “But they use the hospital intermittently and are not as tightly connected to it.”

Even so, they desire a high-quality hospital for their patients and will be willing to help set performance standards to achieve it. But their interest may not extend to patients who are not their own. “If the hospital says you have to also take care of free patients, they may choose not to,” Dr. Wellikson notes. “In fact, sometimes they have their own outpatient centers,” making them direct competitors as their competing practices sap revenue from money-making patients and procedures—all the while sending the sickest and costliest patients to the hospital.

Office-based physicians: The last of Dr. Wellikson’s groups is office-based physicians. These are the doctors who once made daily morning and evening rounds of their hospitalized patients but are now infrequently found at the bedside. “They are the physicians who don’t come to the hospital anymore: primary care physicians, endocrinologists, rheumatologists, neurologists, physicians who do all their surgery as outpatient procedures,” Dr. Wellikson explains.

This upheaval is due to tectonic shifts in both medical economics and lifestyle preferences. “Because the reimbursement for care has gone down, physicians have to see more patients to make the same amount of money,” explains Dr. Wellikson. Turning their hospitalized patients over to hospitalists allows office-based physicians to maximize their income and optimize their time.

“It increases satisfaction, limits the hours you spend in the hospital, and puts some boundaries on your work day,” Dr. Cawley says.

“Doctors want more a predictable lifestyle,” Dr. Whitcomb says. In fact, their absence is already a fait accompli in many community hospitals.

As Dr. Axon succinctly puts it: “The primary care doc has left the building.”

Dr. Cawley believes the new system is a relief to many office-based physicians. “Some do miss going to the hospital and seeing other physicians to network with them. Some miss taking care of their acute in-care patients. But I think most are relieved to not have to go to hospital. They say, ‘No, things are better this way.’”

With so many other physicians withdrawing from hospitals to their offices and clinics, Dr. Wellikson believes hospitalists will become increasingly crucial to the institution’s operation and governance. “Now the home team is going to be more active; how you staff, how you make the hospital more efficient,” he says. “The inside physicians will be much more interactive. That’s why hospital medicine has grown so rapidly.”

The explosive expansion of hospital medicine as a specialty is a direct result of the need to increase efficiency and quality standards in this new hospital atmosphere.

In addition, good home teams create a milieu in which other physicians—the important visitors (cardiologists, surgeons, orthopedists)—will want to work. “My job (as a hospitalist) is to create an environment where you can come in and do your surgery,” Dr. Wellikson points out.

The home team offers something else too: medical expertise. Providing post-operative care is not cost-effective for many surgeons. “The surgical specialists are not paid to manage medical issues,” Dr. Cawley says. “It takes time and if somebody else can manage it, that’s great.” That somebody is often a hospitalist. “There is a quality-control aspect as well,” he adds. “With hospitalists focusing on medical issues, the result is better patient care.”

Melding these groups of physicians with disparate interests and responsibilities is the next challenge for hospital leadership. It is a challenge fraught with potential pitfalls. As Dr. Wellikson explains, “The biggest obstacle is that physicians don’t do change very well.”

Administrators will turn to their institution’s hospitalists (both hospital-employed and contracted) to effect these changes and ensure overall standards and efficiency.

“I think hospitalists are in a position to bridge the gap between administrators and medical staff,” says David Yu, MD, medical director of hospitalist services at Decatur Memorial Hospital in Illinois. “I think that’s why there will be more and more hospitalists in leadership positions. That’s why hospitalists are unique: they have their feet in both worlds.”

Dr. Wellikson believes the home team will step up to the plate and take over many of the leadership duties of the new hospital.

Kenneth Patrick, MD, the ICU director of Chestnut Hill Hospital in Philadelphia, sounds a more cautionary note. Dr. Patrick, a trained hospitalist and intensivist, believes the demise of the old “hospital privilege” model is dissolving ties between physicians and their workplace. “I think younger physicians will be much more transient and more concerned with their position, work hours, and pay,” he says.

He sees a young workforce—whether hospital or office-based—as more disengaged than physicians used to be. “They will meet hospital standards, but not be actively involved in developing them,” he believes. That will be left to a small group of hospital-based physicians “who will voluntarily come forward because it is their civic responsibility. It would be nice if more physicians would work on committees, but they look at them like jury duty and they don’t want to serve.”

Dr. Patrick

“The question everyone asks is ‘What’s in it for me?’” Dr. Yu says. He notes a common sticking point: the requirement for increased documentation, which often means more work for doctors. “I think administrators are going to be in shock if they think practitioners are going to line up and say, ‘Well that’s great for the hospital.’”

The key to cooperation, says Dr. Yu, is the linking of changes to mutual benefit and patient welfare: “The administrators have to communicate that in the long run everyone will gain and it will ultimately lead to better patient care. You have to share your vision, inspire, motivate, and develop a culture of providing quality care. It’s easier said than done, but it’s the essence of medical care.”

What about patients? How do they react when a group of strangers takes over their hospital care rather than the primary care physician they often have gotten to know and trust for years? “Wanting your doctor present is counterbalanced by not having your doctor in the house,” Dr. Axon says. “Now you can see a physician anytime during the day.” And most patients are glad for the tradeoff. Dr. Yu has found the same dynamic with his patients at Decatur Memorial Hospital. “I can just count on one hand patients who were not happy the primary care physician wasn’t there,” he says. “Patients are more concerned with having their problems solved than with who is solving them.” And he makes sure his hospitalist staff never undermines the office based physicians. “We always say we are not better physicians, we are just more available.”

While they may have left the hospital, office-based physicians still will be a large presence in it by advocating for their patients. “If my whole currency is, ‘Do I have hospital privileges?’ then all my decisions are based on that,” Dr. Wellikson says.

Armed with the power of their patient referrals, office-based physicians will be able to demand that hospitals show proof of performance—thus becoming their patients’ ombudsmen. “I’m your shopper for the best healthcare, so the hospital has to step up to the plate and make sure it gets the business,” Dr. Wellikson explains. “They want standards because their patients need the best treatment, and they will have a choice of which hospital to put their patients into. If I now have a choice of three hospitals, I am looking to see that you are the Lexus of healthcare for my patients.”

Looking out for their patients’ interests is not the only way office-based physicians will continue to affect hospitals. As in-patient revenue declines, hospitals must look to the outpatient side to make up the difference. “The hospital is lucky if they break even on the inpatient side; they get the vast majority of money on the outpatient side: testing and procedures that private attendings are sending to the hospital,” Dr. Yu says.

He cautions against alienating those private practitioners by forcing change that is not mutually beneficial. “If you alienate them, you might lose money because they can send their patients to a different institution,” he warns. “These are the same doctors that never admit patients but do order the outpatient ultrasounds, blood tests, and therapies that are all money makers for the hospital. Why would you want to alienate these physicians?”

Dr. Patrick agrees: office-based physicians and hospitalists need each other. “I have to work with the primaries,” he says. “They are my source of referrals.”

There is another group that hospitals must learn to court, according to Dr. Axon: its own hospitalists. “I think you will see more innovative solutions to problems of recruiting hospital-based physicians to perform these functions,” he says. “For that to happen, the doctors will need to get more out of it. Many hospitalist groups are in a quandary; they are expected to do all these extra things, but pay is closely liked to clinical production and the number of patients they see. Those incentives will have to be aligned.”

All of which increases the reliance on—and importance of—those physicians who do work in the hospital—the home team. As Dr. Yu puts it: “I think the hospitalist model, whether you like or hate it, is the wave of the future.” TH

Carol Berczuk is a journalist based in New York.

It used to be so simple. The relationship between doctors and hospitals was a straightforward quid pro quo.

Hospitals granted privileges to physicians to admit and treat their patients, and the physicians returned the favor by assuming unpaid responsibilities like taking call, providing care to uninsured or emergency patients, and serving on administrative committees.

The hospital was like a friendly club whose members exchanged benefits for duties—a win-win situation. No more.

“You used to be part of a fraternity,” explains Win Whitcomb, MD, director of performance improvement at Mercy Medical Center in Springfield, Mass., and a co-founder of SHM. “There were social rewards. There was opportunity for collegial interchange.”

Economic pressure has taken that all away. “The pace of care has greatly intensified, and the financial reward system has deteriorated significantly,” Dr. Whitcomb continues. “We treat larger numbers of uninsured patients with chronic unmanaged illnesses that require intervention. The reward system for physicians to take call and fulfill their obligation to the hospital no longer matches the responsibility.”

To illustrate the change, Dr. Whitcomb offers an example: “We have some days of the month where the call roster for general surgeries has vacancies. A month ago we had to send a patient to another hospital for an appendectomy.”

It is not an isolated instance. “Every hospital is struggling with the fact that many physicians don’t view unassigned call as a part of membership on the staff; they want to be paid for it,” says SHM President Patrick Cawley, MD, executive medical director of the Medical University of South Carolina (MUSC). And extra “pay” for services that used to be rendered gratis is one thing today’s strapped hospitals can little afford.

Committee staffing is another area undergoing change. Attending physicians are simply declining the duty. Neal Axon, MD, a hospitalist and assistant professor of medicine and pediatrics at MUSC, has seen the transformation firsthand. At one hospital his service covered, he saw the following: “At the first staff meeting there were 50 people; there was food, liquor. It was social and attendance was mandatory. You had to make three or four meetings a year to be on medical staff at this hospital.” But then, he says, attendance waned, and in the last year “dropped off precipitously.”

The old ways don’t work so what will replace them? “The point is that both physicians and hospitals need to put something on the table to collaborate,” Dr. Cawley says. “Many are saying that the hospital-physician relationship needs to change, but everyone is still feeling their way through it. What does it mean?”

SHM’s CEO Larry Wellikson, MD, sees a layered structure ahead. “Clearly the system is evolving into three kinds of physicians who use the hospital,” he says. “We are not advocating for it—just saying what it is. This is what is evolving, and hospital staffs need to see this is coming.” His three kinds of physicians are categorized by their relationship to the hospital.

The home team: “The first group is those physicians who work only at the hospital,” he says. “Their professional life is with the hospital as an institution: hospitalists, ER doctors, critical care physicians, and sometimes the anesthesiologists and radiologists. The hospital is the location of their work and provides the tools to do their job. If the hospital works well, they can do their job well. If hospital is dysfunctional, they can’t work well.”

He describes their relationship to the hospital with an anecdote: “When I was regular physician who came to the hospital just to see my patient, if they couldn’t find the chart I would scream and yell about that one patient.” Every physician faced with a missing chart thinks of it as an individual problem. “But now as a hospitalist, I try to fix the system, because all my patients are affected,” he says. “Hospitalists are on the inside trying to make it work.”

Important visitors: The second group, whom Dr. Wellikson calls important visitors, has a totally different relationship with the hospital. These are the cardiologists, orthopedic surgeons, and other medical specialties. “They are very important,” he says. “But they use the hospital intermittently and are not as tightly connected to it.”

Even so, they desire a high-quality hospital for their patients and will be willing to help set performance standards to achieve it. But their interest may not extend to patients who are not their own. “If the hospital says you have to also take care of free patients, they may choose not to,” Dr. Wellikson notes. “In fact, sometimes they have their own outpatient centers,” making them direct competitors as their competing practices sap revenue from money-making patients and procedures—all the while sending the sickest and costliest patients to the hospital.

Office-based physicians: The last of Dr. Wellikson’s groups is office-based physicians. These are the doctors who once made daily morning and evening rounds of their hospitalized patients but are now infrequently found at the bedside. “They are the physicians who don’t come to the hospital anymore: primary care physicians, endocrinologists, rheumatologists, neurologists, physicians who do all their surgery as outpatient procedures,” Dr. Wellikson explains.

This upheaval is due to tectonic shifts in both medical economics and lifestyle preferences. “Because the reimbursement for care has gone down, physicians have to see more patients to make the same amount of money,” explains Dr. Wellikson. Turning their hospitalized patients over to hospitalists allows office-based physicians to maximize their income and optimize their time.

“It increases satisfaction, limits the hours you spend in the hospital, and puts some boundaries on your work day,” Dr. Cawley says.

“Doctors want more a predictable lifestyle,” Dr. Whitcomb says. In fact, their absence is already a fait accompli in many community hospitals.

As Dr. Axon succinctly puts it: “The primary care doc has left the building.”

Dr. Cawley believes the new system is a relief to many office-based physicians. “Some do miss going to the hospital and seeing other physicians to network with them. Some miss taking care of their acute in-care patients. But I think most are relieved to not have to go to hospital. They say, ‘No, things are better this way.’”

With so many other physicians withdrawing from hospitals to their offices and clinics, Dr. Wellikson believes hospitalists will become increasingly crucial to the institution’s operation and governance. “Now the home team is going to be more active; how you staff, how you make the hospital more efficient,” he says. “The inside physicians will be much more interactive. That’s why hospital medicine has grown so rapidly.”

The explosive expansion of hospital medicine as a specialty is a direct result of the need to increase efficiency and quality standards in this new hospital atmosphere.

In addition, good home teams create a milieu in which other physicians—the important visitors (cardiologists, surgeons, orthopedists)—will want to work. “My job (as a hospitalist) is to create an environment where you can come in and do your surgery,” Dr. Wellikson points out.

The home team offers something else too: medical expertise. Providing post-operative care is not cost-effective for many surgeons. “The surgical specialists are not paid to manage medical issues,” Dr. Cawley says. “It takes time and if somebody else can manage it, that’s great.” That somebody is often a hospitalist. “There is a quality-control aspect as well,” he adds. “With hospitalists focusing on medical issues, the result is better patient care.”

Melding these groups of physicians with disparate interests and responsibilities is the next challenge for hospital leadership. It is a challenge fraught with potential pitfalls. As Dr. Wellikson explains, “The biggest obstacle is that physicians don’t do change very well.”

Administrators will turn to their institution’s hospitalists (both hospital-employed and contracted) to effect these changes and ensure overall standards and efficiency.

“I think hospitalists are in a position to bridge the gap between administrators and medical staff,” says David Yu, MD, medical director of hospitalist services at Decatur Memorial Hospital in Illinois. “I think that’s why there will be more and more hospitalists in leadership positions. That’s why hospitalists are unique: they have their feet in both worlds.”

Dr. Wellikson believes the home team will step up to the plate and take over many of the leadership duties of the new hospital.

Kenneth Patrick, MD, the ICU director of Chestnut Hill Hospital in Philadelphia, sounds a more cautionary note. Dr. Patrick, a trained hospitalist and intensivist, believes the demise of the old “hospital privilege” model is dissolving ties between physicians and their workplace. “I think younger physicians will be much more transient and more concerned with their position, work hours, and pay,” he says.

He sees a young workforce—whether hospital or office-based—as more disengaged than physicians used to be. “They will meet hospital standards, but not be actively involved in developing them,” he believes. That will be left to a small group of hospital-based physicians “who will voluntarily come forward because it is their civic responsibility. It would be nice if more physicians would work on committees, but they look at them like jury duty and they don’t want to serve.”

Dr. Patrick

“The question everyone asks is ‘What’s in it for me?’” Dr. Yu says. He notes a common sticking point: the requirement for increased documentation, which often means more work for doctors. “I think administrators are going to be in shock if they think practitioners are going to line up and say, ‘Well that’s great for the hospital.’”

The key to cooperation, says Dr. Yu, is the linking of changes to mutual benefit and patient welfare: “The administrators have to communicate that in the long run everyone will gain and it will ultimately lead to better patient care. You have to share your vision, inspire, motivate, and develop a culture of providing quality care. It’s easier said than done, but it’s the essence of medical care.”

What about patients? How do they react when a group of strangers takes over their hospital care rather than the primary care physician they often have gotten to know and trust for years? “Wanting your doctor present is counterbalanced by not having your doctor in the house,” Dr. Axon says. “Now you can see a physician anytime during the day.” And most patients are glad for the tradeoff. Dr. Yu has found the same dynamic with his patients at Decatur Memorial Hospital. “I can just count on one hand patients who were not happy the primary care physician wasn’t there,” he says. “Patients are more concerned with having their problems solved than with who is solving them.” And he makes sure his hospitalist staff never undermines the office based physicians. “We always say we are not better physicians, we are just more available.”

While they may have left the hospital, office-based physicians still will be a large presence in it by advocating for their patients. “If my whole currency is, ‘Do I have hospital privileges?’ then all my decisions are based on that,” Dr. Wellikson says.

Armed with the power of their patient referrals, office-based physicians will be able to demand that hospitals show proof of performance—thus becoming their patients’ ombudsmen. “I’m your shopper for the best healthcare, so the hospital has to step up to the plate and make sure it gets the business,” Dr. Wellikson explains. “They want standards because their patients need the best treatment, and they will have a choice of which hospital to put their patients into. If I now have a choice of three hospitals, I am looking to see that you are the Lexus of healthcare for my patients.”

Looking out for their patients’ interests is not the only way office-based physicians will continue to affect hospitals. As in-patient revenue declines, hospitals must look to the outpatient side to make up the difference. “The hospital is lucky if they break even on the inpatient side; they get the vast majority of money on the outpatient side: testing and procedures that private attendings are sending to the hospital,” Dr. Yu says.

He cautions against alienating those private practitioners by forcing change that is not mutually beneficial. “If you alienate them, you might lose money because they can send their patients to a different institution,” he warns. “These are the same doctors that never admit patients but do order the outpatient ultrasounds, blood tests, and therapies that are all money makers for the hospital. Why would you want to alienate these physicians?”

Dr. Patrick agrees: office-based physicians and hospitalists need each other. “I have to work with the primaries,” he says. “They are my source of referrals.”

There is another group that hospitals must learn to court, according to Dr. Axon: its own hospitalists. “I think you will see more innovative solutions to problems of recruiting hospital-based physicians to perform these functions,” he says. “For that to happen, the doctors will need to get more out of it. Many hospitalist groups are in a quandary; they are expected to do all these extra things, but pay is closely liked to clinical production and the number of patients they see. Those incentives will have to be aligned.”

All of which increases the reliance on—and importance of—those physicians who do work in the hospital—the home team. As Dr. Yu puts it: “I think the hospitalist model, whether you like or hate it, is the wave of the future.” TH

Carol Berczuk is a journalist based in New York.

Hospitalist Nhi Lan Pham, MD, accepted a signing/starting bonus to relocate to Texas after finishing her residency in internal medicine in the Detroit area in 2007. The accompanying relocation expenses helped Dr. Pham begin her career near her family in Austin, and the flexible work schedule she negotiated allowed her to spend time with her family.

Another hospitalist willing to relocate for the right job used the relocation budget to his advantage. Moving to take the right job cost $2,000. The employer had budgeted $5,000 for relocation expenses, and the physician was able to arrange to have the $3,000 difference added to his signing bonus.

Yet another physician, already established as a hospitalist in an underserved area of his state, was attracted to a new position. However, his original acceptance of a loan repayment from the state as an inducement to work in the underserved region precluded his applying for the new position. His arrangement with the state did not prevent his approaching the new hospital to see what might be possible. It was a wise move; the new hospital agreed to increase the hospitalist’s signing bonus by enough to reimburse the state for the loan repayment.

One foreign-born physician secured a commitment from his recruiter that his employer would sponsor him and his family for green cards. Another candidate agreed to a reduced starting salary in return for help in securing a visa.

With a projected need for 30,000 hospitalists by 2010, hospitalists find themselves in the driver’s seat when it comes to weighing offers. Incentives are increasingly enticing as hospitalist recruiters nationwide struggle to lure top talent.

Incentives, Perks

What does this mean in practical terms? It means not only rising salaries but also incentives and extra perks to attract candidates to this fast-growing specialty.

Financial benefits are widespread, including signing and performance bonuses. Many hospitalists can plan on a guaranteed income. Employers may agree to pay off student loans or reimburse tuition. Malpractice insurance and tail coverage are commonly covered. Some employers also allow part-time or temporary employment to give a new hospitalist an opportunity to decide about the future or to accommodate a personal schedule.

“There is often a laundry list of incentives from which to choose, as well as more of a cafeteria plan that a doctor and employer can customize to meet specific needs,” according to Mark Dotson, MD, senior director of recruitment at Brentwood, Tenn.-based Cogent Healthcare. Cogent is a recruiting firm dedicated to building and managing hospitalist programs.

By far, the most appealing incentives are flexibility of scheduling and workload that allow physicians to coordinate their work schedule with their lifestyle. In fact, Dan Polk, MD, chief of the division of hospital-based medicine at Children’s Memorial Hospital in Chicago, considers flexible scheduling the basis of his plan to retain staff and build job satisfaction.

“We support lifestyle choices and respect life situations,” Dr. Polk says. “We foster the idea of joining a great team, and we make the environment attractive enough to encourage people to stay. We try to work within the team to cover those who need help, such as maternity or family leave, and we compensate for extra time at a different rate. We embrace people who want to work part time or share a job. Our goals are to support people and to make them want to get up in the morning to come to work.”

Part of the attractiveness of schedule flexibility is fewer weekend and night hours. In addition, employers may allow hospitalists to limit their caseload. Some hospitalists, for example, request a cap of 15 to 18 patients a day.

While retaining experienced, motivated staff is a goal of hospitals, lower caseloads mean “more doctors to do the work if doctors work fewer hours,” says Rusty Holman, MD, Cogent’s chief operating officer and SHM’s immediate past president. To meet that need, hospitals are turning to community-based physicians, fellows, and residents to work weekends and evenings. This, in turn, offers the perk of part-time work for those who want more personal time in their schedule.

The demand for nocturnists also is growing (The Hospitalist, January 2008, p. 22). Nocturnists work at night and on weekends and usually work shorter hours. These physicians prefer this schedule so they can have their days free for family or other pursuits. They also enjoy higher compensation, fewer workdays per month, and lower productivity expectations.

In addition to the having the options of part-time hours, temporary work, or job sharing, hospitalists also can negotiate other schedule perks. Some request and receive a two-week-on, two-week-off schedule. Many ask about the shift model, which demands nothing beyond the full eight or 12 hours of work. Still other applicants find a swing shift fits their lifestyle. There are even short-term choices: the hospitalist program at the University of California at Irvine offers recent residents the opportunity to work for one year while deciding about their career. With scheduling choices as part of an incentive package, many hospitalists achieve Dr. Polk’s goal of being eager to come to work each morning.

Physicians are not the only beneficiaries of these perks. Cogent, for example, recruits physician assistants and nurse practitioners when forming hospitalist groups. These employees also enjoy incentives, including tuition for continuing education and the same schedule flexibility as hospitalists.

Seller’s Market

Hospital medicine faces a shortage of qualified applicants. The need for hospitalists far surpasses the supply of physicians, who are in the enviable position of sifting through incentives and perks when selecting a hospitalist job.

This has become a national concern, according to Vikas Parekh, MD, assistant director of the hospitalist program and assistant professor of medicine at the University of Michigan Health Center in Ann Arbor. “We’re not seeing a pool of applicants because top residents are not pursuing hospitalist careers,” Dr. Parekh says.

Alpesh Amin, MD, MBA, a member of SHM’s Board of Directors, concurs but also points out that the number of hospitalist jobs is growing. “The need for a few hundred hospitalists 10 years ago has grown to 20,000 to 30,000 today, thus creating a need much greater than the supply,” says Dr. Amin, professor and chief, general internal medicine, executive director and founder of the hospitalist program at the University of California Irvine.

Why the shortage? First, fewer physicians are choosing to practice general medicine, either as an internist, family practitioner, or hospitalist. A recent study found fewer medical students were planning to concentrate on internal and family medicine, and that those who did planned to go into a subspecialty later.1 Dr. Parekh attributes this to a combination of reasons. “Most internal medicine residents are subspecialty oriented and may have decided their specialty early on,” Dr. Parekh says. “They may choose a subspecialty for financial reasons or prestige,” he continues, “but they may also be unclear about what a hospitalist career really is.”

Second, hospitalist programs have begun to expand from large metropolitan regions to smaller and rural areas. The result is an even greater demand for hospitalists.

Meet the Need

“There are no saturated markets within hospital medicine,” Dr. Holman says. “That is, most groups are always actively recruiting. [Cogent develops] full hospitalist programs, including recruiting, employing, managing, and training for new and existing hospitalist groups.”

Who is being recruited? Many recruiters approach residents who have not chosen a subspecialty to offer a staff position after they finish the residency. Although a recruiting firm may not offer financial aid during the residency, an employer may provide some sort of stipend if the candidate commits to remain on staff for a specified time after residency. “Recruit and retain” is the operative phrase in these cases.

Recruiters also are approaching generalists just entering the market to point out the advantages of avoiding the startup costs of establishing an outpatient practice. Further, many hospitalists are emerging from the ranks of solo practitioners interested in the financial and personal advantages of belonging to hospitalist groups. Not only does that eliminate the practice overhead (including the burden of regulatory compliance), but it also may offer additional administrative and academic opportunities. As Dr. Amin says, “There are more MD-MBA combos out there.”

Are incentives the answer to the shortage? Perhaps for now. With time, hospital medicine’s built-in perks may end the shortage and the need for added incentives. TH

Ann Kepler is a medical writer based in Chicago.

Reference

1. Croasdale M. Primary care doctors in demand; signing bonuses and higher pay for some. American Medical News. June 19, 2006. Available at www.ama-assn.org/amednews/site/free/prl10619.htm. Last accessed March 19, 2008.

Hospitalist Nhi Lan Pham, MD, accepted a signing/starting bonus to relocate to Texas after finishing her residency in internal medicine in the Detroit area in 2007. The accompanying relocation expenses helped Dr. Pham begin her career near her family in Austin, and the flexible work schedule she negotiated allowed her to spend time with her family.

Another hospitalist willing to relocate for the right job used the relocation budget to his advantage. Moving to take the right job cost $2,000. The employer had budgeted $5,000 for relocation expenses, and the physician was able to arrange to have the $3,000 difference added to his signing bonus.

Yet another physician, already established as a hospitalist in an underserved area of his state, was attracted to a new position. However, his original acceptance of a loan repayment from the state as an inducement to work in the underserved region precluded his applying for the new position. His arrangement with the state did not prevent his approaching the new hospital to see what might be possible. It was a wise move; the new hospital agreed to increase the hospitalist’s signing bonus by enough to reimburse the state for the loan repayment.

One foreign-born physician secured a commitment from his recruiter that his employer would sponsor him and his family for green cards. Another candidate agreed to a reduced starting salary in return for help in securing a visa.

With a projected need for 30,000 hospitalists by 2010, hospitalists find themselves in the driver’s seat when it comes to weighing offers. Incentives are increasingly enticing as hospitalist recruiters nationwide struggle to lure top talent.

Incentives, Perks

What does this mean in practical terms? It means not only rising salaries but also incentives and extra perks to attract candidates to this fast-growing specialty.

Financial benefits are widespread, including signing and performance bonuses. Many hospitalists can plan on a guaranteed income. Employers may agree to pay off student loans or reimburse tuition. Malpractice insurance and tail coverage are commonly covered. Some employers also allow part-time or temporary employment to give a new hospitalist an opportunity to decide about the future or to accommodate a personal schedule.

“There is often a laundry list of incentives from which to choose, as well as more of a cafeteria plan that a doctor and employer can customize to meet specific needs,” according to Mark Dotson, MD, senior director of recruitment at Brentwood, Tenn.-based Cogent Healthcare. Cogent is a recruiting firm dedicated to building and managing hospitalist programs.

By far, the most appealing incentives are flexibility of scheduling and workload that allow physicians to coordinate their work schedule with their lifestyle. In fact, Dan Polk, MD, chief of the division of hospital-based medicine at Children’s Memorial Hospital in Chicago, considers flexible scheduling the basis of his plan to retain staff and build job satisfaction.

“We support lifestyle choices and respect life situations,” Dr. Polk says. “We foster the idea of joining a great team, and we make the environment attractive enough to encourage people to stay. We try to work within the team to cover those who need help, such as maternity or family leave, and we compensate for extra time at a different rate. We embrace people who want to work part time or share a job. Our goals are to support people and to make them want to get up in the morning to come to work.”

Part of the attractiveness of schedule flexibility is fewer weekend and night hours. In addition, employers may allow hospitalists to limit their caseload. Some hospitalists, for example, request a cap of 15 to 18 patients a day.

While retaining experienced, motivated staff is a goal of hospitals, lower caseloads mean “more doctors to do the work if doctors work fewer hours,” says Rusty Holman, MD, Cogent’s chief operating officer and SHM’s immediate past president. To meet that need, hospitals are turning to community-based physicians, fellows, and residents to work weekends and evenings. This, in turn, offers the perk of part-time work for those who want more personal time in their schedule.

The demand for nocturnists also is growing (The Hospitalist, January 2008, p. 22). Nocturnists work at night and on weekends and usually work shorter hours. These physicians prefer this schedule so they can have their days free for family or other pursuits. They also enjoy higher compensation, fewer workdays per month, and lower productivity expectations.

In addition to the having the options of part-time hours, temporary work, or job sharing, hospitalists also can negotiate other schedule perks. Some request and receive a two-week-on, two-week-off schedule. Many ask about the shift model, which demands nothing beyond the full eight or 12 hours of work. Still other applicants find a swing shift fits their lifestyle. There are even short-term choices: the hospitalist program at the University of California at Irvine offers recent residents the opportunity to work for one year while deciding about their career. With scheduling choices as part of an incentive package, many hospitalists achieve Dr. Polk’s goal of being eager to come to work each morning.

Physicians are not the only beneficiaries of these perks. Cogent, for example, recruits physician assistants and nurse practitioners when forming hospitalist groups. These employees also enjoy incentives, including tuition for continuing education and the same schedule flexibility as hospitalists.

Seller’s Market

Hospital medicine faces a shortage of qualified applicants. The need for hospitalists far surpasses the supply of physicians, who are in the enviable position of sifting through incentives and perks when selecting a hospitalist job.

This has become a national concern, according to Vikas Parekh, MD, assistant director of the hospitalist program and assistant professor of medicine at the University of Michigan Health Center in Ann Arbor. “We’re not seeing a pool of applicants because top residents are not pursuing hospitalist careers,” Dr. Parekh says.

Alpesh Amin, MD, MBA, a member of SHM’s Board of Directors, concurs but also points out that the number of hospitalist jobs is growing. “The need for a few hundred hospitalists 10 years ago has grown to 20,000 to 30,000 today, thus creating a need much greater than the supply,” says Dr. Amin, professor and chief, general internal medicine, executive director and founder of the hospitalist program at the University of California Irvine.

Why the shortage? First, fewer physicians are choosing to practice general medicine, either as an internist, family practitioner, or hospitalist. A recent study found fewer medical students were planning to concentrate on internal and family medicine, and that those who did planned to go into a subspecialty later.1 Dr. Parekh attributes this to a combination of reasons. “Most internal medicine residents are subspecialty oriented and may have decided their specialty early on,” Dr. Parekh says. “They may choose a subspecialty for financial reasons or prestige,” he continues, “but they may also be unclear about what a hospitalist career really is.”

Second, hospitalist programs have begun to expand from large metropolitan regions to smaller and rural areas. The result is an even greater demand for hospitalists.

Meet the Need

“There are no saturated markets within hospital medicine,” Dr. Holman says. “That is, most groups are always actively recruiting. [Cogent develops] full hospitalist programs, including recruiting, employing, managing, and training for new and existing hospitalist groups.”

Who is being recruited? Many recruiters approach residents who have not chosen a subspecialty to offer a staff position after they finish the residency. Although a recruiting firm may not offer financial aid during the residency, an employer may provide some sort of stipend if the candidate commits to remain on staff for a specified time after residency. “Recruit and retain” is the operative phrase in these cases.

Recruiters also are approaching generalists just entering the market to point out the advantages of avoiding the startup costs of establishing an outpatient practice. Further, many hospitalists are emerging from the ranks of solo practitioners interested in the financial and personal advantages of belonging to hospitalist groups. Not only does that eliminate the practice overhead (including the burden of regulatory compliance), but it also may offer additional administrative and academic opportunities. As Dr. Amin says, “There are more MD-MBA combos out there.”

Are incentives the answer to the shortage? Perhaps for now. With time, hospital medicine’s built-in perks may end the shortage and the need for added incentives. TH

Ann Kepler is a medical writer based in Chicago.

Reference

1. Croasdale M. Primary care doctors in demand; signing bonuses and higher pay for some. American Medical News. June 19, 2006. Available at www.ama-assn.org/amednews/site/free/prl10619.htm. Last accessed March 19, 2008.

Hospitalist Nhi Lan Pham, MD, accepted a signing/starting bonus to relocate to Texas after finishing her residency in internal medicine in the Detroit area in 2007. The accompanying relocation expenses helped Dr. Pham begin her career near her family in Austin, and the flexible work schedule she negotiated allowed her to spend time with her family.

Another hospitalist willing to relocate for the right job used the relocation budget to his advantage. Moving to take the right job cost $2,000. The employer had budgeted $5,000 for relocation expenses, and the physician was able to arrange to have the $3,000 difference added to his signing bonus.

Yet another physician, already established as a hospitalist in an underserved area of his state, was attracted to a new position. However, his original acceptance of a loan repayment from the state as an inducement to work in the underserved region precluded his applying for the new position. His arrangement with the state did not prevent his approaching the new hospital to see what might be possible. It was a wise move; the new hospital agreed to increase the hospitalist’s signing bonus by enough to reimburse the state for the loan repayment.

One foreign-born physician secured a commitment from his recruiter that his employer would sponsor him and his family for green cards. Another candidate agreed to a reduced starting salary in return for help in securing a visa.

With a projected need for 30,000 hospitalists by 2010, hospitalists find themselves in the driver’s seat when it comes to weighing offers. Incentives are increasingly enticing as hospitalist recruiters nationwide struggle to lure top talent.

Incentives, Perks

What does this mean in practical terms? It means not only rising salaries but also incentives and extra perks to attract candidates to this fast-growing specialty.

Financial benefits are widespread, including signing and performance bonuses. Many hospitalists can plan on a guaranteed income. Employers may agree to pay off student loans or reimburse tuition. Malpractice insurance and tail coverage are commonly covered. Some employers also allow part-time or temporary employment to give a new hospitalist an opportunity to decide about the future or to accommodate a personal schedule.

“There is often a laundry list of incentives from which to choose, as well as more of a cafeteria plan that a doctor and employer can customize to meet specific needs,” according to Mark Dotson, MD, senior director of recruitment at Brentwood, Tenn.-based Cogent Healthcare. Cogent is a recruiting firm dedicated to building and managing hospitalist programs.

By far, the most appealing incentives are flexibility of scheduling and workload that allow physicians to coordinate their work schedule with their lifestyle. In fact, Dan Polk, MD, chief of the division of hospital-based medicine at Children’s Memorial Hospital in Chicago, considers flexible scheduling the basis of his plan to retain staff and build job satisfaction.

“We support lifestyle choices and respect life situations,” Dr. Polk says. “We foster the idea of joining a great team, and we make the environment attractive enough to encourage people to stay. We try to work within the team to cover those who need help, such as maternity or family leave, and we compensate for extra time at a different rate. We embrace people who want to work part time or share a job. Our goals are to support people and to make them want to get up in the morning to come to work.”

Part of the attractiveness of schedule flexibility is fewer weekend and night hours. In addition, employers may allow hospitalists to limit their caseload. Some hospitalists, for example, request a cap of 15 to 18 patients a day.

While retaining experienced, motivated staff is a goal of hospitals, lower caseloads mean “more doctors to do the work if doctors work fewer hours,” says Rusty Holman, MD, Cogent’s chief operating officer and SHM’s immediate past president. To meet that need, hospitals are turning to community-based physicians, fellows, and residents to work weekends and evenings. This, in turn, offers the perk of part-time work for those who want more personal time in their schedule.

The demand for nocturnists also is growing (The Hospitalist, January 2008, p. 22). Nocturnists work at night and on weekends and usually work shorter hours. These physicians prefer this schedule so they can have their days free for family or other pursuits. They also enjoy higher compensation, fewer workdays per month, and lower productivity expectations.

In addition to the having the options of part-time hours, temporary work, or job sharing, hospitalists also can negotiate other schedule perks. Some request and receive a two-week-on, two-week-off schedule. Many ask about the shift model, which demands nothing beyond the full eight or 12 hours of work. Still other applicants find a swing shift fits their lifestyle. There are even short-term choices: the hospitalist program at the University of California at Irvine offers recent residents the opportunity to work for one year while deciding about their career. With scheduling choices as part of an incentive package, many hospitalists achieve Dr. Polk’s goal of being eager to come to work each morning.

Physicians are not the only beneficiaries of these perks. Cogent, for example, recruits physician assistants and nurse practitioners when forming hospitalist groups. These employees also enjoy incentives, including tuition for continuing education and the same schedule flexibility as hospitalists.

Seller’s Market

Hospital medicine faces a shortage of qualified applicants. The need for hospitalists far surpasses the supply of physicians, who are in the enviable position of sifting through incentives and perks when selecting a hospitalist job.

This has become a national concern, according to Vikas Parekh, MD, assistant director of the hospitalist program and assistant professor of medicine at the University of Michigan Health Center in Ann Arbor. “We’re not seeing a pool of applicants because top residents are not pursuing hospitalist careers,” Dr. Parekh says.

Alpesh Amin, MD, MBA, a member of SHM’s Board of Directors, concurs but also points out that the number of hospitalist jobs is growing. “The need for a few hundred hospitalists 10 years ago has grown to 20,000 to 30,000 today, thus creating a need much greater than the supply,” says Dr. Amin, professor and chief, general internal medicine, executive director and founder of the hospitalist program at the University of California Irvine.

Why the shortage? First, fewer physicians are choosing to practice general medicine, either as an internist, family practitioner, or hospitalist. A recent study found fewer medical students were planning to concentrate on internal and family medicine, and that those who did planned to go into a subspecialty later.1 Dr. Parekh attributes this to a combination of reasons. “Most internal medicine residents are subspecialty oriented and may have decided their specialty early on,” Dr. Parekh says. “They may choose a subspecialty for financial reasons or prestige,” he continues, “but they may also be unclear about what a hospitalist career really is.”

Second, hospitalist programs have begun to expand from large metropolitan regions to smaller and rural areas. The result is an even greater demand for hospitalists.

Meet the Need

“There are no saturated markets within hospital medicine,” Dr. Holman says. “That is, most groups are always actively recruiting. [Cogent develops] full hospitalist programs, including recruiting, employing, managing, and training for new and existing hospitalist groups.”

Who is being recruited? Many recruiters approach residents who have not chosen a subspecialty to offer a staff position after they finish the residency. Although a recruiting firm may not offer financial aid during the residency, an employer may provide some sort of stipend if the candidate commits to remain on staff for a specified time after residency. “Recruit and retain” is the operative phrase in these cases.

Recruiters also are approaching generalists just entering the market to point out the advantages of avoiding the startup costs of establishing an outpatient practice. Further, many hospitalists are emerging from the ranks of solo practitioners interested in the financial and personal advantages of belonging to hospitalist groups. Not only does that eliminate the practice overhead (including the burden of regulatory compliance), but it also may offer additional administrative and academic opportunities. As Dr. Amin says, “There are more MD-MBA combos out there.”

Are incentives the answer to the shortage? Perhaps for now. With time, hospital medicine’s built-in perks may end the shortage and the need for added incentives. TH

Ann Kepler is a medical writer based in Chicago.

Reference

1. Croasdale M. Primary care doctors in demand; signing bonuses and higher pay for some. American Medical News. June 19, 2006. Available at www.ama-assn.org/amednews/site/free/prl10619.htm. Last accessed March 19, 2008.

The pendulum of expert opinion is swinging away from routinely recommending beta-blockers to prevent cardiac events in non-cardiac surgery patients. We won’t be abandoning the perioperative use of beta-blockers altogether, but we will probably be using them more selectively than in the past.

The latest factor driving the trend is the online publication in May 2008 of the results of the Perioperative Ischemic Evaluation (POISE) trial,¹ the largest placebo-controlled trial of perioperative beta-blocker use to date. In brief, in a cohort of patients with atherosclerotic disease or at risk for it who were undergoing noncardiac surgery, fewer patients who received extended-release metoprolol succinate had a myocardial infarction, but more of them died or had a stroke compared with those receiving placebo. (Extended-release metoprolol succinate is available in the United States as Toprol-XL and generically.)

Not so long ago, the pendulum was going the other way. After two small trials in the 1990s concluded that beta-blockers reduced the risk of perioperative cardiac events in selected patients with known or suspected coronary disease,^2,3 their perioperative use was subsequently endorsed by the Leapfrog Group and the Agency for Healthcare Research and Quality. The National Quality Forum included perioperative beta-blockade in its “Safe Practices for Better Healthcare 2006 update,”^4,5 and the Physician Consortium for Performance Improvement and the Surgical Care Improvement Project both listed it as a quality measure.

Since then, this practice has been closely studied, especially as concomitant research has failed to demonstrate that pre-operative coronary revascularization improves outcomes, even in the presence of ischemic disease. But evidence has been accumulating that routine use of beta-blockers may not benefit as many patients as was hoped, and may actually cause harm. The 2007 joint American College of Cardiology (ACC) and American Heart Association (AHA) guidelines on perioperative cardiovascular evaluation and care for noncardiac surgery gives its strongest recommendation (class I: benefit clearly outweighs risk) for perioperative beta-blocker use only for patients at high risk: those with known ischemic heart disease undergoing vascular surgery and those who are already on these drugs before surgery.⁶

However, there are still gaps in our knowledge. Perhaps, with proper implementation, we may be able to use beta-blockers to improve outcomes in patients at intermediate risk as well. In this paper, we review the rationale and the evidence for and against perioperative use of beta-blockers and provide practical guidance for internists and hospitalists.

WHY CARDIAC EVENTS OCCUR AFTER SURGERY

Adverse cardiovascular events such as myocardial infarction and unstable angina are the leading causes of death after surgery.⁷ Such events occur in approximately 1% of patients older than 50 years undergoing elective inpatient surgery, but this number may be higher (approximately 5%) in those with known or suspected coronary disease.^8,9 Perioperative cardiac events can also be harbingers of further complications, dramatically increasing hospital length of stay.¹⁰

Some ischemic events are caused by physiologic derangements involving the balance between inflammatory mediators, sympathetic tone, and oxygen supply and demand that occur under the stress of surgery. Others are more “traditional” in etiology, involving acute plaque rupture, thrombosis, and occlusion. Studies have consistently found a correlation between perioperative ischemia and cardiac events (both in-hospital and long-term) and death.^11–17 Other studies suggest that most perioperative cardiac infarcts are non-Q-wave events,¹⁸ and most events occur within the first few days after surgery, particularly the first 48 hours, when the effects of anesthetics, pain, fluid shifts, and physiologic derangements are greatest.

Factors that may trigger acute occlusion in the perioperative period include abrupt changes in sympathetic tone, increased levels of cortisol and catecholamines, and tissue hypoxia. Other potential triggers activated or increased by the stress of surgery include coagulation factors such as alterations in platelet function; inflammatory factors such as tumor necrosis factor alpha, interleukin 1, interleukin 6, and C-reactive protein; and metabolism of free fatty acids (which contribute to increased oxygen demand as well as endothelial dysfunction).^9,19,20

A 1996 autopsy study found that 38 (90%) of 42 patients who died of a perioperative infarct had evidence of acute plaque rupture or plaque hemorrhage on coronary sectioning, findings corroborated in another, similar study.^21,22 These studies suggest that multiple causes contribute to perioperative myocardial infarction, and a single strategy may not suffice for prevention.

IF BETA-BLOCKERS PROTECT, HOW DO THEY DO IT?

Beta-blockers have several effects that should, in theory, protect against cardiac events during and after surgery.²³ They reduce cardiac oxygen demand by reducing the force of contraction and the heart rate, and they increase the duration of diastole, when the heart muscle is perfused. They are also antiarrhythmic, and they may limit free radical production, metalloproteinase activity, and myocardial plaque inflammation.²⁴

Some researchers have speculated that using beta-blockers long-term may alter intra-cellular signaling processes, for example decreasing the expression of receptors that receive signals for cell death, which in turn may affect the response to reperfusion cell injury and death. If this is true, there may be an advantage to starting beta-blockers well in advance of surgery.²⁵

EARLY CLINICAL EVIDENCE IN FAVOR OF PERIOPERATIVE BETA-BLOCKER USE

Evidence in patients at high risk

Mangano et al,² in a study published in 1996, randomized 200 patients with known coronary disease or established risk factors for it who were undergoing noncardiac surgery to receive the beta-blocker atenolol orally and intravenously or placebo in the immediate perioperative period. Fewer patients in the atenolol group died in the first 6 months after hospital discharge (0 vs 8%, P < .001), the first year (3% vs 14%, P = .005), and the first 2 years (10% vs 21%, P = .019). However, there was no difference in short-term outcomes, and the study excluded patients who died in the immediate postoperative period. If these patients had been included in the analysis, the difference in the death rate at 2 years would not have been statistically significant.²⁶ Other critical findings: more patients in the atenolol group were using angiotensin-converting enzyme inhibitors and beta-blockers when they were discharged, and the placebo group had slightly more patients with prior myocardial infarction or diabetes.²⁷ (Atenolol is available in the United Sates as Tenormin and generically.)

Poldermans et al,³ in a study published in 1999, randomized 112 vascular surgery patients to receive either oral bisoprolol or placebo. These patients were selected from a larger cohort of 1,351 patients on the basis of high-risk clinical features and abnormal results on dobutamine echocardiography. Bisoprolol was started at least 1 week before surgery (range 7–89 days, mean 37 days), and patients were reevaluated before surgery so that the dose could be titrated to a goal heart rate of less than 60 beats per minute. After surgery, the drug was continued for another 30 days. The study was stopped early because the bisoprolol group had a 90% lower rate of non-fatal myocardial infarction and cardiac death at 30 days. Despite the study’s limitation (eg, enrolling selected patients and using an unblinded protocol), these compelling findings made a strong case for the use of beta-blockers perioperatively in patients at high risk, ie, those with ischemic heart disease who are undergoing major vascular surgery. (Bisoprolol is available in the United States as Zebeta and generically)

Evidence in patients at intermediate risk

Boersma et al²⁸ performed a follow-up to the study by Poldermans et al, published in 2001, in which they analyzed characteristics of all 1,351 patients who had been originally considered for enrollment. Using regression analysis, they identified seven clinical risk factors that predicted adverse cardiac events: angina, prior myocardial infarction, congestive heart failure, prior stroke, diabetes, renal failure, and age 70 years or older. Furthermore, for the entire cohort, patients receiving beta-blockers had a lower risk of cardiac complications (0.8%) than those not receiving beta-blockers (2.3%). In particular, the patients at intermediate risk (defined as having one or two risk factors) had a very low event rate regardless of stress test results, provided they were on beta-blockers: their risk of death or myocardial infarction was 0.9%, compared with 3.0% for those not on beta-blockers.

The authors concluded that dobutamine stress testing may not be necessary in patients at intermediate risk if beta-blockers are appropriately prescribed. However, others took issue with their data and conclusions, arguing that there have been so few trials that the data are still inconclusive and inadequate to ascertain the benefit of perioperative beta-blockade, particularly in patients not at high risk.^26,29

The Revised Cardiac Risk Index. Although the Boersma risk-factor index is not used in general practice, numerous experts^27,20–32 recommend a similar one, the Revised Cardiac Risk Index, devised by Lee et al.⁸ This index consists of six risk factors, each of which is worth one point:

• Congestive heart failure, based on history or examination
• Myocardial infarction, symptomatic ischemic heart disease, or a positive stress test
• Renal insufficiency (ie, serum creatinine level > 2 mg/dL)
• History of stroke or transient ischemic attack
• Diabetes requiring insulin
• High-risk surgery (defined as intrathoracic, intra-abdominal, or suprainguinal vascular surgery).

Patients with three or more points are considered to be at high risk, and those with one or two points are considered to be at intermediate risk. The ACC/AHA 2007 guidelines⁶ use a modified version of this index that considers the issue of surgical risk separately from the other five clinical conditions.

Devereaux et al³³ performed a meta-analysis, published in 2005, of 22 studies of perioperative beta-blockade. They concluded that beta-blockers had no discernable benefit in any outcome measured, including deaths from any cause, deaths from cardiovascular causes, other cardiac events, hypotension, bradycardia, and bronchospasm. However, they based this conclusion on the use of a 99% confidence interval for each relative risk, which they believed was justified because the trials were small and the numbers of events were only moderate. When the outcomes are assessed using the more common 95% confidence interval, benefit was detected in the combined end point of cardiovascular death, nonfatal myocardial infarction, and nonfatal cardiac arrest.

Yang et al,³⁴ Brady et al,³⁵ and Juul et al³⁶ performed three subsequent randomized trials that added to the controversy. Most of the patients in these trials were at intermediate or low risk, and none of the trials found a significant benefit with perioperative beta-blocker use. However, the protocols in these studies were different from the one in the study by Poldermans et al,³ which had found perioperative beta-blockade to be beneficial. Whereas patients in that earlier study started taking a beta-blocker at least 1 week before surgery (and on average much earlier), had their dose aggressively titrated to a target heart rate, and continued taking it for 30 days afterward, the protocols in the later trials called for the drug to be started within 24 hours before surgery and continued for only a short time afterward.

Lindenauer et al,³⁷ in a retrospective study published in 2005, found that fewer surgical patients who received beta-blockers in the hospital died in the hospital. The researchers used an administrative database of more than 780,000 patients who underwent noncardiac surgery, and they used propensity-score matching to compare the postoperative mortality rates of patients who received beta-blockers and a matched group in the same large cohort who did not. Beta-blockers were associated with a lower morality rate in patients in whom the Revised Cardiac Risk Index score was 3 or greater. However, although there was a trend toward a lower rate with beta-blocker use in patients whose score was 2 (ie, at intermediate risk), the difference was not statistically significant, and patients with a score of 0 or 1 saw no benefit and were possibly harmed.

The authors admitted that their study had a number of limitations, including a retrospective design and the use of an administrative database for information regarding risk index conditions and comorbidities. In addition, because they assumed that any patient who received a beta-blocker on the first 2 hospital days was receiving appropriate perioperative treatment, they may have incorrectly estimated the number of patients who actually received these drugs as a risk-reduction strategy. For instance, some patients at low risk could have received beta-blockers for treatment of a specific event, which would be reflected as an increase in event rates for this group. They also had no data on what medications the patients received before they were hospitalized or whether the dose was titrated effectively. The study excluded all patients with congestive heart failure and chronic obstructive pulmonary disease, who may be candidates for beta-blockers in actual practice. In fact, a recent observational study in patients with severe left ventricular dysfunction suggested that these drugs substantially reduced the incidence of death in the short term and the long term.³⁸ Finally, half the surgeries were nonelective, which makes extrapolation of their risk profile by the Revised Cardiac Risk Index difficult, since Lee et al excluded patients undergoing emergency surgery from the cohorts from which they derived and validated their index criteria.

Nevertheless, the authors concluded that patients at intermediate risk derive no benefit from perioperative beta-blocker use, and that the odds ratio for death was actually higher in patients with no risk factors who received a beta-blocker.

DOES PERIOPERATIVE BETA-BLOCKER USE CAUSE HARM?

The published data on whether perioperative beta-blocker use harms patients are conflicting and up to now have been limited.

Stone et al³⁹ reported a substantial incidence of bradycardia requiring atropine in patients treated with a single dose of a beta-blocker preoperatively, but the complications were not clearly characterized.

The Perioperative Beta-Blockade trial.³⁵ Significantly more patients given short-acting metoprolol had intraoperative falls in blood pressure and heart rate, and more required inotropic support during surgery, although the treating anesthesiologists refused to be blinded in that study. (Short-acting metoprolol is available in the United States as Lopressor and generically.)

Devereaux et al,³³ in their meta-analysis, found a higher risk of bradycardia requiring treatment (but not a higher risk of hypotension) in beta-blocker users in nine studies, including the study by Stone et al and the Perioperative Beta-Blockade trial (relative risk 2.27, 95% confidence interval 1.36–3.80).

Conversely, at least three other studies found no difference in rates of intraoperative events.^36,40,41 There are few data on the incidence of other complications such as perioperative pulmonary edema and bronchospasm.

POISE: THE FIRST LARGE RANDOMIZED TRIAL

In May 2008, results were published from POISE, the first large randomized controlled trial of perioperative beta-blockade.¹ An impressive 8,351 patients—most of them at intermediate risk—were randomized to receive extended-release metoprolol succinate or placebo starting just before surgery and continuing for 30 days afterward.

Although the incidence of the primary composite end point (cardiovascular death, nonfatal myocardial infarction, and nonfatal cardiac arrest) was lower at 30 days in the metoprolol group than in the placebo group (5.8% vs 6.9%, hazard ratio 0.83, P = .04), other findings were worrisome: more metoprolol recipients died of any cause (3.1% vs 2.3%, P = .03) or had a stroke (1.0% vs 0.5%, P = .005). The major contributor to the higher mortality rate in this group appears to have been sepsis.

How beta-blockers might promote death by sepsis is unclear. The authors offered two possible explanations: perhaps beta-blocker-induced hypotension predisposes patients to infection and sepsis, or perhaps the slower heart rate and lower force of contraction induced by beta-blockers could mask normal responses to systemic infection, which in turn could delay recognition and treatment or impede the normal immune response. These mechanisms, like others, are speculative.

The risks of other adverse outcomes such as bradycardia and hypotension were substantially higher in the metoprolol group. The authors also pointed out that most of the patients who suffered nonfatal strokes were subsequently disabled or incapacitated, while most of those who suffered nonfatal cardiac events did not progress to further cardiac problems.

This new study has not yet been rigorously debated, but it will likely come under scrutiny for its dosing regimen (extended-release metoprolol succinate 100 mg or placebo 2–4 hours before surgery; another 100 mg or placebo 6 hours after surgery or sooner if the heart rate was 80 beats per minute or more and the systolic blood pressure 100 mm Hg or higher; and then 200 mg or placebo 12 hours after the second dose and every 24 hours thereafter for 30 days). This was fairly aggressive, especially for patients who have never received a beta-blocker before. In contrast, the protocol for the Perioperative Beta-Blockade trial called for only 25 to 50 mg of short-acting metoprolol twice a day. Another criticism is that the medication was started only a few hours before surgery, although there is no current standard practice for either the dose or when the treatment should be started. The population had a fairly high rate of cerebrovascular disease (perhaps predisposing to stroke whenever blood pressure dropped), and 10% of patients were undergoing urgent or emergency surgery, which carries a higher risk of morbidity.

ANY ROLE FOR BETA-BLOCKERS IN THOSE AT INTERMEDIATE RISK?

Thus, in the past decade, the appropriate perioperative use of beta-blockers, which, after the findings by Mangano et al and Poldermans et al, were seen as potentially beneficial for any patient at risk of coronary disease, with little suggestion of harm, has become more clearly defined, and the risks are more evident. The most compelling evidence in favor of using them comes from patients with ischemic heart disease undergoing vascular surgery; the 2007 ACC/AHA guidelines recommend that this group be offered beta-blockers in the absence of a contraindication (class I recommendation: benefit clearly outweighs risk).⁶ The guidelines also point out that these drugs should be continued in patients already taking them for cardiac indications before surgery, because ischemia may be precipitated if a beta-blocker is abruptly discontinued.^42,43

Additionally, the guidelines recommend considering beta-blockers for vascular surgery patients at high cardiac risk (with a Revised Cardiac Risk Index score of 3 or more), even if they are not known to have ischemic heart disease. This is a class IIa recommendation (the benefit outweighs the risk, but more studies are required).

The guidelines also recommend that beta-blockers be considered for patients who have a score of 0 if they are undergoing vascular surgery (class IIb recommendation) or a score of 1 if they are undergoing vascular surgery (class IIa recommendation) or intermediate-risk surgery (class IIb recommendation). However, in view of the POISE results, these recommendations need to be carefully scrutinized.

These data notwithstanding, beta-blockers still might be beneficial in perioperative patients at intermediate risk.

Start beta-blockers sooner?

To help patients at intermediate risk (such as those with diabetes without known heart disease), we may need to do what Poldermans et al did³: instead of seeing patients only once a day or two before surgery, we may need to do the preoperative assessment as much as a month before and, if necessary, start a beta-blocker at a low dose, titrate it to a goal heart rate, and follow the patient closely up until surgery and afterward.

The importance of heart-rate control was illustrated in a recent cohort study of perioperative beta-blockers in vascular surgery patients,⁴⁴ in which higher beta-blocker doses, carefully monitored, were associated with less ischemia and cardiac enzyme release. In addition, long-term mortality rates were lower in patients with lower heart rates. And Poldermans et al⁴⁵ recently performed a study in more than 700 intermediate-risk patients who were divided into two groups, one that underwent preoperative stress testing and one that did not. Beta-blockers were given to both groups, and doses were titrated to a goal heart rate of less than 65. The patients with optimally controlled heart rates had the lowest event rates.

However, the logistics of such a program would be challenging. For the most part, internists and hospitalists involved in perioperative assessment do not control the timing of referral or surgery, and adjustments cannot be made for patients whose preoperative clinic visit falls only a few days before surgery. Instituting a second or third visit to assess the efficacy of beta-blockade burdens the patient and may not be practical.

Are all beta-blockers equivalent?

An additional factor is the choice of agent. While the most significant studies of perioperative beta-blockade have used beta-1 receptor-selective agents (ie, metoprolol, atenolol, and bisoprolol), there is no prospective evidence that any particular agent is superior. However, a recent retrospective analysis of elderly surgical patients did suggest that longer-acting beta-blockers may be preferable: patients who had been on atenolol in the year before surgery had a 20% lower risk of postoperative myocardial infarction or death than those who had been on short-acting metoprolol, with no difference in noncardiac outcomes.⁴⁶

The pendulum of expert opinion is swinging away from routinely recommending beta-blockers to prevent cardiac events in non-cardiac surgery patients. We won’t be abandoning the perioperative use of beta-blockers altogether, but we will probably be using them more selectively than in the past.

The latest factor driving the trend is the online publication in May 2008 of the results of the Perioperative Ischemic Evaluation (POISE) trial,¹ the largest placebo-controlled trial of perioperative beta-blocker use to date. In brief, in a cohort of patients with atherosclerotic disease or at risk for it who were undergoing noncardiac surgery, fewer patients who received extended-release metoprolol succinate had a myocardial infarction, but more of them died or had a stroke compared with those receiving placebo. (Extended-release metoprolol succinate is available in the United States as Toprol-XL and generically.)

Not so long ago, the pendulum was going the other way. After two small trials in the 1990s concluded that beta-blockers reduced the risk of perioperative cardiac events in selected patients with known or suspected coronary disease,^2,3 their perioperative use was subsequently endorsed by the Leapfrog Group and the Agency for Healthcare Research and Quality. The National Quality Forum included perioperative beta-blockade in its “Safe Practices for Better Healthcare 2006 update,”^4,5 and the Physician Consortium for Performance Improvement and the Surgical Care Improvement Project both listed it as a quality measure.

Since then, this practice has been closely studied, especially as concomitant research has failed to demonstrate that pre-operative coronary revascularization improves outcomes, even in the presence of ischemic disease. But evidence has been accumulating that routine use of beta-blockers may not benefit as many patients as was hoped, and may actually cause harm. The 2007 joint American College of Cardiology (ACC) and American Heart Association (AHA) guidelines on perioperative cardiovascular evaluation and care for noncardiac surgery gives its strongest recommendation (class I: benefit clearly outweighs risk) for perioperative beta-blocker use only for patients at high risk: those with known ischemic heart disease undergoing vascular surgery and those who are already on these drugs before surgery.⁶

However, there are still gaps in our knowledge. Perhaps, with proper implementation, we may be able to use beta-blockers to improve outcomes in patients at intermediate risk as well. In this paper, we review the rationale and the evidence for and against perioperative use of beta-blockers and provide practical guidance for internists and hospitalists.

WHY CARDIAC EVENTS OCCUR AFTER SURGERY

Adverse cardiovascular events such as myocardial infarction and unstable angina are the leading causes of death after surgery.⁷ Such events occur in approximately 1% of patients older than 50 years undergoing elective inpatient surgery, but this number may be higher (approximately 5%) in those with known or suspected coronary disease.^8,9 Perioperative cardiac events can also be harbingers of further complications, dramatically increasing hospital length of stay.¹⁰

Some ischemic events are caused by physiologic derangements involving the balance between inflammatory mediators, sympathetic tone, and oxygen supply and demand that occur under the stress of surgery. Others are more “traditional” in etiology, involving acute plaque rupture, thrombosis, and occlusion. Studies have consistently found a correlation between perioperative ischemia and cardiac events (both in-hospital and long-term) and death.^11–17 Other studies suggest that most perioperative cardiac infarcts are non-Q-wave events,¹⁸ and most events occur within the first few days after surgery, particularly the first 48 hours, when the effects of anesthetics, pain, fluid shifts, and physiologic derangements are greatest.

Factors that may trigger acute occlusion in the perioperative period include abrupt changes in sympathetic tone, increased levels of cortisol and catecholamines, and tissue hypoxia. Other potential triggers activated or increased by the stress of surgery include coagulation factors such as alterations in platelet function; inflammatory factors such as tumor necrosis factor alpha, interleukin 1, interleukin 6, and C-reactive protein; and metabolism of free fatty acids (which contribute to increased oxygen demand as well as endothelial dysfunction).^9,19,20

A 1996 autopsy study found that 38 (90%) of 42 patients who died of a perioperative infarct had evidence of acute plaque rupture or plaque hemorrhage on coronary sectioning, findings corroborated in another, similar study.^21,22 These studies suggest that multiple causes contribute to perioperative myocardial infarction, and a single strategy may not suffice for prevention.

IF BETA-BLOCKERS PROTECT, HOW DO THEY DO IT?

Beta-blockers have several effects that should, in theory, protect against cardiac events during and after surgery.²³ They reduce cardiac oxygen demand by reducing the force of contraction and the heart rate, and they increase the duration of diastole, when the heart muscle is perfused. They are also antiarrhythmic, and they may limit free radical production, metalloproteinase activity, and myocardial plaque inflammation.²⁴

Some researchers have speculated that using beta-blockers long-term may alter intra-cellular signaling processes, for example decreasing the expression of receptors that receive signals for cell death, which in turn may affect the response to reperfusion cell injury and death. If this is true, there may be an advantage to starting beta-blockers well in advance of surgery.²⁵

EARLY CLINICAL EVIDENCE IN FAVOR OF PERIOPERATIVE BETA-BLOCKER USE

Evidence in patients at high risk

Mangano et al,² in a study published in 1996, randomized 200 patients with known coronary disease or established risk factors for it who were undergoing noncardiac surgery to receive the beta-blocker atenolol orally and intravenously or placebo in the immediate perioperative period. Fewer patients in the atenolol group died in the first 6 months after hospital discharge (0 vs 8%, P < .001), the first year (3% vs 14%, P = .005), and the first 2 years (10% vs 21%, P = .019). However, there was no difference in short-term outcomes, and the study excluded patients who died in the immediate postoperative period. If these patients had been included in the analysis, the difference in the death rate at 2 years would not have been statistically significant.²⁶ Other critical findings: more patients in the atenolol group were using angiotensin-converting enzyme inhibitors and beta-blockers when they were discharged, and the placebo group had slightly more patients with prior myocardial infarction or diabetes.²⁷ (Atenolol is available in the United Sates as Tenormin and generically.)

Poldermans et al,³ in a study published in 1999, randomized 112 vascular surgery patients to receive either oral bisoprolol or placebo. These patients were selected from a larger cohort of 1,351 patients on the basis of high-risk clinical features and abnormal results on dobutamine echocardiography. Bisoprolol was started at least 1 week before surgery (range 7–89 days, mean 37 days), and patients were reevaluated before surgery so that the dose could be titrated to a goal heart rate of less than 60 beats per minute. After surgery, the drug was continued for another 30 days. The study was stopped early because the bisoprolol group had a 90% lower rate of non-fatal myocardial infarction and cardiac death at 30 days. Despite the study’s limitation (eg, enrolling selected patients and using an unblinded protocol), these compelling findings made a strong case for the use of beta-blockers perioperatively in patients at high risk, ie, those with ischemic heart disease who are undergoing major vascular surgery. (Bisoprolol is available in the United States as Zebeta and generically)

Evidence in patients at intermediate risk

Boersma et al²⁸ performed a follow-up to the study by Poldermans et al, published in 2001, in which they analyzed characteristics of all 1,351 patients who had been originally considered for enrollment. Using regression analysis, they identified seven clinical risk factors that predicted adverse cardiac events: angina, prior myocardial infarction, congestive heart failure, prior stroke, diabetes, renal failure, and age 70 years or older. Furthermore, for the entire cohort, patients receiving beta-blockers had a lower risk of cardiac complications (0.8%) than those not receiving beta-blockers (2.3%). In particular, the patients at intermediate risk (defined as having one or two risk factors) had a very low event rate regardless of stress test results, provided they were on beta-blockers: their risk of death or myocardial infarction was 0.9%, compared with 3.0% for those not on beta-blockers.

The authors concluded that dobutamine stress testing may not be necessary in patients at intermediate risk if beta-blockers are appropriately prescribed. However, others took issue with their data and conclusions, arguing that there have been so few trials that the data are still inconclusive and inadequate to ascertain the benefit of perioperative beta-blockade, particularly in patients not at high risk.^26,29

The Revised Cardiac Risk Index. Although the Boersma risk-factor index is not used in general practice, numerous experts^27,20–32 recommend a similar one, the Revised Cardiac Risk Index, devised by Lee et al.⁸ This index consists of six risk factors, each of which is worth one point:

• Congestive heart failure, based on history or examination
• Myocardial infarction, symptomatic ischemic heart disease, or a positive stress test
• Renal insufficiency (ie, serum creatinine level > 2 mg/dL)
• History of stroke or transient ischemic attack
• Diabetes requiring insulin
• High-risk surgery (defined as intrathoracic, intra-abdominal, or suprainguinal vascular surgery).

Patients with three or more points are considered to be at high risk, and those with one or two points are considered to be at intermediate risk. The ACC/AHA 2007 guidelines⁶ use a modified version of this index that considers the issue of surgical risk separately from the other five clinical conditions.

Devereaux et al³³ performed a meta-analysis, published in 2005, of 22 studies of perioperative beta-blockade. They concluded that beta-blockers had no discernable benefit in any outcome measured, including deaths from any cause, deaths from cardiovascular causes, other cardiac events, hypotension, bradycardia, and bronchospasm. However, they based this conclusion on the use of a 99% confidence interval for each relative risk, which they believed was justified because the trials were small and the numbers of events were only moderate. When the outcomes are assessed using the more common 95% confidence interval, benefit was detected in the combined end point of cardiovascular death, nonfatal myocardial infarction, and nonfatal cardiac arrest.

Yang et al,³⁴ Brady et al,³⁵ and Juul et al³⁶ performed three subsequent randomized trials that added to the controversy. Most of the patients in these trials were at intermediate or low risk, and none of the trials found a significant benefit with perioperative beta-blocker use. However, the protocols in these studies were different from the one in the study by Poldermans et al,³ which had found perioperative beta-blockade to be beneficial. Whereas patients in that earlier study started taking a beta-blocker at least 1 week before surgery (and on average much earlier), had their dose aggressively titrated to a target heart rate, and continued taking it for 30 days afterward, the protocols in the later trials called for the drug to be started within 24 hours before surgery and continued for only a short time afterward.

Lindenauer et al,³⁷ in a retrospective study published in 2005, found that fewer surgical patients who received beta-blockers in the hospital died in the hospital. The researchers used an administrative database of more than 780,000 patients who underwent noncardiac surgery, and they used propensity-score matching to compare the postoperative mortality rates of patients who received beta-blockers and a matched group in the same large cohort who did not. Beta-blockers were associated with a lower morality rate in patients in whom the Revised Cardiac Risk Index score was 3 or greater. However, although there was a trend toward a lower rate with beta-blocker use in patients whose score was 2 (ie, at intermediate risk), the difference was not statistically significant, and patients with a score of 0 or 1 saw no benefit and were possibly harmed.

The authors admitted that their study had a number of limitations, including a retrospective design and the use of an administrative database for information regarding risk index conditions and comorbidities. In addition, because they assumed that any patient who received a beta-blocker on the first 2 hospital days was receiving appropriate perioperative treatment, they may have incorrectly estimated the number of patients who actually received these drugs as a risk-reduction strategy. For instance, some patients at low risk could have received beta-blockers for treatment of a specific event, which would be reflected as an increase in event rates for this group. They also had no data on what medications the patients received before they were hospitalized or whether the dose was titrated effectively. The study excluded all patients with congestive heart failure and chronic obstructive pulmonary disease, who may be candidates for beta-blockers in actual practice. In fact, a recent observational study in patients with severe left ventricular dysfunction suggested that these drugs substantially reduced the incidence of death in the short term and the long term.³⁸ Finally, half the surgeries were nonelective, which makes extrapolation of their risk profile by the Revised Cardiac Risk Index difficult, since Lee et al excluded patients undergoing emergency surgery from the cohorts from which they derived and validated their index criteria.

Nevertheless, the authors concluded that patients at intermediate risk derive no benefit from perioperative beta-blocker use, and that the odds ratio for death was actually higher in patients with no risk factors who received a beta-blocker.

DOES PERIOPERATIVE BETA-BLOCKER USE CAUSE HARM?

The published data on whether perioperative beta-blocker use harms patients are conflicting and up to now have been limited.

Stone et al³⁹ reported a substantial incidence of bradycardia requiring atropine in patients treated with a single dose of a beta-blocker preoperatively, but the complications were not clearly characterized.

The Perioperative Beta-Blockade trial.³⁵ Significantly more patients given short-acting metoprolol had intraoperative falls in blood pressure and heart rate, and more required inotropic support during surgery, although the treating anesthesiologists refused to be blinded in that study. (Short-acting metoprolol is available in the United States as Lopressor and generically.)

Devereaux et al,³³ in their meta-analysis, found a higher risk of bradycardia requiring treatment (but not a higher risk of hypotension) in beta-blocker users in nine studies, including the study by Stone et al and the Perioperative Beta-Blockade trial (relative risk 2.27, 95% confidence interval 1.36–3.80).

Conversely, at least three other studies found no difference in rates of intraoperative events.^36,40,41 There are few data on the incidence of other complications such as perioperative pulmonary edema and bronchospasm.

POISE: THE FIRST LARGE RANDOMIZED TRIAL

In May 2008, results were published from POISE, the first large randomized controlled trial of perioperative beta-blockade.¹ An impressive 8,351 patients—most of them at intermediate risk—were randomized to receive extended-release metoprolol succinate or placebo starting just before surgery and continuing for 30 days afterward.

Although the incidence of the primary composite end point (cardiovascular death, nonfatal myocardial infarction, and nonfatal cardiac arrest) was lower at 30 days in the metoprolol group than in the placebo group (5.8% vs 6.9%, hazard ratio 0.83, P = .04), other findings were worrisome: more metoprolol recipients died of any cause (3.1% vs 2.3%, P = .03) or had a stroke (1.0% vs 0.5%, P = .005). The major contributor to the higher mortality rate in this group appears to have been sepsis.

How beta-blockers might promote death by sepsis is unclear. The authors offered two possible explanations: perhaps beta-blocker-induced hypotension predisposes patients to infection and sepsis, or perhaps the slower heart rate and lower force of contraction induced by beta-blockers could mask normal responses to systemic infection, which in turn could delay recognition and treatment or impede the normal immune response. These mechanisms, like others, are speculative.

The risks of other adverse outcomes such as bradycardia and hypotension were substantially higher in the metoprolol group. The authors also pointed out that most of the patients who suffered nonfatal strokes were subsequently disabled or incapacitated, while most of those who suffered nonfatal cardiac events did not progress to further cardiac problems.

This new study has not yet been rigorously debated, but it will likely come under scrutiny for its dosing regimen (extended-release metoprolol succinate 100 mg or placebo 2–4 hours before surgery; another 100 mg or placebo 6 hours after surgery or sooner if the heart rate was 80 beats per minute or more and the systolic blood pressure 100 mm Hg or higher; and then 200 mg or placebo 12 hours after the second dose and every 24 hours thereafter for 30 days). This was fairly aggressive, especially for patients who have never received a beta-blocker before. In contrast, the protocol for the Perioperative Beta-Blockade trial called for only 25 to 50 mg of short-acting metoprolol twice a day. Another criticism is that the medication was started only a few hours before surgery, although there is no current standard practice for either the dose or when the treatment should be started. The population had a fairly high rate of cerebrovascular disease (perhaps predisposing to stroke whenever blood pressure dropped), and 10% of patients were undergoing urgent or emergency surgery, which carries a higher risk of morbidity.

ANY ROLE FOR BETA-BLOCKERS IN THOSE AT INTERMEDIATE RISK?

Thus, in the past decade, the appropriate perioperative use of beta-blockers, which, after the findings by Mangano et al and Poldermans et al, were seen as potentially beneficial for any patient at risk of coronary disease, with little suggestion of harm, has become more clearly defined, and the risks are more evident. The most compelling evidence in favor of using them comes from patients with ischemic heart disease undergoing vascular surgery; the 2007 ACC/AHA guidelines recommend that this group be offered beta-blockers in the absence of a contraindication (class I recommendation: benefit clearly outweighs risk).⁶ The guidelines also point out that these drugs should be continued in patients already taking them for cardiac indications before surgery, because ischemia may be precipitated if a beta-blocker is abruptly discontinued.^42,43

Additionally, the guidelines recommend considering beta-blockers for vascular surgery patients at high cardiac risk (with a Revised Cardiac Risk Index score of 3 or more), even if they are not known to have ischemic heart disease. This is a class IIa recommendation (the benefit outweighs the risk, but more studies are required).

The guidelines also recommend that beta-blockers be considered for patients who have a score of 0 if they are undergoing vascular surgery (class IIb recommendation) or a score of 1 if they are undergoing vascular surgery (class IIa recommendation) or intermediate-risk surgery (class IIb recommendation). However, in view of the POISE results, these recommendations need to be carefully scrutinized.

These data notwithstanding, beta-blockers still might be beneficial in perioperative patients at intermediate risk.

Start beta-blockers sooner?

To help patients at intermediate risk (such as those with diabetes without known heart disease), we may need to do what Poldermans et al did³: instead of seeing patients only once a day or two before surgery, we may need to do the preoperative assessment as much as a month before and, if necessary, start a beta-blocker at a low dose, titrate it to a goal heart rate, and follow the patient closely up until surgery and afterward.

The importance of heart-rate control was illustrated in a recent cohort study of perioperative beta-blockers in vascular surgery patients,⁴⁴ in which higher beta-blocker doses, carefully monitored, were associated with less ischemia and cardiac enzyme release. In addition, long-term mortality rates were lower in patients with lower heart rates. And Poldermans et al⁴⁵ recently performed a study in more than 700 intermediate-risk patients who were divided into two groups, one that underwent preoperative stress testing and one that did not. Beta-blockers were given to both groups, and doses were titrated to a goal heart rate of less than 65. The patients with optimally controlled heart rates had the lowest event rates.

However, the logistics of such a program would be challenging. For the most part, internists and hospitalists involved in perioperative assessment do not control the timing of referral or surgery, and adjustments cannot be made for patients whose preoperative clinic visit falls only a few days before surgery. Instituting a second or third visit to assess the efficacy of beta-blockade burdens the patient and may not be practical.

Are all beta-blockers equivalent?

An additional factor is the choice of agent. While the most significant studies of perioperative beta-blockade have used beta-1 receptor-selective agents (ie, metoprolol, atenolol, and bisoprolol), there is no prospective evidence that any particular agent is superior. However, a recent retrospective analysis of elderly surgical patients did suggest that longer-acting beta-blockers may be preferable: patients who had been on atenolol in the year before surgery had a 20% lower risk of postoperative myocardial infarction or death than those who had been on short-acting metoprolol, with no difference in noncardiac outcomes.⁴⁶

The pendulum of expert opinion is swinging away from routinely recommending beta-blockers to prevent cardiac events in non-cardiac surgery patients. We won’t be abandoning the perioperative use of beta-blockers altogether, but we will probably be using them more selectively than in the past.

The latest factor driving the trend is the online publication in May 2008 of the results of the Perioperative Ischemic Evaluation (POISE) trial,¹ the largest placebo-controlled trial of perioperative beta-blocker use to date. In brief, in a cohort of patients with atherosclerotic disease or at risk for it who were undergoing noncardiac surgery, fewer patients who received extended-release metoprolol succinate had a myocardial infarction, but more of them died or had a stroke compared with those receiving placebo. (Extended-release metoprolol succinate is available in the United States as Toprol-XL and generically.)

Not so long ago, the pendulum was going the other way. After two small trials in the 1990s concluded that beta-blockers reduced the risk of perioperative cardiac events in selected patients with known or suspected coronary disease,^2,3 their perioperative use was subsequently endorsed by the Leapfrog Group and the Agency for Healthcare Research and Quality. The National Quality Forum included perioperative beta-blockade in its “Safe Practices for Better Healthcare 2006 update,”^4,5 and the Physician Consortium for Performance Improvement and the Surgical Care Improvement Project both listed it as a quality measure.

Since then, this practice has been closely studied, especially as concomitant research has failed to demonstrate that pre-operative coronary revascularization improves outcomes, even in the presence of ischemic disease. But evidence has been accumulating that routine use of beta-blockers may not benefit as many patients as was hoped, and may actually cause harm. The 2007 joint American College of Cardiology (ACC) and American Heart Association (AHA) guidelines on perioperative cardiovascular evaluation and care for noncardiac surgery gives its strongest recommendation (class I: benefit clearly outweighs risk) for perioperative beta-blocker use only for patients at high risk: those with known ischemic heart disease undergoing vascular surgery and those who are already on these drugs before surgery.⁶

However, there are still gaps in our knowledge. Perhaps, with proper implementation, we may be able to use beta-blockers to improve outcomes in patients at intermediate risk as well. In this paper, we review the rationale and the evidence for and against perioperative use of beta-blockers and provide practical guidance for internists and hospitalists.

WHY CARDIAC EVENTS OCCUR AFTER SURGERY

Adverse cardiovascular events such as myocardial infarction and unstable angina are the leading causes of death after surgery.⁷ Such events occur in approximately 1% of patients older than 50 years undergoing elective inpatient surgery, but this number may be higher (approximately 5%) in those with known or suspected coronary disease.^8,9 Perioperative cardiac events can also be harbingers of further complications, dramatically increasing hospital length of stay.¹⁰

Some ischemic events are caused by physiologic derangements involving the balance between inflammatory mediators, sympathetic tone, and oxygen supply and demand that occur under the stress of surgery. Others are more “traditional” in etiology, involving acute plaque rupture, thrombosis, and occlusion. Studies have consistently found a correlation between perioperative ischemia and cardiac events (both in-hospital and long-term) and death.^11–17 Other studies suggest that most perioperative cardiac infarcts are non-Q-wave events,¹⁸ and most events occur within the first few days after surgery, particularly the first 48 hours, when the effects of anesthetics, pain, fluid shifts, and physiologic derangements are greatest.

Factors that may trigger acute occlusion in the perioperative period include abrupt changes in sympathetic tone, increased levels of cortisol and catecholamines, and tissue hypoxia. Other potential triggers activated or increased by the stress of surgery include coagulation factors such as alterations in platelet function; inflammatory factors such as tumor necrosis factor alpha, interleukin 1, interleukin 6, and C-reactive protein; and metabolism of free fatty acids (which contribute to increased oxygen demand as well as endothelial dysfunction).^9,19,20

A 1996 autopsy study found that 38 (90%) of 42 patients who died of a perioperative infarct had evidence of acute plaque rupture or plaque hemorrhage on coronary sectioning, findings corroborated in another, similar study.^21,22 These studies suggest that multiple causes contribute to perioperative myocardial infarction, and a single strategy may not suffice for prevention.

IF BETA-BLOCKERS PROTECT, HOW DO THEY DO IT?

Beta-blockers have several effects that should, in theory, protect against cardiac events during and after surgery.²³ They reduce cardiac oxygen demand by reducing the force of contraction and the heart rate, and they increase the duration of diastole, when the heart muscle is perfused. They are also antiarrhythmic, and they may limit free radical production, metalloproteinase activity, and myocardial plaque inflammation.²⁴

Some researchers have speculated that using beta-blockers long-term may alter intra-cellular signaling processes, for example decreasing the expression of receptors that receive signals for cell death, which in turn may affect the response to reperfusion cell injury and death. If this is true, there may be an advantage to starting beta-blockers well in advance of surgery.²⁵

EARLY CLINICAL EVIDENCE IN FAVOR OF PERIOPERATIVE BETA-BLOCKER USE

Evidence in patients at high risk

Mangano et al,² in a study published in 1996, randomized 200 patients with known coronary disease or established risk factors for it who were undergoing noncardiac surgery to receive the beta-blocker atenolol orally and intravenously or placebo in the immediate perioperative period. Fewer patients in the atenolol group died in the first 6 months after hospital discharge (0 vs 8%, P < .001), the first year (3% vs 14%, P = .005), and the first 2 years (10% vs 21%, P = .019). However, there was no difference in short-term outcomes, and the study excluded patients who died in the immediate postoperative period. If these patients had been included in the analysis, the difference in the death rate at 2 years would not have been statistically significant.²⁶ Other critical findings: more patients in the atenolol group were using angiotensin-converting enzyme inhibitors and beta-blockers when they were discharged, and the placebo group had slightly more patients with prior myocardial infarction or diabetes.²⁷ (Atenolol is available in the United Sates as Tenormin and generically.)

Poldermans et al,³ in a study published in 1999, randomized 112 vascular surgery patients to receive either oral bisoprolol or placebo. These patients were selected from a larger cohort of 1,351 patients on the basis of high-risk clinical features and abnormal results on dobutamine echocardiography. Bisoprolol was started at least 1 week before surgery (range 7–89 days, mean 37 days), and patients were reevaluated before surgery so that the dose could be titrated to a goal heart rate of less than 60 beats per minute. After surgery, the drug was continued for another 30 days. The study was stopped early because the bisoprolol group had a 90% lower rate of non-fatal myocardial infarction and cardiac death at 30 days. Despite the study’s limitation (eg, enrolling selected patients and using an unblinded protocol), these compelling findings made a strong case for the use of beta-blockers perioperatively in patients at high risk, ie, those with ischemic heart disease who are undergoing major vascular surgery. (Bisoprolol is available in the United States as Zebeta and generically)

Evidence in patients at intermediate risk

Boersma et al²⁸ performed a follow-up to the study by Poldermans et al, published in 2001, in which they analyzed characteristics of all 1,351 patients who had been originally considered for enrollment. Using regression analysis, they identified seven clinical risk factors that predicted adverse cardiac events: angina, prior myocardial infarction, congestive heart failure, prior stroke, diabetes, renal failure, and age 70 years or older. Furthermore, for the entire cohort, patients receiving beta-blockers had a lower risk of cardiac complications (0.8%) than those not receiving beta-blockers (2.3%). In particular, the patients at intermediate risk (defined as having one or two risk factors) had a very low event rate regardless of stress test results, provided they were on beta-blockers: their risk of death or myocardial infarction was 0.9%, compared with 3.0% for those not on beta-blockers.

The authors concluded that dobutamine stress testing may not be necessary in patients at intermediate risk if beta-blockers are appropriately prescribed. However, others took issue with their data and conclusions, arguing that there have been so few trials that the data are still inconclusive and inadequate to ascertain the benefit of perioperative beta-blockade, particularly in patients not at high risk.^26,29

The Revised Cardiac Risk Index. Although the Boersma risk-factor index is not used in general practice, numerous experts^27,20–32 recommend a similar one, the Revised Cardiac Risk Index, devised by Lee et al.⁸ This index consists of six risk factors, each of which is worth one point:

• Congestive heart failure, based on history or examination
• Myocardial infarction, symptomatic ischemic heart disease, or a positive stress test
• Renal insufficiency (ie, serum creatinine level > 2 mg/dL)
• History of stroke or transient ischemic attack
• Diabetes requiring insulin
• High-risk surgery (defined as intrathoracic, intra-abdominal, or suprainguinal vascular surgery).

Patients with three or more points are considered to be at high risk, and those with one or two points are considered to be at intermediate risk. The ACC/AHA 2007 guidelines⁶ use a modified version of this index that considers the issue of surgical risk separately from the other five clinical conditions.

Devereaux et al³³ performed a meta-analysis, published in 2005, of 22 studies of perioperative beta-blockade. They concluded that beta-blockers had no discernable benefit in any outcome measured, including deaths from any cause, deaths from cardiovascular causes, other cardiac events, hypotension, bradycardia, and bronchospasm. However, they based this conclusion on the use of a 99% confidence interval for each relative risk, which they believed was justified because the trials were small and the numbers of events were only moderate. When the outcomes are assessed using the more common 95% confidence interval, benefit was detected in the combined end point of cardiovascular death, nonfatal myocardial infarction, and nonfatal cardiac arrest.

Yang et al,³⁴ Brady et al,³⁵ and Juul et al³⁶ performed three subsequent randomized trials that added to the controversy. Most of the patients in these trials were at intermediate or low risk, and none of the trials found a significant benefit with perioperative beta-blocker use. However, the protocols in these studies were different from the one in the study by Poldermans et al,³ which had found perioperative beta-blockade to be beneficial. Whereas patients in that earlier study started taking a beta-blocker at least 1 week before surgery (and on average much earlier), had their dose aggressively titrated to a target heart rate, and continued taking it for 30 days afterward, the protocols in the later trials called for the drug to be started within 24 hours before surgery and continued for only a short time afterward.

Lindenauer et al,³⁷ in a retrospective study published in 2005, found that fewer surgical patients who received beta-blockers in the hospital died in the hospital. The researchers used an administrative database of more than 780,000 patients who underwent noncardiac surgery, and they used propensity-score matching to compare the postoperative mortality rates of patients who received beta-blockers and a matched group in the same large cohort who did not. Beta-blockers were associated with a lower morality rate in patients in whom the Revised Cardiac Risk Index score was 3 or greater. However, although there was a trend toward a lower rate with beta-blocker use in patients whose score was 2 (ie, at intermediate risk), the difference was not statistically significant, and patients with a score of 0 or 1 saw no benefit and were possibly harmed.

The authors admitted that their study had a number of limitations, including a retrospective design and the use of an administrative database for information regarding risk index conditions and comorbidities. In addition, because they assumed that any patient who received a beta-blocker on the first 2 hospital days was receiving appropriate perioperative treatment, they may have incorrectly estimated the number of patients who actually received these drugs as a risk-reduction strategy. For instance, some patients at low risk could have received beta-blockers for treatment of a specific event, which would be reflected as an increase in event rates for this group. They also had no data on what medications the patients received before they were hospitalized or whether the dose was titrated effectively. The study excluded all patients with congestive heart failure and chronic obstructive pulmonary disease, who may be candidates for beta-blockers in actual practice. In fact, a recent observational study in patients with severe left ventricular dysfunction suggested that these drugs substantially reduced the incidence of death in the short term and the long term.³⁸ Finally, half the surgeries were nonelective, which makes extrapolation of their risk profile by the Revised Cardiac Risk Index difficult, since Lee et al excluded patients undergoing emergency surgery from the cohorts from which they derived and validated their index criteria.

Nevertheless, the authors concluded that patients at intermediate risk derive no benefit from perioperative beta-blocker use, and that the odds ratio for death was actually higher in patients with no risk factors who received a beta-blocker.

DOES PERIOPERATIVE BETA-BLOCKER USE CAUSE HARM?

The published data on whether perioperative beta-blocker use harms patients are conflicting and up to now have been limited.

Stone et al³⁹ reported a substantial incidence of bradycardia requiring atropine in patients treated with a single dose of a beta-blocker preoperatively, but the complications were not clearly characterized.

The Perioperative Beta-Blockade trial.³⁵ Significantly more patients given short-acting metoprolol had intraoperative falls in blood pressure and heart rate, and more required inotropic support during surgery, although the treating anesthesiologists refused to be blinded in that study. (Short-acting metoprolol is available in the United States as Lopressor and generically.)

Devereaux et al,³³ in their meta-analysis, found a higher risk of bradycardia requiring treatment (but not a higher risk of hypotension) in beta-blocker users in nine studies, including the study by Stone et al and the Perioperative Beta-Blockade trial (relative risk 2.27, 95% confidence interval 1.36–3.80).

Conversely, at least three other studies found no difference in rates of intraoperative events.^36,40,41 There are few data on the incidence of other complications such as perioperative pulmonary edema and bronchospasm.

POISE: THE FIRST LARGE RANDOMIZED TRIAL

In May 2008, results were published from POISE, the first large randomized controlled trial of perioperative beta-blockade.¹ An impressive 8,351 patients—most of them at intermediate risk—were randomized to receive extended-release metoprolol succinate or placebo starting just before surgery and continuing for 30 days afterward.

Although the incidence of the primary composite end point (cardiovascular death, nonfatal myocardial infarction, and nonfatal cardiac arrest) was lower at 30 days in the metoprolol group than in the placebo group (5.8% vs 6.9%, hazard ratio 0.83, P = .04), other findings were worrisome: more metoprolol recipients died of any cause (3.1% vs 2.3%, P = .03) or had a stroke (1.0% vs 0.5%, P = .005). The major contributor to the higher mortality rate in this group appears to have been sepsis.

How beta-blockers might promote death by sepsis is unclear. The authors offered two possible explanations: perhaps beta-blocker-induced hypotension predisposes patients to infection and sepsis, or perhaps the slower heart rate and lower force of contraction induced by beta-blockers could mask normal responses to systemic infection, which in turn could delay recognition and treatment or impede the normal immune response. These mechanisms, like others, are speculative.

The risks of other adverse outcomes such as bradycardia and hypotension were substantially higher in the metoprolol group. The authors also pointed out that most of the patients who suffered nonfatal strokes were subsequently disabled or incapacitated, while most of those who suffered nonfatal cardiac events did not progress to further cardiac problems.

This new study has not yet been rigorously debated, but it will likely come under scrutiny for its dosing regimen (extended-release metoprolol succinate 100 mg or placebo 2–4 hours before surgery; another 100 mg or placebo 6 hours after surgery or sooner if the heart rate was 80 beats per minute or more and the systolic blood pressure 100 mm Hg or higher; and then 200 mg or placebo 12 hours after the second dose and every 24 hours thereafter for 30 days). This was fairly aggressive, especially for patients who have never received a beta-blocker before. In contrast, the protocol for the Perioperative Beta-Blockade trial called for only 25 to 50 mg of short-acting metoprolol twice a day. Another criticism is that the medication was started only a few hours before surgery, although there is no current standard practice for either the dose or when the treatment should be started. The population had a fairly high rate of cerebrovascular disease (perhaps predisposing to stroke whenever blood pressure dropped), and 10% of patients were undergoing urgent or emergency surgery, which carries a higher risk of morbidity.

ANY ROLE FOR BETA-BLOCKERS IN THOSE AT INTERMEDIATE RISK?

Thus, in the past decade, the appropriate perioperative use of beta-blockers, which, after the findings by Mangano et al and Poldermans et al, were seen as potentially beneficial for any patient at risk of coronary disease, with little suggestion of harm, has become more clearly defined, and the risks are more evident. The most compelling evidence in favor of using them comes from patients with ischemic heart disease undergoing vascular surgery; the 2007 ACC/AHA guidelines recommend that this group be offered beta-blockers in the absence of a contraindication (class I recommendation: benefit clearly outweighs risk).⁶ The guidelines also point out that these drugs should be continued in patients already taking them for cardiac indications before surgery, because ischemia may be precipitated if a beta-blocker is abruptly discontinued.^42,43

Additionally, the guidelines recommend considering beta-blockers for vascular surgery patients at high cardiac risk (with a Revised Cardiac Risk Index score of 3 or more), even if they are not known to have ischemic heart disease. This is a class IIa recommendation (the benefit outweighs the risk, but more studies are required).

The guidelines also recommend that beta-blockers be considered for patients who have a score of 0 if they are undergoing vascular surgery (class IIb recommendation) or a score of 1 if they are undergoing vascular surgery (class IIa recommendation) or intermediate-risk surgery (class IIb recommendation). However, in view of the POISE results, these recommendations need to be carefully scrutinized.

These data notwithstanding, beta-blockers still might be beneficial in perioperative patients at intermediate risk.

Start beta-blockers sooner?

To help patients at intermediate risk (such as those with diabetes without known heart disease), we may need to do what Poldermans et al did³: instead of seeing patients only once a day or two before surgery, we may need to do the preoperative assessment as much as a month before and, if necessary, start a beta-blocker at a low dose, titrate it to a goal heart rate, and follow the patient closely up until surgery and afterward.

The importance of heart-rate control was illustrated in a recent cohort study of perioperative beta-blockers in vascular surgery patients,⁴⁴ in which higher beta-blocker doses, carefully monitored, were associated with less ischemia and cardiac enzyme release. In addition, long-term mortality rates were lower in patients with lower heart rates. And Poldermans et al⁴⁵ recently performed a study in more than 700 intermediate-risk patients who were divided into two groups, one that underwent preoperative stress testing and one that did not. Beta-blockers were given to both groups, and doses were titrated to a goal heart rate of less than 65. The patients with optimally controlled heart rates had the lowest event rates.

However, the logistics of such a program would be challenging. For the most part, internists and hospitalists involved in perioperative assessment do not control the timing of referral or surgery, and adjustments cannot be made for patients whose preoperative clinic visit falls only a few days before surgery. Instituting a second or third visit to assess the efficacy of beta-blockade burdens the patient and may not be practical.

Are all beta-blockers equivalent?

An additional factor is the choice of agent. While the most significant studies of perioperative beta-blockade have used beta-1 receptor-selective agents (ie, metoprolol, atenolol, and bisoprolol), there is no prospective evidence that any particular agent is superior. However, a recent retrospective analysis of elderly surgical patients did suggest that longer-acting beta-blockers may be preferable: patients who had been on atenolol in the year before surgery had a 20% lower risk of postoperative myocardial infarction or death than those who had been on short-acting metoprolol, with no difference in noncardiac outcomes.⁴⁶

Ezetimibe (Zetia) was licensed by the US Food and Drug Administration in 2002 on the basis of its ability to reduce low-density lipoprotein cholesterol (LDL-C) levels. The reductions are mild, approximately 15%,¹ which is comparable to the effects of a stringent diet and exercise or of a statin in titrated doses.

See related commentary

However, there was no evidence that ezetimbe, which has a unique mechanism of action, delivers a benefit in terms of clinical outcomes. Despite this, the use of ezetimibe (alone or in fixed-dose combination with simvastatin, a preparation sold as Vytorin) grew rapidly, generating annual sales of $5.2 billion. Clinicians and the manufacturer (Merck/Schering-Plough) broadly assumed that LDL-C reduction would carry ezetimibe’s day as clinical trials emerged.

The assumption seemed reasonable, since evidence from the past 3 decades has established a clear link between lowering LDL-C levels via diverse mechanisms and positive clinical outcomes, particularly lower rates of cardiovascular disease and death. Indeed, LDL-C measurement is now a focus of cardiovascular risk assessment and management, as reflected in national treatment guidelines.

THE ENHANCE TRIAL: EZETIMIBE FAILS A KEY TEST

Unexpectedly, ezetimibe failed its first step in clinical trial validation, the Ezetimibe and Simvastatin in Hypercholesterolemia Enhances Atherosclerosis Regression (ENHANCE) trial.² Apart from the scientifically irrelevant political regulatory intrigue generated by the sponsor’s conduct in this trial, ENHANCE’s findings challenge us to confront issues of what we assume vs what we really know, and how to interpret the complex results of clinical trials.

To be fair to the trial’s investigators, ENHANCE achieved its objective of enrolling a population with a very high LDL-C level, which is ezetimibe’s target and has been widely used in the study of atherosclerosis progression as a marker of potential drug benefit. Nevertheless, and even though the LDL-C level 2 years later was 52 mg/dL lower in the group receiving ezetimibe/simvastatin than in the group receiving simvastatin alone (Zocor), at LDL-C levels that are typically associated with atherosclerosis progression (140–190 mg/dL), ezetimibe failed to reduce the progression of atherosclerosis.

Supplementary appendix to Kastelein JJ, et al. Simvastatin with or without ezetimibe in familial hypercholesterolemia. N Engl J Med 2008; 358:1431–1443. doi:10.1056/NEJMoa0800742. Copyright 2008, Massachusetts Medical Society.

These trends are particularly worrisome, given that the ezetimibe/simvastatin group achieved a greater reduction in C-reactive protein levels, which typically has resulted in superior outcomes in atherosclerosis³ and clinical effects⁴ in combination with LDL-C reduction.

In view of these findings, should clinicians stand firm and continue to use ezetimibe? Or should we reevaluate our position and await more data about this unique, first-in-class compound?

WISHFUL POST HOC HYPOTHESES

In this issue of the Cleveland Clinic Journal of Medicine, Dr. Michael Davidson,⁵ a respected lipid expert but one invested in ezetimibe’s development, assures us that all is in order and that the results of ENHANCE can be explained away by several arguments, most notably that most of the trial’s participants had previously received lipid-lowering treatment, which obscured the effects of ezetimibe. Moreover, he argues that ezetimibe’s mechanism of action is well understood and that the drug is safe and well tolerated and thus should remain a first-line treatment for hyperlipidemia.

These arguments may eventually prove to be correct, but as of now they are merely wishful post hoc hypotheses awaiting more data apart from ENHANCE. Negative clinical trials do occur as a matter of chance, but we should be cautious in any attempts to explain away a trial that was designed, executed, and reported as conceived simply because the results do not match our expectations.

Confronted with ENHANCE, the astute clinician should ask three questions: Do we really understand ezetimibe’s mechanism of action? Do other lines of evidence indicate the drug is beneficial? And how reliable is the arterial thickness as a surrogate end point?

DO WE UNDERSTAND EZETIMIBE’S MECHANISM OF ACTION?

Do we understand ezetimibe’s full mechanism of action? Not really.

True, ezetimibe inhibits cholesterol transport, a process that is integral both to cholesterol’s enteric absorption and to its systemic clearance. But although Dr. Davidson asserts that ezetimibe has cellular effects similar to those of statins, in fact it has the opposite effect on HMG-coA reductase, and no effects on LDL receptors.⁶

Furthermore, although initial studies suggested that ezetimibe inhibits enteric cholesterol absorption by inhibiting the Niemann-Pick C1L1 (NPC1L1) receptor, more recent investigations call this into serious question and point more definitively at a receptor known as scavenger receptor-B1 (SR-B1). As stated in a recent editorial, “SR-B1 in the apical site of enterocytes is the primary high-affinity site of cholesterol uptake and ezetimibe can inhibit this process. Moreover, the [possibility is ruled out] of NPC1L1 being a major player in this cholesterol uptake. This is at variance with the view of the colleagues from Schering-Plough who claim the same for NPC1L1.”⁷

SR-B1 is also a high-affinity receptor for high-density lipoprotein⁸ and thus is active in the antiatherosclerotic process of reverse cholesterol transport, inhibition of which significantly accelerates the development of atherosclerosis.⁹

Additionally, in vitro and thus unrelated to the effects of changing cholesterol concentration, ezetimibe down-regulates SR-B1 and another key cholesterol transporter protein called ABCA1.¹⁰ Further, ezetimibe induces down-regulation of raft protein domains, including CD36,¹¹ another effect opposite to that of statins.

These little-recognized effects of ezetimibe are among many that are completely unrelated to enteric cholesterol absorption. Yet, they are likely to be active within the liver and systemically where these proteins reside, and they are putatively proatherosclerotic. Contrary to often-cited opinion, ezetimibe is systemically absorbed, with 11% of the compound excreted in the urine.¹² Thus, the compound is systemically available to exert these same actions in the liver and elsewhere. Moreover, the absorbed drug is glucuronidated and is extensively recirculated in the liver in a form (its glucuronide) that is more potent than the parent compound.

In sum, present opinion is that ezetimibe inhibits lipid transport and interacts with a variety of receptors, not only in the gut but also systemically at the cell membrane and also inside the cell, focally disrupting several tightly regulated biologic processes.⁷ Thus, although ezetimibe reduces serum LDL-C levels via its effect in the gut, this effect may well be offset or even overridden systemically by other, unmeasurable effects, leading to counterintuitive results in terms of atherosclerosis or clinical events.

This would not be the first time a lipid-lowering drug has disappointed us: torcetrapib, another transport inhibitor, dramatically raises serum high-density lipoprotein cholesterol levels and reduces LDL-C but was found not only to have no effect on atherosclerosis, but also to potentiate adverse clinical outcomes.

The net impact of these other actions of ezetimibe is not known. We will discover its true clinical effects only through studies of endothelial function, atherosclerosis, and clinical cardiovascular outcomes. ENHANCE, which looked at atherosclerosis, is thus our strongest signal to date on the net effect of ezetimibe.

DO OTHER LINES OF EVIDENCE INDICATE EZETIMIBE IS BENEFICIAL?

Can we be reassured that ENHANCE’s results are spurious on the basis of other lines of evidence? Again, not really.

Experiments in animals, particularly in mice,¹³ have shown that ezetimibe may be antiatherosclerotic, although mice are considered the “worst model”⁷ for the study of ezetimibe, and notably, LDL-C levels were lowered far more in these experiments than they are clinically. Enthusiasm for these animal models should be tempered by interspecies variability in ezetimibe’s “off-target” effects and in the recent failure of other lipid transport drugs in human trials (torcetrapib and ACAT inhibitors) that had shown initial success in animals. No animal model is established for evaluating drugs of ezetimibe’s class, given its complex mechanism of action.

In human studies, the only other surrogate of the net effect of ezetimibe is endothelial function. Among several randomized clinical trials of ezetimibe,^14–18 only one was designed to compare the effects of ezetimibe alone, ezetimibe plus a statin, and a statin by itself in titrated or in maximum doses.¹⁵ After 4 weeks of therapy, all groups had lower LDL-C levels. However, ezetimibe monotherapy and ezetimibe/simvastatin combination therapy had no detectable effect on the arterial response to acetylcholine, but atorvastatin (Lipitor) monotherapy did. To be fair, the other (very small) trials showed mixed results, thus keeping the hypothesis of ezetimibe’s benefit alive, but with nothing close to a clear signal of benefit.

IS ARTERIAL THICKNESS RELIABLE AS A SURROGATE END POINT?

Was the principal problem in ENHANCE the use of carotid intima-media thickness as the primary end point? No.

This issue has received a lot of attention, much of which I believe is misinformed. No trial end point is infallible, including carotid intima-media thickness, and one must remain open to the possibility of chance findings. However, it has been a relatively reasonable end point in trials of diverse cardiovascular preventive strategies, including lipid-lowering, blood-pressure-lowering, and lifestyle interventions and as a directional biomarker of clinical atherosclerotic events.

We should be cautious about comparing data on carotid intima-media thickness from different trials, as Dr. Davidson attempts to do, in view of methodologic and population differences: each trial must be considered independently. Of greatest concern in ENHANCE is the consistency among intima-media thickness end points, including strong trends toward adverse effects in the most diseased carotid and femoral segments.

Moreover, ENHANCE’s detractors contend that the carotid intima-media thickness of the studied population was normal, citing this as evidence of delipidation from prior treatment. Although not impossible (as shown by the work of Zhao and colleagues in the setting of prolonged, intense lipid-lowering therapy¹⁹), at the moment this hypothesis is a matter of conjecture in the ENHANCE participants, particularly because their LDL-C levels were still quite elevated during the trial and conceivably even before randomization.

But these patients were not normal: they were typical patients with familial hypercholesterolemia with extremely elevated LDL-C levels and abnormally thick arteries for their age. Population screening estimates show that, for age and sex, the carotid intima-media thickness values in ENHANCE would lie in the upper quartile of those in the general population.²⁰ Moreover, their mean value is consistent with that in similar-aged groups of patients with familial hypercholesterolemia, even with lower rates of prior statin pretreatment.²¹

The most convincing evidence for the validity of the ENHANCE findings comes from the published subgroup data (Figure 1). In participants whose baseline carotid intima-media thickness was above the median at baseline, the thickness increased more with ezetimibe/simvastatin than with simvastatin alone. The same was true in the subgroup with above-average LDL-C levels at baseline. The subgroups with no prior statin treatment, low-dose prior statin treatment, and high-dose prior statin showed no heterogeneity of response: their carotid intima-media thickness increased more with ezetimibe/simvastatin than with simvastatin alone. None of these differences was statistically significant; however, these prespecified subgroup data seemingly invalidate arguments against the ENHANCE results based on carotid intima-media thickness findings.

In this context, ENHANCE can only be interpreted as a strong initial negative signal, a “red flag” about ezetimibe’s net health benefits.

WHAT NEXT?

The proper present focus of this debate is not on LDL-C but rather on ezetimibe, its unique mechanism of action, and on the need for more evidence about this complex compound.

At present, ezetimibe’s mechanism of action is not fully understood, and its benefit—for now, only mild LDL-C reduction—is too uncertain for us to be spending $5.2 billion a year for it. Its manufacturer is fortunate that the drug is even licensed, given the current and seemingly appropriate regulatory changes under which drugs introducing new therapeutic classes are scrutinized more closely for benefits and risks. “Safe and well tolerated,” as contended by Dr. Davidson, is not nearly enough: drugs must show clinically important benefits. We still know too little about this drug, the manufacturer of which has invested far more in marketing than in science, a point on which Dr. Davidson and I agree.

In 2008, ezetimibe is an appropriate candidate for testing in clinical trials, and in years to come it may be worthy of clinical attention—if rigorous and objectively conducted clinical trials prove its worth. At present, clinical equipoise dictates that ezetimibe is not an appropriate alternative to a statin in titrated doses, to the addition of other lipid-lowering drugs to a statin, to greater attention to drug adherence, or to lifestyle modification.

For the moment, given the ENHANCE results, the clinical usefulness of ezetimibe still remains to be proven. Much more evidence is needed before we can confidently reembrace the clinical use of ezetimibe.

Ezetimibe (Zetia) was licensed by the US Food and Drug Administration in 2002 on the basis of its ability to reduce low-density lipoprotein cholesterol (LDL-C) levels. The reductions are mild, approximately 15%,¹ which is comparable to the effects of a stringent diet and exercise or of a statin in titrated doses.

See related commentary

However, there was no evidence that ezetimbe, which has a unique mechanism of action, delivers a benefit in terms of clinical outcomes. Despite this, the use of ezetimibe (alone or in fixed-dose combination with simvastatin, a preparation sold as Vytorin) grew rapidly, generating annual sales of $5.2 billion. Clinicians and the manufacturer (Merck/Schering-Plough) broadly assumed that LDL-C reduction would carry ezetimibe’s day as clinical trials emerged.

The assumption seemed reasonable, since evidence from the past 3 decades has established a clear link between lowering LDL-C levels via diverse mechanisms and positive clinical outcomes, particularly lower rates of cardiovascular disease and death. Indeed, LDL-C measurement is now a focus of cardiovascular risk assessment and management, as reflected in national treatment guidelines.

THE ENHANCE TRIAL: EZETIMIBE FAILS A KEY TEST

Unexpectedly, ezetimibe failed its first step in clinical trial validation, the Ezetimibe and Simvastatin in Hypercholesterolemia Enhances Atherosclerosis Regression (ENHANCE) trial.² Apart from the scientifically irrelevant political regulatory intrigue generated by the sponsor’s conduct in this trial, ENHANCE’s findings challenge us to confront issues of what we assume vs what we really know, and how to interpret the complex results of clinical trials.

To be fair to the trial’s investigators, ENHANCE achieved its objective of enrolling a population with a very high LDL-C level, which is ezetimibe’s target and has been widely used in the study of atherosclerosis progression as a marker of potential drug benefit. Nevertheless, and even though the LDL-C level 2 years later was 52 mg/dL lower in the group receiving ezetimibe/simvastatin than in the group receiving simvastatin alone (Zocor), at LDL-C levels that are typically associated with atherosclerosis progression (140–190 mg/dL), ezetimibe failed to reduce the progression of atherosclerosis.

Supplementary appendix to Kastelein JJ, et al. Simvastatin with or without ezetimibe in familial hypercholesterolemia. N Engl J Med 2008; 358:1431–1443. doi:10.1056/NEJMoa0800742. Copyright 2008, Massachusetts Medical Society.

These trends are particularly worrisome, given that the ezetimibe/simvastatin group achieved a greater reduction in C-reactive protein levels, which typically has resulted in superior outcomes in atherosclerosis³ and clinical effects⁴ in combination with LDL-C reduction.

In view of these findings, should clinicians stand firm and continue to use ezetimibe? Or should we reevaluate our position and await more data about this unique, first-in-class compound?

WISHFUL POST HOC HYPOTHESES

In this issue of the Cleveland Clinic Journal of Medicine, Dr. Michael Davidson,⁵ a respected lipid expert but one invested in ezetimibe’s development, assures us that all is in order and that the results of ENHANCE can be explained away by several arguments, most notably that most of the trial’s participants had previously received lipid-lowering treatment, which obscured the effects of ezetimibe. Moreover, he argues that ezetimibe’s mechanism of action is well understood and that the drug is safe and well tolerated and thus should remain a first-line treatment for hyperlipidemia.

These arguments may eventually prove to be correct, but as of now they are merely wishful post hoc hypotheses awaiting more data apart from ENHANCE. Negative clinical trials do occur as a matter of chance, but we should be cautious in any attempts to explain away a trial that was designed, executed, and reported as conceived simply because the results do not match our expectations.

Confronted with ENHANCE, the astute clinician should ask three questions: Do we really understand ezetimibe’s mechanism of action? Do other lines of evidence indicate the drug is beneficial? And how reliable is the arterial thickness as a surrogate end point?

DO WE UNDERSTAND EZETIMIBE’S MECHANISM OF ACTION?

Do we understand ezetimibe’s full mechanism of action? Not really.

True, ezetimibe inhibits cholesterol transport, a process that is integral both to cholesterol’s enteric absorption and to its systemic clearance. But although Dr. Davidson asserts that ezetimibe has cellular effects similar to those of statins, in fact it has the opposite effect on HMG-coA reductase, and no effects on LDL receptors.⁶

Furthermore, although initial studies suggested that ezetimibe inhibits enteric cholesterol absorption by inhibiting the Niemann-Pick C1L1 (NPC1L1) receptor, more recent investigations call this into serious question and point more definitively at a receptor known as scavenger receptor-B1 (SR-B1). As stated in a recent editorial, “SR-B1 in the apical site of enterocytes is the primary high-affinity site of cholesterol uptake and ezetimibe can inhibit this process. Moreover, the [possibility is ruled out] of NPC1L1 being a major player in this cholesterol uptake. This is at variance with the view of the colleagues from Schering-Plough who claim the same for NPC1L1.”⁷

SR-B1 is also a high-affinity receptor for high-density lipoprotein⁸ and thus is active in the antiatherosclerotic process of reverse cholesterol transport, inhibition of which significantly accelerates the development of atherosclerosis.⁹

Additionally, in vitro and thus unrelated to the effects of changing cholesterol concentration, ezetimibe down-regulates SR-B1 and another key cholesterol transporter protein called ABCA1.¹⁰ Further, ezetimibe induces down-regulation of raft protein domains, including CD36,¹¹ another effect opposite to that of statins.

These little-recognized effects of ezetimibe are among many that are completely unrelated to enteric cholesterol absorption. Yet, they are likely to be active within the liver and systemically where these proteins reside, and they are putatively proatherosclerotic. Contrary to often-cited opinion, ezetimibe is systemically absorbed, with 11% of the compound excreted in the urine.¹² Thus, the compound is systemically available to exert these same actions in the liver and elsewhere. Moreover, the absorbed drug is glucuronidated and is extensively recirculated in the liver in a form (its glucuronide) that is more potent than the parent compound.

In sum, present opinion is that ezetimibe inhibits lipid transport and interacts with a variety of receptors, not only in the gut but also systemically at the cell membrane and also inside the cell, focally disrupting several tightly regulated biologic processes.⁷ Thus, although ezetimibe reduces serum LDL-C levels via its effect in the gut, this effect may well be offset or even overridden systemically by other, unmeasurable effects, leading to counterintuitive results in terms of atherosclerosis or clinical events.

This would not be the first time a lipid-lowering drug has disappointed us: torcetrapib, another transport inhibitor, dramatically raises serum high-density lipoprotein cholesterol levels and reduces LDL-C but was found not only to have no effect on atherosclerosis, but also to potentiate adverse clinical outcomes.

The net impact of these other actions of ezetimibe is not known. We will discover its true clinical effects only through studies of endothelial function, atherosclerosis, and clinical cardiovascular outcomes. ENHANCE, which looked at atherosclerosis, is thus our strongest signal to date on the net effect of ezetimibe.

DO OTHER LINES OF EVIDENCE INDICATE EZETIMIBE IS BENEFICIAL?

Can we be reassured that ENHANCE’s results are spurious on the basis of other lines of evidence? Again, not really.

Experiments in animals, particularly in mice,¹³ have shown that ezetimibe may be antiatherosclerotic, although mice are considered the “worst model”⁷ for the study of ezetimibe, and notably, LDL-C levels were lowered far more in these experiments than they are clinically. Enthusiasm for these animal models should be tempered by interspecies variability in ezetimibe’s “off-target” effects and in the recent failure of other lipid transport drugs in human trials (torcetrapib and ACAT inhibitors) that had shown initial success in animals. No animal model is established for evaluating drugs of ezetimibe’s class, given its complex mechanism of action.

In human studies, the only other surrogate of the net effect of ezetimibe is endothelial function. Among several randomized clinical trials of ezetimibe,^14–18 only one was designed to compare the effects of ezetimibe alone, ezetimibe plus a statin, and a statin by itself in titrated or in maximum doses.¹⁵ After 4 weeks of therapy, all groups had lower LDL-C levels. However, ezetimibe monotherapy and ezetimibe/simvastatin combination therapy had no detectable effect on the arterial response to acetylcholine, but atorvastatin (Lipitor) monotherapy did. To be fair, the other (very small) trials showed mixed results, thus keeping the hypothesis of ezetimibe’s benefit alive, but with nothing close to a clear signal of benefit.

IS ARTERIAL THICKNESS RELIABLE AS A SURROGATE END POINT?

Was the principal problem in ENHANCE the use of carotid intima-media thickness as the primary end point? No.

This issue has received a lot of attention, much of which I believe is misinformed. No trial end point is infallible, including carotid intima-media thickness, and one must remain open to the possibility of chance findings. However, it has been a relatively reasonable end point in trials of diverse cardiovascular preventive strategies, including lipid-lowering, blood-pressure-lowering, and lifestyle interventions and as a directional biomarker of clinical atherosclerotic events.

We should be cautious about comparing data on carotid intima-media thickness from different trials, as Dr. Davidson attempts to do, in view of methodologic and population differences: each trial must be considered independently. Of greatest concern in ENHANCE is the consistency among intima-media thickness end points, including strong trends toward adverse effects in the most diseased carotid and femoral segments.

Moreover, ENHANCE’s detractors contend that the carotid intima-media thickness of the studied population was normal, citing this as evidence of delipidation from prior treatment. Although not impossible (as shown by the work of Zhao and colleagues in the setting of prolonged, intense lipid-lowering therapy¹⁹), at the moment this hypothesis is a matter of conjecture in the ENHANCE participants, particularly because their LDL-C levels were still quite elevated during the trial and conceivably even before randomization.

But these patients were not normal: they were typical patients with familial hypercholesterolemia with extremely elevated LDL-C levels and abnormally thick arteries for their age. Population screening estimates show that, for age and sex, the carotid intima-media thickness values in ENHANCE would lie in the upper quartile of those in the general population.²⁰ Moreover, their mean value is consistent with that in similar-aged groups of patients with familial hypercholesterolemia, even with lower rates of prior statin pretreatment.²¹

The most convincing evidence for the validity of the ENHANCE findings comes from the published subgroup data (Figure 1). In participants whose baseline carotid intima-media thickness was above the median at baseline, the thickness increased more with ezetimibe/simvastatin than with simvastatin alone. The same was true in the subgroup with above-average LDL-C levels at baseline. The subgroups with no prior statin treatment, low-dose prior statin treatment, and high-dose prior statin showed no heterogeneity of response: their carotid intima-media thickness increased more with ezetimibe/simvastatin than with simvastatin alone. None of these differences was statistically significant; however, these prespecified subgroup data seemingly invalidate arguments against the ENHANCE results based on carotid intima-media thickness findings.

In this context, ENHANCE can only be interpreted as a strong initial negative signal, a “red flag” about ezetimibe’s net health benefits.

WHAT NEXT?

The proper present focus of this debate is not on LDL-C but rather on ezetimibe, its unique mechanism of action, and on the need for more evidence about this complex compound.

At present, ezetimibe’s mechanism of action is not fully understood, and its benefit—for now, only mild LDL-C reduction—is too uncertain for us to be spending $5.2 billion a year for it. Its manufacturer is fortunate that the drug is even licensed, given the current and seemingly appropriate regulatory changes under which drugs introducing new therapeutic classes are scrutinized more closely for benefits and risks. “Safe and well tolerated,” as contended by Dr. Davidson, is not nearly enough: drugs must show clinically important benefits. We still know too little about this drug, the manufacturer of which has invested far more in marketing than in science, a point on which Dr. Davidson and I agree.

In 2008, ezetimibe is an appropriate candidate for testing in clinical trials, and in years to come it may be worthy of clinical attention—if rigorous and objectively conducted clinical trials prove its worth. At present, clinical equipoise dictates that ezetimibe is not an appropriate alternative to a statin in titrated doses, to the addition of other lipid-lowering drugs to a statin, to greater attention to drug adherence, or to lifestyle modification.

For the moment, given the ENHANCE results, the clinical usefulness of ezetimibe still remains to be proven. Much more evidence is needed before we can confidently reembrace the clinical use of ezetimibe.

Ezetimibe (Zetia) was licensed by the US Food and Drug Administration in 2002 on the basis of its ability to reduce low-density lipoprotein cholesterol (LDL-C) levels. The reductions are mild, approximately 15%,¹ which is comparable to the effects of a stringent diet and exercise or of a statin in titrated doses.

See related commentary

However, there was no evidence that ezetimbe, which has a unique mechanism of action, delivers a benefit in terms of clinical outcomes. Despite this, the use of ezetimibe (alone or in fixed-dose combination with simvastatin, a preparation sold as Vytorin) grew rapidly, generating annual sales of $5.2 billion. Clinicians and the manufacturer (Merck/Schering-Plough) broadly assumed that LDL-C reduction would carry ezetimibe’s day as clinical trials emerged.

The assumption seemed reasonable, since evidence from the past 3 decades has established a clear link between lowering LDL-C levels via diverse mechanisms and positive clinical outcomes, particularly lower rates of cardiovascular disease and death. Indeed, LDL-C measurement is now a focus of cardiovascular risk assessment and management, as reflected in national treatment guidelines.

THE ENHANCE TRIAL: EZETIMIBE FAILS A KEY TEST

Unexpectedly, ezetimibe failed its first step in clinical trial validation, the Ezetimibe and Simvastatin in Hypercholesterolemia Enhances Atherosclerosis Regression (ENHANCE) trial.² Apart from the scientifically irrelevant political regulatory intrigue generated by the sponsor’s conduct in this trial, ENHANCE’s findings challenge us to confront issues of what we assume vs what we really know, and how to interpret the complex results of clinical trials.

To be fair to the trial’s investigators, ENHANCE achieved its objective of enrolling a population with a very high LDL-C level, which is ezetimibe’s target and has been widely used in the study of atherosclerosis progression as a marker of potential drug benefit. Nevertheless, and even though the LDL-C level 2 years later was 52 mg/dL lower in the group receiving ezetimibe/simvastatin than in the group receiving simvastatin alone (Zocor), at LDL-C levels that are typically associated with atherosclerosis progression (140–190 mg/dL), ezetimibe failed to reduce the progression of atherosclerosis.

Supplementary appendix to Kastelein JJ, et al. Simvastatin with or without ezetimibe in familial hypercholesterolemia. N Engl J Med 2008; 358:1431–1443. doi:10.1056/NEJMoa0800742. Copyright 2008, Massachusetts Medical Society.

These trends are particularly worrisome, given that the ezetimibe/simvastatin group achieved a greater reduction in C-reactive protein levels, which typically has resulted in superior outcomes in atherosclerosis³ and clinical effects⁴ in combination with LDL-C reduction.

In view of these findings, should clinicians stand firm and continue to use ezetimibe? Or should we reevaluate our position and await more data about this unique, first-in-class compound?

WISHFUL POST HOC HYPOTHESES

In this issue of the Cleveland Clinic Journal of Medicine, Dr. Michael Davidson,⁵ a respected lipid expert but one invested in ezetimibe’s development, assures us that all is in order and that the results of ENHANCE can be explained away by several arguments, most notably that most of the trial’s participants had previously received lipid-lowering treatment, which obscured the effects of ezetimibe. Moreover, he argues that ezetimibe’s mechanism of action is well understood and that the drug is safe and well tolerated and thus should remain a first-line treatment for hyperlipidemia.

These arguments may eventually prove to be correct, but as of now they are merely wishful post hoc hypotheses awaiting more data apart from ENHANCE. Negative clinical trials do occur as a matter of chance, but we should be cautious in any attempts to explain away a trial that was designed, executed, and reported as conceived simply because the results do not match our expectations.

Confronted with ENHANCE, the astute clinician should ask three questions: Do we really understand ezetimibe’s mechanism of action? Do other lines of evidence indicate the drug is beneficial? And how reliable is the arterial thickness as a surrogate end point?

DO WE UNDERSTAND EZETIMIBE’S MECHANISM OF ACTION?

Do we understand ezetimibe’s full mechanism of action? Not really.

True, ezetimibe inhibits cholesterol transport, a process that is integral both to cholesterol’s enteric absorption and to its systemic clearance. But although Dr. Davidson asserts that ezetimibe has cellular effects similar to those of statins, in fact it has the opposite effect on HMG-coA reductase, and no effects on LDL receptors.⁶

Furthermore, although initial studies suggested that ezetimibe inhibits enteric cholesterol absorption by inhibiting the Niemann-Pick C1L1 (NPC1L1) receptor, more recent investigations call this into serious question and point more definitively at a receptor known as scavenger receptor-B1 (SR-B1). As stated in a recent editorial, “SR-B1 in the apical site of enterocytes is the primary high-affinity site of cholesterol uptake and ezetimibe can inhibit this process. Moreover, the [possibility is ruled out] of NPC1L1 being a major player in this cholesterol uptake. This is at variance with the view of the colleagues from Schering-Plough who claim the same for NPC1L1.”⁷

SR-B1 is also a high-affinity receptor for high-density lipoprotein⁸ and thus is active in the antiatherosclerotic process of reverse cholesterol transport, inhibition of which significantly accelerates the development of atherosclerosis.⁹

Additionally, in vitro and thus unrelated to the effects of changing cholesterol concentration, ezetimibe down-regulates SR-B1 and another key cholesterol transporter protein called ABCA1.¹⁰ Further, ezetimibe induces down-regulation of raft protein domains, including CD36,¹¹ another effect opposite to that of statins.

These little-recognized effects of ezetimibe are among many that are completely unrelated to enteric cholesterol absorption. Yet, they are likely to be active within the liver and systemically where these proteins reside, and they are putatively proatherosclerotic. Contrary to often-cited opinion, ezetimibe is systemically absorbed, with 11% of the compound excreted in the urine.¹² Thus, the compound is systemically available to exert these same actions in the liver and elsewhere. Moreover, the absorbed drug is glucuronidated and is extensively recirculated in the liver in a form (its glucuronide) that is more potent than the parent compound.

In sum, present opinion is that ezetimibe inhibits lipid transport and interacts with a variety of receptors, not only in the gut but also systemically at the cell membrane and also inside the cell, focally disrupting several tightly regulated biologic processes.⁷ Thus, although ezetimibe reduces serum LDL-C levels via its effect in the gut, this effect may well be offset or even overridden systemically by other, unmeasurable effects, leading to counterintuitive results in terms of atherosclerosis or clinical events.

This would not be the first time a lipid-lowering drug has disappointed us: torcetrapib, another transport inhibitor, dramatically raises serum high-density lipoprotein cholesterol levels and reduces LDL-C but was found not only to have no effect on atherosclerosis, but also to potentiate adverse clinical outcomes.

The net impact of these other actions of ezetimibe is not known. We will discover its true clinical effects only through studies of endothelial function, atherosclerosis, and clinical cardiovascular outcomes. ENHANCE, which looked at atherosclerosis, is thus our strongest signal to date on the net effect of ezetimibe.

DO OTHER LINES OF EVIDENCE INDICATE EZETIMIBE IS BENEFICIAL?

Can we be reassured that ENHANCE’s results are spurious on the basis of other lines of evidence? Again, not really.

Experiments in animals, particularly in mice,¹³ have shown that ezetimibe may be antiatherosclerotic, although mice are considered the “worst model”⁷ for the study of ezetimibe, and notably, LDL-C levels were lowered far more in these experiments than they are clinically. Enthusiasm for these animal models should be tempered by interspecies variability in ezetimibe’s “off-target” effects and in the recent failure of other lipid transport drugs in human trials (torcetrapib and ACAT inhibitors) that had shown initial success in animals. No animal model is established for evaluating drugs of ezetimibe’s class, given its complex mechanism of action.

In human studies, the only other surrogate of the net effect of ezetimibe is endothelial function. Among several randomized clinical trials of ezetimibe,^14–18 only one was designed to compare the effects of ezetimibe alone, ezetimibe plus a statin, and a statin by itself in titrated or in maximum doses.¹⁵ After 4 weeks of therapy, all groups had lower LDL-C levels. However, ezetimibe monotherapy and ezetimibe/simvastatin combination therapy had no detectable effect on the arterial response to acetylcholine, but atorvastatin (Lipitor) monotherapy did. To be fair, the other (very small) trials showed mixed results, thus keeping the hypothesis of ezetimibe’s benefit alive, but with nothing close to a clear signal of benefit.

IS ARTERIAL THICKNESS RELIABLE AS A SURROGATE END POINT?

Was the principal problem in ENHANCE the use of carotid intima-media thickness as the primary end point? No.

This issue has received a lot of attention, much of which I believe is misinformed. No trial end point is infallible, including carotid intima-media thickness, and one must remain open to the possibility of chance findings. However, it has been a relatively reasonable end point in trials of diverse cardiovascular preventive strategies, including lipid-lowering, blood-pressure-lowering, and lifestyle interventions and as a directional biomarker of clinical atherosclerotic events.

We should be cautious about comparing data on carotid intima-media thickness from different trials, as Dr. Davidson attempts to do, in view of methodologic and population differences: each trial must be considered independently. Of greatest concern in ENHANCE is the consistency among intima-media thickness end points, including strong trends toward adverse effects in the most diseased carotid and femoral segments.

Moreover, ENHANCE’s detractors contend that the carotid intima-media thickness of the studied population was normal, citing this as evidence of delipidation from prior treatment. Although not impossible (as shown by the work of Zhao and colleagues in the setting of prolonged, intense lipid-lowering therapy¹⁹), at the moment this hypothesis is a matter of conjecture in the ENHANCE participants, particularly because their LDL-C levels were still quite elevated during the trial and conceivably even before randomization.

But these patients were not normal: they were typical patients with familial hypercholesterolemia with extremely elevated LDL-C levels and abnormally thick arteries for their age. Population screening estimates show that, for age and sex, the carotid intima-media thickness values in ENHANCE would lie in the upper quartile of those in the general population.²⁰ Moreover, their mean value is consistent with that in similar-aged groups of patients with familial hypercholesterolemia, even with lower rates of prior statin pretreatment.²¹

The most convincing evidence for the validity of the ENHANCE findings comes from the published subgroup data (Figure 1). In participants whose baseline carotid intima-media thickness was above the median at baseline, the thickness increased more with ezetimibe/simvastatin than with simvastatin alone. The same was true in the subgroup with above-average LDL-C levels at baseline. The subgroups with no prior statin treatment, low-dose prior statin treatment, and high-dose prior statin showed no heterogeneity of response: their carotid intima-media thickness increased more with ezetimibe/simvastatin than with simvastatin alone. None of these differences was statistically significant; however, these prespecified subgroup data seemingly invalidate arguments against the ENHANCE results based on carotid intima-media thickness findings.

In this context, ENHANCE can only be interpreted as a strong initial negative signal, a “red flag” about ezetimibe’s net health benefits.

WHAT NEXT?

The proper present focus of this debate is not on LDL-C but rather on ezetimibe, its unique mechanism of action, and on the need for more evidence about this complex compound.

At present, ezetimibe’s mechanism of action is not fully understood, and its benefit—for now, only mild LDL-C reduction—is too uncertain for us to be spending $5.2 billion a year for it. Its manufacturer is fortunate that the drug is even licensed, given the current and seemingly appropriate regulatory changes under which drugs introducing new therapeutic classes are scrutinized more closely for benefits and risks. “Safe and well tolerated,” as contended by Dr. Davidson, is not nearly enough: drugs must show clinically important benefits. We still know too little about this drug, the manufacturer of which has invested far more in marketing than in science, a point on which Dr. Davidson and I agree.

In 2008, ezetimibe is an appropriate candidate for testing in clinical trials, and in years to come it may be worthy of clinical attention—if rigorous and objectively conducted clinical trials prove its worth. At present, clinical equipoise dictates that ezetimibe is not an appropriate alternative to a statin in titrated doses, to the addition of other lipid-lowering drugs to a statin, to greater attention to drug adherence, or to lifestyle modification.

For the moment, given the ENHANCE results, the clinical usefulness of ezetimibe still remains to be proven. Much more evidence is needed before we can confidently reembrace the clinical use of ezetimibe.