Dimitriy Levin, MD

Clinical question: Which clinical decision rule—Wells rule, simplified Wells rule, revised Geneva score, or simplified revised Geneva score—is the best for evaluating a patient with a possible acute pulmonary embolism?

Background: The use of standardized clinical decision rules to determine the probability of an acute pulmonary embolism (PE) has significantly improved the diagnostic evaluation of patients with suspected PE. Several clinical decision rules are available and widely used, but they have not been previously directly compared.

Study design: Prospective cohort.

Setting: Seven hospitals in the Netherlands.

Synopsis: A total of 807 patients with suspected first episode of acute PE had a sequential workup with clinical probability assessment and D-dimer testing. When PE was considered unlikely according to all four clinical decision rules and a normal D-dimer result, PE was excluded. In the remaining patients, a CT scan was used to confirm or exclude the diagnosis.

The prevalence of PE was 23%. Combined with a normal D-dimer, the decision rules excluded PE in 22% to 24% of patients. Thirty percent of patients had discordant decision rule outcomes, but PE was not detected by CT in any of these patients when combined with a normal D-dimer.

This study has practical limitations because management was based on a combination of four decision rules and D-dimer testing rather than only one rule and D-dimer testing, which is the more realistic clinical approach.

Bottom line: When used correctly and in conjunction with a D-dimer result, the Wells rule, simplified Wells rule, revised Geneva score, and simplified revised Geneva score all perform similarly in the exclusion of acute PE.

Citation: Douma RA, Mos IC, Erkens PM, et al. Performance of 4 clinical decision rules in the diagnostic management of acute pulmonary embolism: a prospective cohort study. Ann Intern Med. 2011;154:709-718.

For more of physician reviews of HM-related literature, check out this month's"In the Literature".

Clinical question: Which clinical decision rule—Wells rule, simplified Wells rule, revised Geneva score, or simplified revised Geneva score—is the best for evaluating a patient with a possible acute pulmonary embolism?

Background: The use of standardized clinical decision rules to determine the probability of an acute pulmonary embolism (PE) has significantly improved the diagnostic evaluation of patients with suspected PE. Several clinical decision rules are available and widely used, but they have not been previously directly compared.

Study design: Prospective cohort.

Setting: Seven hospitals in the Netherlands.

Synopsis: A total of 807 patients with suspected first episode of acute PE had a sequential workup with clinical probability assessment and D-dimer testing. When PE was considered unlikely according to all four clinical decision rules and a normal D-dimer result, PE was excluded. In the remaining patients, a CT scan was used to confirm or exclude the diagnosis.

The prevalence of PE was 23%. Combined with a normal D-dimer, the decision rules excluded PE in 22% to 24% of patients. Thirty percent of patients had discordant decision rule outcomes, but PE was not detected by CT in any of these patients when combined with a normal D-dimer.

This study has practical limitations because management was based on a combination of four decision rules and D-dimer testing rather than only one rule and D-dimer testing, which is the more realistic clinical approach.

Bottom line: When used correctly and in conjunction with a D-dimer result, the Wells rule, simplified Wells rule, revised Geneva score, and simplified revised Geneva score all perform similarly in the exclusion of acute PE.

Citation: Douma RA, Mos IC, Erkens PM, et al. Performance of 4 clinical decision rules in the diagnostic management of acute pulmonary embolism: a prospective cohort study. Ann Intern Med. 2011;154:709-718.

For more of physician reviews of HM-related literature, check out this month's"In the Literature".

Clinical question: Which clinical decision rule—Wells rule, simplified Wells rule, revised Geneva score, or simplified revised Geneva score—is the best for evaluating a patient with a possible acute pulmonary embolism?

Background: The use of standardized clinical decision rules to determine the probability of an acute pulmonary embolism (PE) has significantly improved the diagnostic evaluation of patients with suspected PE. Several clinical decision rules are available and widely used, but they have not been previously directly compared.

Study design: Prospective cohort.

Setting: Seven hospitals in the Netherlands.

Synopsis: A total of 807 patients with suspected first episode of acute PE had a sequential workup with clinical probability assessment and D-dimer testing. When PE was considered unlikely according to all four clinical decision rules and a normal D-dimer result, PE was excluded. In the remaining patients, a CT scan was used to confirm or exclude the diagnosis.

The prevalence of PE was 23%. Combined with a normal D-dimer, the decision rules excluded PE in 22% to 24% of patients. Thirty percent of patients had discordant decision rule outcomes, but PE was not detected by CT in any of these patients when combined with a normal D-dimer.

This study has practical limitations because management was based on a combination of four decision rules and D-dimer testing rather than only one rule and D-dimer testing, which is the more realistic clinical approach.

Bottom line: When used correctly and in conjunction with a D-dimer result, the Wells rule, simplified Wells rule, revised Geneva score, and simplified revised Geneva score all perform similarly in the exclusion of acute PE.

Citation: Douma RA, Mos IC, Erkens PM, et al. Performance of 4 clinical decision rules in the diagnostic management of acute pulmonary embolism: a prospective cohort study. Ann Intern Med. 2011;154:709-718.

For more of physician reviews of HM-related literature, check out this month's"In the Literature".

Clinical question: When do venous thromboembolism (VTE) events occur after cancer surgery?

Background: Cancer is a known risk factor for VTE. Prophylaxis for VTE after cancer surgery is commonly stopped at the time of hospital discharge despite evidence for extended-duration treatment.

Study design: Retrospective cohort.

Setting: Patients reported to the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) database.

Synopsis: The authors examined the records of 46,656 patients who underwent surgery for one of nine specified cancers. Overall VTE rate was 1.6% (1.0% deep venous thrombosis and 0.6% pulmonary embolism), with 33.4% of VTE events occurring after hospital discharge. VTE risk was highest after esophagogastric and hepatopancreaticobiliary surgery, followed by lung, rectum, ovary/uterus, colon, and prostate. Breast and thyroid/parathyroid surgeries had the lowest incidence of VTE. VTE was associated with increased 30-day mortality. Use of VTE prophylaxis during or after hospitalization was not recorded.

Bottom line: Elevated VTE risk persists following hospital discharge after cancer surgery and consideration should be given to extended-duration thromboprophylaxis. Optimal duration of prophylaxis and its risks and benefits remain poorly defined.

Citation: Merkow RP, Bilimoria KY, McCarter MD, et al. Post-discharge venous thromboembolism after cancer surgery: extending the case for extended prophylaxis. Ann Surg. 2011;254:131-137.

For more physician reviews of HM-related literature, visit our website.

Clinical question: When do venous thromboembolism (VTE) events occur after cancer surgery?

Background: Cancer is a known risk factor for VTE. Prophylaxis for VTE after cancer surgery is commonly stopped at the time of hospital discharge despite evidence for extended-duration treatment.

Study design: Retrospective cohort.

Setting: Patients reported to the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) database.

Synopsis: The authors examined the records of 46,656 patients who underwent surgery for one of nine specified cancers. Overall VTE rate was 1.6% (1.0% deep venous thrombosis and 0.6% pulmonary embolism), with 33.4% of VTE events occurring after hospital discharge. VTE risk was highest after esophagogastric and hepatopancreaticobiliary surgery, followed by lung, rectum, ovary/uterus, colon, and prostate. Breast and thyroid/parathyroid surgeries had the lowest incidence of VTE. VTE was associated with increased 30-day mortality. Use of VTE prophylaxis during or after hospitalization was not recorded.

Bottom line: Elevated VTE risk persists following hospital discharge after cancer surgery and consideration should be given to extended-duration thromboprophylaxis. Optimal duration of prophylaxis and its risks and benefits remain poorly defined.

Citation: Merkow RP, Bilimoria KY, McCarter MD, et al. Post-discharge venous thromboembolism after cancer surgery: extending the case for extended prophylaxis. Ann Surg. 2011;254:131-137.

For more physician reviews of HM-related literature, visit our website.

Clinical question: When do venous thromboembolism (VTE) events occur after cancer surgery?

Background: Cancer is a known risk factor for VTE. Prophylaxis for VTE after cancer surgery is commonly stopped at the time of hospital discharge despite evidence for extended-duration treatment.

Study design: Retrospective cohort.

Setting: Patients reported to the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) database.

Synopsis: The authors examined the records of 46,656 patients who underwent surgery for one of nine specified cancers. Overall VTE rate was 1.6% (1.0% deep venous thrombosis and 0.6% pulmonary embolism), with 33.4% of VTE events occurring after hospital discharge. VTE risk was highest after esophagogastric and hepatopancreaticobiliary surgery, followed by lung, rectum, ovary/uterus, colon, and prostate. Breast and thyroid/parathyroid surgeries had the lowest incidence of VTE. VTE was associated with increased 30-day mortality. Use of VTE prophylaxis during or after hospitalization was not recorded.

Bottom line: Elevated VTE risk persists following hospital discharge after cancer surgery and consideration should be given to extended-duration thromboprophylaxis. Optimal duration of prophylaxis and its risks and benefits remain poorly defined.

Citation: Merkow RP, Bilimoria KY, McCarter MD, et al. Post-discharge venous thromboembolism after cancer surgery: extending the case for extended prophylaxis. Ann Surg. 2011;254:131-137.

For more physician reviews of HM-related literature, visit our website.

In This Edition

Literature At A Glance

A guide to this month’s studies

1. Use of a Care-Transitions Intervention Reduces 30-Day Hospital Readmissions
2. Routine Chest Radiographs after Pneumonia to Rule Out Lung Cancer Have Low Diagnostic Yield
3. Hospitalist Care Shifts Costs to the Outpatient Environment
4. Stopping Smoking at Any Time before Surgery Is Safe
5. Hospitalization for Infection Increases Risk of Stroke
6. Longer Duration of Perioperative Antibiotics May Be Beneficial after Cardiac Surgery
7. Stroke Unit Care Combined with Early Supported Discharge Improves Outcomes
8. Criteria May Help Identify Patients at Risk for Infective Endocarditis

Use of a Care-Transitions Intervention Reduces 30-Day Hospital Readmissions

Clinical question: Does use of a specific care-transitions intervention (CTI) reduce 30-day hospital readmissions in a nonintegrated healthcare system?

Background: Previous interventions addressing improved communication between members of the healthcare team, patients, and families at time of discharge show promise for reducing hospital readmissions. Although these interventions revealed positive results, no research has been completed within a system where healthcare is integrated across settings.

Study design: Quasi-experimental prospective cohort study.

Setting: Six Rhode Island acute-care hospitals, including two community hospitals, three teaching hospitals, and a tertiary-care center and teaching hospital. Facilities ranged from 129 beds to 719 beds.

Synopsis: The CTI is a patient-centered intervention occurring across 30 days. The intervention includes a home visit by a coach within three days of hospital discharge, a telephone call within seven to 10 days of discharge, and a final telephone call no later than 30 days after admission. During these contacts, coaches encourage patient and family participation in care, and active communication with their primary-care provider regarding their disease state. A convenience sample of fee-for-service Medicare beneficiaries was identified by admission diagnoses of acute myocardial infarction, congestive heart failure, or specific pulmonary conditions. Overall, 74% participants completed the entire intervention. The odds of a hospital readmission were significantly lower in the intervention population compared with those who did not receive the intervention (OR 0.61; 95% CI, 0.42-0.88).

Study design: Study design was limited by ability to provide coaching (only 8% of total population was approached), and therefore may not be representative of a typical integrated healthcare setting. In addition, the sample consisted of a convenience sample, which may limit generalizability.

Bottom line: The CTI appears to decrease the rate of 30-day hospital readmissions in Medicare patients with certain cardiac and pulmonary diagnoses.

Citation: Voss R, Gardner R, Baier R, Butterfield K, Lehrman S, Gravenstein S. The care transitions intervention: translating from efficacy to effectiveness. Arch Intern Med. 2011;171:1232-1237.

Routine Chest Radiographs after Pneumonia to Rule Out Lung Cancer Have Low Diagnostic Yield

Clinical question: What is the incidence of new lung cancer detected on routine post-pneumonia chest radiographs?

Background: Routine chest radiographs have been recommended four to eight weeks after resolution of pneumonia to exclude underlying lung cancer. The diagnostic yield of this practice is uncertain.

Study design: Population-based cohort.

Setting: Seven emergency departments and six hospitals in Edmonton, Alberta, Canada.

Synopsis: Authors enrolled 3,398 patients with clinical and radiographic evidence of pneumonia. Of these, 59% were aged 50 and older, 52% were male, 17% were current smokers, 18% had COPD, and 49% were treated as inpatients. At 90-day follow-up, 1.1% of patients received a new diagnosis of lung cancer, with incidence steadily increasing to 2.2% at three-year follow-up. In multivariate analysis, age 50 and older, male sex, and current smoking were independent predictors of post-pneumonia new lung cancer diagnosis. Limiting follow-up chest radiographs to patients aged 50 and older would have detected 98% of new lung cancers and improved diagnostic yield to 2.8%.

Bottom line: Routine post-pneumonia chest radiographs for lung cancer screening have low diagnostic yield that is only marginally improved by selecting high-risk populations.

Citation: Tang KL, Eurich DT, Minhas-Sandhu JK, Marrie TJ, Majumdar SR. Incidence, correlates, and chest radiographic yield of new lung cancer diagnosis in 3398 patients with pneumonia. Arch Intern Med. 2011;171:1193-1198.

Hospitalist Care Shifts Costs to the Outpatient Environment

Clinical question: How does hospitalist care affect medical utilization costs after hospital discharge?

Background: The number of patients cared for by hospitalists is growing rapidly. Some studies have shown hospitalists to decrease length of stay and inpatient costs. The impact of shorter hospitalization on outpatient medical utilization and costs is not known.

Study design: Population-based national cohort.

Setting: Hospitalized Medicare patients.

Synopsis: In this study of 58,125 Medicare admissions at 454 hospitals, hospitalist care was associated with a 0.64-day shorter adjusted length of stay and $282 lower hospital charges compared with patients cared for by their primary-care physicians (PCPs). This was offset by $332 higher Medicare spending in the 30 days following hospitalization. Patients cared for by hospitalists were less likely to be discharged home (OR 0.82, 95% CI, 0.78-0.86), and were more likely to require emergency department visits (OR 1.18, 95% CI, 1.12-1.24) and readmissions (OR 1.08, 95% CI, 1.02-1.14). The authors postulate that shorter length of stay associated with hospitalist care is achieved at the expense of shifting costs to the outpatient environment. The discharged patients are sicker and, as a result, require more skilled care and repeat hospital visits.

Bottom line: Hospitalist care may be associated with higher overall costs and more medical utilization.

Citation: Kuo YF, Goodwin JS. Association of hospitalist care with medical utilization after discharge: evidence of cost shift from a cohort study. Ann Intern Med. 2011;155:152-159.

Stopping Smoking at Any Time before Surgery Is Safe

Clinical question: Is smoking cessation within eight weeks of surgery safe?

Background: Smoking cessation before surgery can decrease the risk of surgical complications. However, several studies found increased risk for perioperative complications when smoking was stopped within eight weeks of surgery. These findings created uncertainty about general safety of tobacco cessation counseling before surgery.

Study design: Systematic review and meta-analysis.

Setting: Smokers undergoing any type of surgery.

Synopsis: The authors identified nine studies involving 889 patients that compared smokers who quit within eight weeks of surgery with those who continued to smoke. There was considerable heterogeneity in the studies but no overall difference in perioperative complications between those who quit smoking and those who continued to smoke (OR 0.78, 95% CI, 0.57-1.07). The subset of studies examining pulmonary complications also found no difference (OR 1.18, 95% CI, 0.95-1.46).

Bottom line: Smoking cessation at any time before surgery appears to be safe.

Citation: Myers K, Hajek P, Hinds C, McRobbie H. Stopping smoking shortly before surgery and postoperative complications: a systematic review and meta-analysis. Arch Intern Med. 2011;171:983-989.

Hospitalization for Infection Increases Risk of Stroke

Clinical question: Can infection act as a precipitant for acute ischemic stroke?

Background: Little is known about precipitants of acute ischemic stroke. Severe infections have been shown to promote hypercoagulability and platelet activation, and to induce endothelial dysfunction. Authors postulated that infections severe enough to warrant hospitalization might transiently increase the risk for stroke.

Study design: Case-crossover analysis of data from a multicenter prospective cohort (Cardiovascular Health Study).

Setting: Medicare patients in four communities.

Synopsis: During a median follow-up of 12.2 years, 669 strokes occurred in 5,639 study participants. Hospitalization for infection within 14 days was associated with increased risk of stroke (OR 8.0, 95% CI, 1.6-77.3), and the risk remained elevated for hospitalizations within 90 days (OR 3.4, 95% CI, 1.8-6.5). The findings remained significant after adjusting for comorbidities, including age, sex, race, smoking, and diabetes. The number of patients hospitalized for infection before stroke was small—eight within 14 days, and 29 within 90 days.

Bottom line: Infection severe enough to require hospitalization may act as a trigger for acute ischemic stroke.

Citation: Elkind MS, Carty CL, O’Meara ES, et al. Hospitalization for infection and risk of acute ischemic stroke: the Cardiovascular Health Study. Stroke. 2011;42:1851-1856.

Longer Duration of Perioperative Antibiotics May Be Beneficial after Cardiac Surgery

Clinical question: Is antibiotic prophylaxis for 24 or more hours better than shorter duration of treatment after cardiac surgery?

Background: Sternal surgical site infections are a serious complication of cardiac surgery. The optimal duration of perioperative antibiotic prophylaxis is not known, with recommendations ranging from a single dose to 72 hours. The Society of Thoracic Surgeons’ recommendation for 24 to 72 hours of prophylaxis is not based on a systematic review and meta-analysis.

Study design: Systematic review and meta-analysis.

Setting: Adult patients undergoing open-heart surgery who received perioperative antibiotic prophylaxis.

Synopsis: Authors identified 12 trials encompassing 7,893 patients. Compared with prophylaxis of ≥24 hours, prophylaxis of <24 hours was associated with a higher risk of sternal surgical site infections (RR 1.38, 95% CI, 1.13-1.69) and deep infections (RR 1.68, 95% CI, 1.12-2.53). There was no difference in mortality, other infections, or adverse events. Most studies had methodological limitations with a high risk for bias.

Bottom line: Perioperative antibiotic prophylaxis of ≥24 hours reduces sternal surgical infections.

Citation: Mertz D, Johnstone J, Loeb M. Does duration of perioperative antibiotic prophylaxis matter in cardiac surgery? A systematic review and meta-analysis. Ann Surg. 2011;254:48-54.

Stroke Unit Care Combined with Early Supported Discharge Improves Outcomes

Clinical question: Does early supported discharge (ESD) improve outcomes more than conventional follow-up in stroke patients?

Background: ESD is a mobile team that coordinates follow-up and rehabilitation. Previous studies have shown it to be beneficial in patients with mild to moderate disability at one year, but long-term effects of ESD are not known.

Study design: Randomized controlled trial.

Setting: Single center in Norway.

Synopsis: Stroke-unit patients were recruited and received standard care or ESD after discharge. All 320 patients received standard acute care. The proportion of patients with modified Rankin Score (mRS) of ≤2 was not significantly different in the two groups but identified a trend toward improvement in the intervention group (38% vs. 30%, P=0.106). More patients receiving conventional follow-up died or were institutionalized (P=0.032) but mortality rates at five years were similar (ESD 46% vs. 51%). Secondary outcomes (Scandinavian Stroke Scale, Barthel Index, Frenchay Activity Index, and Mini Mental Status Examination) were not statistically different. Predictors of good outcome in the ESD group included young age, low mRS, and living with others.

This study recruited patients from 1995 to 1997 and followed the patients for five years. Limitations to the applicability include advances in stroke rehabilitation in the last 10 years. The cost of a mobile multidisciplinary team consisting of a physiotherapist, occupational therapist, nurse, and part-time physician was not discussed and may limit the availability to many patients.

Bottom line: Early supported discharge may increase the proportion of patients living at home five years after stroke.

Citation: Fjaertoft H, Rohweder G, Indredavik B. Stroke unit care combined with early supported discharge improves 5-year outcome: a randomized controlled trial. Stroke. 2011;42:1707-1711.

Criteria May Help Identify Patients at Risk for Infective Endocarditis

Clinical question: Which patients with Staphylococcus aureus bacteremia benefit the most from transesophageal echocardiography?

Background: Infective endocarditis is a serious complication of S. aureus bacteremia (SAB), occurring in 5% to 17% of patients with documented SAB. It has been recommended to perform transesophageal echocardiography (TEE) in all patients with SAB. Large variation exists in rates of TEE, and identifying patients at low risk for endocarditis may help with more appropriate utilization of this test.

Study design: Retrospective cohort analysis.

Setting: Two university-based German tertiary hospitals (INSTINCT cohort) and one North American university-based hospital from October 1994 to December 2009 (SABG cohort).

Synopsis: A total of 736 cases of nosocomial SAB were analyzed. Age, source of infection, and 30-day and 90-day case fatality rates were similar between the two cohorts. Patients were followed during the index hospitalization and for three months after discharge.

Patients with infective endocarditis were more likely to have prolonged bacteremia; a permanent intracardiac device, such as a pacemaker or a heart valve; be recipients of hemodialysis; and have osteomyelitis. Of the 83 patients who did not fulfill any of the prediction criteria, no cases of infective endocarditis were found.

Bottom line: A set of simple criteria may help identify patients with nosocomial SAB who are at risk for infective endocarditis. The subset of patients who do not meet any of these criteria may not need diagnostic evaluation with TEE.

Citation: Kaasch, AJ, Fowler VG Jr., Rieg S, et al. Use of a simple criteria set for guiding echocardiography in nosocomial Staphylococcus aureus bacteremia. Clin Infect Dis. 2011;53:1-9.

In This Edition

Literature At A Glance

A guide to this month’s studies

1. Use of a Care-Transitions Intervention Reduces 30-Day Hospital Readmissions
2. Routine Chest Radiographs after Pneumonia to Rule Out Lung Cancer Have Low Diagnostic Yield
3. Hospitalist Care Shifts Costs to the Outpatient Environment
4. Stopping Smoking at Any Time before Surgery Is Safe
5. Hospitalization for Infection Increases Risk of Stroke
6. Longer Duration of Perioperative Antibiotics May Be Beneficial after Cardiac Surgery
7. Stroke Unit Care Combined with Early Supported Discharge Improves Outcomes
8. Criteria May Help Identify Patients at Risk for Infective Endocarditis

Use of a Care-Transitions Intervention Reduces 30-Day Hospital Readmissions

Clinical question: Does use of a specific care-transitions intervention (CTI) reduce 30-day hospital readmissions in a nonintegrated healthcare system?

Background: Previous interventions addressing improved communication between members of the healthcare team, patients, and families at time of discharge show promise for reducing hospital readmissions. Although these interventions revealed positive results, no research has been completed within a system where healthcare is integrated across settings.

Study design: Quasi-experimental prospective cohort study.

Setting: Six Rhode Island acute-care hospitals, including two community hospitals, three teaching hospitals, and a tertiary-care center and teaching hospital. Facilities ranged from 129 beds to 719 beds.

Synopsis: The CTI is a patient-centered intervention occurring across 30 days. The intervention includes a home visit by a coach within three days of hospital discharge, a telephone call within seven to 10 days of discharge, and a final telephone call no later than 30 days after admission. During these contacts, coaches encourage patient and family participation in care, and active communication with their primary-care provider regarding their disease state. A convenience sample of fee-for-service Medicare beneficiaries was identified by admission diagnoses of acute myocardial infarction, congestive heart failure, or specific pulmonary conditions. Overall, 74% participants completed the entire intervention. The odds of a hospital readmission were significantly lower in the intervention population compared with those who did not receive the intervention (OR 0.61; 95% CI, 0.42-0.88).

Study design: Study design was limited by ability to provide coaching (only 8% of total population was approached), and therefore may not be representative of a typical integrated healthcare setting. In addition, the sample consisted of a convenience sample, which may limit generalizability.

Bottom line: The CTI appears to decrease the rate of 30-day hospital readmissions in Medicare patients with certain cardiac and pulmonary diagnoses.

Citation: Voss R, Gardner R, Baier R, Butterfield K, Lehrman S, Gravenstein S. The care transitions intervention: translating from efficacy to effectiveness. Arch Intern Med. 2011;171:1232-1237.

Routine Chest Radiographs after Pneumonia to Rule Out Lung Cancer Have Low Diagnostic Yield

Clinical question: What is the incidence of new lung cancer detected on routine post-pneumonia chest radiographs?

Background: Routine chest radiographs have been recommended four to eight weeks after resolution of pneumonia to exclude underlying lung cancer. The diagnostic yield of this practice is uncertain.

Study design: Population-based cohort.

Setting: Seven emergency departments and six hospitals in Edmonton, Alberta, Canada.

Synopsis: Authors enrolled 3,398 patients with clinical and radiographic evidence of pneumonia. Of these, 59% were aged 50 and older, 52% were male, 17% were current smokers, 18% had COPD, and 49% were treated as inpatients. At 90-day follow-up, 1.1% of patients received a new diagnosis of lung cancer, with incidence steadily increasing to 2.2% at three-year follow-up. In multivariate analysis, age 50 and older, male sex, and current smoking were independent predictors of post-pneumonia new lung cancer diagnosis. Limiting follow-up chest radiographs to patients aged 50 and older would have detected 98% of new lung cancers and improved diagnostic yield to 2.8%.

Bottom line: Routine post-pneumonia chest radiographs for lung cancer screening have low diagnostic yield that is only marginally improved by selecting high-risk populations.

Citation: Tang KL, Eurich DT, Minhas-Sandhu JK, Marrie TJ, Majumdar SR. Incidence, correlates, and chest radiographic yield of new lung cancer diagnosis in 3398 patients with pneumonia. Arch Intern Med. 2011;171:1193-1198.

Hospitalist Care Shifts Costs to the Outpatient Environment

Clinical question: How does hospitalist care affect medical utilization costs after hospital discharge?

Background: The number of patients cared for by hospitalists is growing rapidly. Some studies have shown hospitalists to decrease length of stay and inpatient costs. The impact of shorter hospitalization on outpatient medical utilization and costs is not known.

Study design: Population-based national cohort.

Setting: Hospitalized Medicare patients.

Synopsis: In this study of 58,125 Medicare admissions at 454 hospitals, hospitalist care was associated with a 0.64-day shorter adjusted length of stay and $282 lower hospital charges compared with patients cared for by their primary-care physicians (PCPs). This was offset by $332 higher Medicare spending in the 30 days following hospitalization. Patients cared for by hospitalists were less likely to be discharged home (OR 0.82, 95% CI, 0.78-0.86), and were more likely to require emergency department visits (OR 1.18, 95% CI, 1.12-1.24) and readmissions (OR 1.08, 95% CI, 1.02-1.14). The authors postulate that shorter length of stay associated with hospitalist care is achieved at the expense of shifting costs to the outpatient environment. The discharged patients are sicker and, as a result, require more skilled care and repeat hospital visits.

Bottom line: Hospitalist care may be associated with higher overall costs and more medical utilization.

Citation: Kuo YF, Goodwin JS. Association of hospitalist care with medical utilization after discharge: evidence of cost shift from a cohort study. Ann Intern Med. 2011;155:152-159.

Stopping Smoking at Any Time before Surgery Is Safe

Clinical question: Is smoking cessation within eight weeks of surgery safe?

Background: Smoking cessation before surgery can decrease the risk of surgical complications. However, several studies found increased risk for perioperative complications when smoking was stopped within eight weeks of surgery. These findings created uncertainty about general safety of tobacco cessation counseling before surgery.

Study design: Systematic review and meta-analysis.

Setting: Smokers undergoing any type of surgery.

Synopsis: The authors identified nine studies involving 889 patients that compared smokers who quit within eight weeks of surgery with those who continued to smoke. There was considerable heterogeneity in the studies but no overall difference in perioperative complications between those who quit smoking and those who continued to smoke (OR 0.78, 95% CI, 0.57-1.07). The subset of studies examining pulmonary complications also found no difference (OR 1.18, 95% CI, 0.95-1.46).

Bottom line: Smoking cessation at any time before surgery appears to be safe.

Citation: Myers K, Hajek P, Hinds C, McRobbie H. Stopping smoking shortly before surgery and postoperative complications: a systematic review and meta-analysis. Arch Intern Med. 2011;171:983-989.

Hospitalization for Infection Increases Risk of Stroke

Clinical question: Can infection act as a precipitant for acute ischemic stroke?

Background: Little is known about precipitants of acute ischemic stroke. Severe infections have been shown to promote hypercoagulability and platelet activation, and to induce endothelial dysfunction. Authors postulated that infections severe enough to warrant hospitalization might transiently increase the risk for stroke.

Study design: Case-crossover analysis of data from a multicenter prospective cohort (Cardiovascular Health Study).

Setting: Medicare patients in four communities.

Synopsis: During a median follow-up of 12.2 years, 669 strokes occurred in 5,639 study participants. Hospitalization for infection within 14 days was associated with increased risk of stroke (OR 8.0, 95% CI, 1.6-77.3), and the risk remained elevated for hospitalizations within 90 days (OR 3.4, 95% CI, 1.8-6.5). The findings remained significant after adjusting for comorbidities, including age, sex, race, smoking, and diabetes. The number of patients hospitalized for infection before stroke was small—eight within 14 days, and 29 within 90 days.

Bottom line: Infection severe enough to require hospitalization may act as a trigger for acute ischemic stroke.

Citation: Elkind MS, Carty CL, O’Meara ES, et al. Hospitalization for infection and risk of acute ischemic stroke: the Cardiovascular Health Study. Stroke. 2011;42:1851-1856.

Longer Duration of Perioperative Antibiotics May Be Beneficial after Cardiac Surgery

Clinical question: Is antibiotic prophylaxis for 24 or more hours better than shorter duration of treatment after cardiac surgery?

Background: Sternal surgical site infections are a serious complication of cardiac surgery. The optimal duration of perioperative antibiotic prophylaxis is not known, with recommendations ranging from a single dose to 72 hours. The Society of Thoracic Surgeons’ recommendation for 24 to 72 hours of prophylaxis is not based on a systematic review and meta-analysis.

Study design: Systematic review and meta-analysis.

Setting: Adult patients undergoing open-heart surgery who received perioperative antibiotic prophylaxis.

Synopsis: Authors identified 12 trials encompassing 7,893 patients. Compared with prophylaxis of ≥24 hours, prophylaxis of <24 hours was associated with a higher risk of sternal surgical site infections (RR 1.38, 95% CI, 1.13-1.69) and deep infections (RR 1.68, 95% CI, 1.12-2.53). There was no difference in mortality, other infections, or adverse events. Most studies had methodological limitations with a high risk for bias.

Bottom line: Perioperative antibiotic prophylaxis of ≥24 hours reduces sternal surgical infections.

Citation: Mertz D, Johnstone J, Loeb M. Does duration of perioperative antibiotic prophylaxis matter in cardiac surgery? A systematic review and meta-analysis. Ann Surg. 2011;254:48-54.

Stroke Unit Care Combined with Early Supported Discharge Improves Outcomes

Clinical question: Does early supported discharge (ESD) improve outcomes more than conventional follow-up in stroke patients?

Background: ESD is a mobile team that coordinates follow-up and rehabilitation. Previous studies have shown it to be beneficial in patients with mild to moderate disability at one year, but long-term effects of ESD are not known.

Study design: Randomized controlled trial.

Setting: Single center in Norway.

Synopsis: Stroke-unit patients were recruited and received standard care or ESD after discharge. All 320 patients received standard acute care. The proportion of patients with modified Rankin Score (mRS) of ≤2 was not significantly different in the two groups but identified a trend toward improvement in the intervention group (38% vs. 30%, P=0.106). More patients receiving conventional follow-up died or were institutionalized (P=0.032) but mortality rates at five years were similar (ESD 46% vs. 51%). Secondary outcomes (Scandinavian Stroke Scale, Barthel Index, Frenchay Activity Index, and Mini Mental Status Examination) were not statistically different. Predictors of good outcome in the ESD group included young age, low mRS, and living with others.

This study recruited patients from 1995 to 1997 and followed the patients for five years. Limitations to the applicability include advances in stroke rehabilitation in the last 10 years. The cost of a mobile multidisciplinary team consisting of a physiotherapist, occupational therapist, nurse, and part-time physician was not discussed and may limit the availability to many patients.

Bottom line: Early supported discharge may increase the proportion of patients living at home five years after stroke.

Citation: Fjaertoft H, Rohweder G, Indredavik B. Stroke unit care combined with early supported discharge improves 5-year outcome: a randomized controlled trial. Stroke. 2011;42:1707-1711.

Criteria May Help Identify Patients at Risk for Infective Endocarditis

Clinical question: Which patients with Staphylococcus aureus bacteremia benefit the most from transesophageal echocardiography?

Background: Infective endocarditis is a serious complication of S. aureus bacteremia (SAB), occurring in 5% to 17% of patients with documented SAB. It has been recommended to perform transesophageal echocardiography (TEE) in all patients with SAB. Large variation exists in rates of TEE, and identifying patients at low risk for endocarditis may help with more appropriate utilization of this test.

Study design: Retrospective cohort analysis.

Setting: Two university-based German tertiary hospitals (INSTINCT cohort) and one North American university-based hospital from October 1994 to December 2009 (SABG cohort).

Synopsis: A total of 736 cases of nosocomial SAB were analyzed. Age, source of infection, and 30-day and 90-day case fatality rates were similar between the two cohorts. Patients were followed during the index hospitalization and for three months after discharge.

Patients with infective endocarditis were more likely to have prolonged bacteremia; a permanent intracardiac device, such as a pacemaker or a heart valve; be recipients of hemodialysis; and have osteomyelitis. Of the 83 patients who did not fulfill any of the prediction criteria, no cases of infective endocarditis were found.

Bottom line: A set of simple criteria may help identify patients with nosocomial SAB who are at risk for infective endocarditis. The subset of patients who do not meet any of these criteria may not need diagnostic evaluation with TEE.

Citation: Kaasch, AJ, Fowler VG Jr., Rieg S, et al. Use of a simple criteria set for guiding echocardiography in nosocomial Staphylococcus aureus bacteremia. Clin Infect Dis. 2011;53:1-9.

In This Edition

Literature At A Glance

A guide to this month’s studies

1. Use of a Care-Transitions Intervention Reduces 30-Day Hospital Readmissions
2. Routine Chest Radiographs after Pneumonia to Rule Out Lung Cancer Have Low Diagnostic Yield
3. Hospitalist Care Shifts Costs to the Outpatient Environment
4. Stopping Smoking at Any Time before Surgery Is Safe
5. Hospitalization for Infection Increases Risk of Stroke
6. Longer Duration of Perioperative Antibiotics May Be Beneficial after Cardiac Surgery
7. Stroke Unit Care Combined with Early Supported Discharge Improves Outcomes
8. Criteria May Help Identify Patients at Risk for Infective Endocarditis

Use of a Care-Transitions Intervention Reduces 30-Day Hospital Readmissions

Clinical question: Does use of a specific care-transitions intervention (CTI) reduce 30-day hospital readmissions in a nonintegrated healthcare system?

Background: Previous interventions addressing improved communication between members of the healthcare team, patients, and families at time of discharge show promise for reducing hospital readmissions. Although these interventions revealed positive results, no research has been completed within a system where healthcare is integrated across settings.

Study design: Quasi-experimental prospective cohort study.

Setting: Six Rhode Island acute-care hospitals, including two community hospitals, three teaching hospitals, and a tertiary-care center and teaching hospital. Facilities ranged from 129 beds to 719 beds.

Synopsis: The CTI is a patient-centered intervention occurring across 30 days. The intervention includes a home visit by a coach within three days of hospital discharge, a telephone call within seven to 10 days of discharge, and a final telephone call no later than 30 days after admission. During these contacts, coaches encourage patient and family participation in care, and active communication with their primary-care provider regarding their disease state. A convenience sample of fee-for-service Medicare beneficiaries was identified by admission diagnoses of acute myocardial infarction, congestive heart failure, or specific pulmonary conditions. Overall, 74% participants completed the entire intervention. The odds of a hospital readmission were significantly lower in the intervention population compared with those who did not receive the intervention (OR 0.61; 95% CI, 0.42-0.88).

Study design: Study design was limited by ability to provide coaching (only 8% of total population was approached), and therefore may not be representative of a typical integrated healthcare setting. In addition, the sample consisted of a convenience sample, which may limit generalizability.

Bottom line: The CTI appears to decrease the rate of 30-day hospital readmissions in Medicare patients with certain cardiac and pulmonary diagnoses.

Citation: Voss R, Gardner R, Baier R, Butterfield K, Lehrman S, Gravenstein S. The care transitions intervention: translating from efficacy to effectiveness. Arch Intern Med. 2011;171:1232-1237.

Routine Chest Radiographs after Pneumonia to Rule Out Lung Cancer Have Low Diagnostic Yield

Clinical question: What is the incidence of new lung cancer detected on routine post-pneumonia chest radiographs?

Background: Routine chest radiographs have been recommended four to eight weeks after resolution of pneumonia to exclude underlying lung cancer. The diagnostic yield of this practice is uncertain.

Study design: Population-based cohort.

Setting: Seven emergency departments and six hospitals in Edmonton, Alberta, Canada.

Synopsis: Authors enrolled 3,398 patients with clinical and radiographic evidence of pneumonia. Of these, 59% were aged 50 and older, 52% were male, 17% were current smokers, 18% had COPD, and 49% were treated as inpatients. At 90-day follow-up, 1.1% of patients received a new diagnosis of lung cancer, with incidence steadily increasing to 2.2% at three-year follow-up. In multivariate analysis, age 50 and older, male sex, and current smoking were independent predictors of post-pneumonia new lung cancer diagnosis. Limiting follow-up chest radiographs to patients aged 50 and older would have detected 98% of new lung cancers and improved diagnostic yield to 2.8%.

Bottom line: Routine post-pneumonia chest radiographs for lung cancer screening have low diagnostic yield that is only marginally improved by selecting high-risk populations.

Citation: Tang KL, Eurich DT, Minhas-Sandhu JK, Marrie TJ, Majumdar SR. Incidence, correlates, and chest radiographic yield of new lung cancer diagnosis in 3398 patients with pneumonia. Arch Intern Med. 2011;171:1193-1198.

Hospitalist Care Shifts Costs to the Outpatient Environment

Clinical question: How does hospitalist care affect medical utilization costs after hospital discharge?

Background: The number of patients cared for by hospitalists is growing rapidly. Some studies have shown hospitalists to decrease length of stay and inpatient costs. The impact of shorter hospitalization on outpatient medical utilization and costs is not known.

Study design: Population-based national cohort.

Setting: Hospitalized Medicare patients.

Synopsis: In this study of 58,125 Medicare admissions at 454 hospitals, hospitalist care was associated with a 0.64-day shorter adjusted length of stay and $282 lower hospital charges compared with patients cared for by their primary-care physicians (PCPs). This was offset by $332 higher Medicare spending in the 30 days following hospitalization. Patients cared for by hospitalists were less likely to be discharged home (OR 0.82, 95% CI, 0.78-0.86), and were more likely to require emergency department visits (OR 1.18, 95% CI, 1.12-1.24) and readmissions (OR 1.08, 95% CI, 1.02-1.14). The authors postulate that shorter length of stay associated with hospitalist care is achieved at the expense of shifting costs to the outpatient environment. The discharged patients are sicker and, as a result, require more skilled care and repeat hospital visits.

Bottom line: Hospitalist care may be associated with higher overall costs and more medical utilization.

Citation: Kuo YF, Goodwin JS. Association of hospitalist care with medical utilization after discharge: evidence of cost shift from a cohort study. Ann Intern Med. 2011;155:152-159.

Stopping Smoking at Any Time before Surgery Is Safe

Clinical question: Is smoking cessation within eight weeks of surgery safe?

Background: Smoking cessation before surgery can decrease the risk of surgical complications. However, several studies found increased risk for perioperative complications when smoking was stopped within eight weeks of surgery. These findings created uncertainty about general safety of tobacco cessation counseling before surgery.

Study design: Systematic review and meta-analysis.

Setting: Smokers undergoing any type of surgery.

Synopsis: The authors identified nine studies involving 889 patients that compared smokers who quit within eight weeks of surgery with those who continued to smoke. There was considerable heterogeneity in the studies but no overall difference in perioperative complications between those who quit smoking and those who continued to smoke (OR 0.78, 95% CI, 0.57-1.07). The subset of studies examining pulmonary complications also found no difference (OR 1.18, 95% CI, 0.95-1.46).

Bottom line: Smoking cessation at any time before surgery appears to be safe.

Citation: Myers K, Hajek P, Hinds C, McRobbie H. Stopping smoking shortly before surgery and postoperative complications: a systematic review and meta-analysis. Arch Intern Med. 2011;171:983-989.

Hospitalization for Infection Increases Risk of Stroke

Clinical question: Can infection act as a precipitant for acute ischemic stroke?

Background: Little is known about precipitants of acute ischemic stroke. Severe infections have been shown to promote hypercoagulability and platelet activation, and to induce endothelial dysfunction. Authors postulated that infections severe enough to warrant hospitalization might transiently increase the risk for stroke.

Study design: Case-crossover analysis of data from a multicenter prospective cohort (Cardiovascular Health Study).

Setting: Medicare patients in four communities.

Synopsis: During a median follow-up of 12.2 years, 669 strokes occurred in 5,639 study participants. Hospitalization for infection within 14 days was associated with increased risk of stroke (OR 8.0, 95% CI, 1.6-77.3), and the risk remained elevated for hospitalizations within 90 days (OR 3.4, 95% CI, 1.8-6.5). The findings remained significant after adjusting for comorbidities, including age, sex, race, smoking, and diabetes. The number of patients hospitalized for infection before stroke was small—eight within 14 days, and 29 within 90 days.

Bottom line: Infection severe enough to require hospitalization may act as a trigger for acute ischemic stroke.

Citation: Elkind MS, Carty CL, O’Meara ES, et al. Hospitalization for infection and risk of acute ischemic stroke: the Cardiovascular Health Study. Stroke. 2011;42:1851-1856.

Longer Duration of Perioperative Antibiotics May Be Beneficial after Cardiac Surgery

Clinical question: Is antibiotic prophylaxis for 24 or more hours better than shorter duration of treatment after cardiac surgery?

Background: Sternal surgical site infections are a serious complication of cardiac surgery. The optimal duration of perioperative antibiotic prophylaxis is not known, with recommendations ranging from a single dose to 72 hours. The Society of Thoracic Surgeons’ recommendation for 24 to 72 hours of prophylaxis is not based on a systematic review and meta-analysis.

Study design: Systematic review and meta-analysis.

Setting: Adult patients undergoing open-heart surgery who received perioperative antibiotic prophylaxis.

Synopsis: Authors identified 12 trials encompassing 7,893 patients. Compared with prophylaxis of ≥24 hours, prophylaxis of <24 hours was associated with a higher risk of sternal surgical site infections (RR 1.38, 95% CI, 1.13-1.69) and deep infections (RR 1.68, 95% CI, 1.12-2.53). There was no difference in mortality, other infections, or adverse events. Most studies had methodological limitations with a high risk for bias.

Bottom line: Perioperative antibiotic prophylaxis of ≥24 hours reduces sternal surgical infections.

Citation: Mertz D, Johnstone J, Loeb M. Does duration of perioperative antibiotic prophylaxis matter in cardiac surgery? A systematic review and meta-analysis. Ann Surg. 2011;254:48-54.

Stroke Unit Care Combined with Early Supported Discharge Improves Outcomes

Clinical question: Does early supported discharge (ESD) improve outcomes more than conventional follow-up in stroke patients?

Background: ESD is a mobile team that coordinates follow-up and rehabilitation. Previous studies have shown it to be beneficial in patients with mild to moderate disability at one year, but long-term effects of ESD are not known.

Study design: Randomized controlled trial.

Setting: Single center in Norway.

Synopsis: Stroke-unit patients were recruited and received standard care or ESD after discharge. All 320 patients received standard acute care. The proportion of patients with modified Rankin Score (mRS) of ≤2 was not significantly different in the two groups but identified a trend toward improvement in the intervention group (38% vs. 30%, P=0.106). More patients receiving conventional follow-up died or were institutionalized (P=0.032) but mortality rates at five years were similar (ESD 46% vs. 51%). Secondary outcomes (Scandinavian Stroke Scale, Barthel Index, Frenchay Activity Index, and Mini Mental Status Examination) were not statistically different. Predictors of good outcome in the ESD group included young age, low mRS, and living with others.

This study recruited patients from 1995 to 1997 and followed the patients for five years. Limitations to the applicability include advances in stroke rehabilitation in the last 10 years. The cost of a mobile multidisciplinary team consisting of a physiotherapist, occupational therapist, nurse, and part-time physician was not discussed and may limit the availability to many patients.

Bottom line: Early supported discharge may increase the proportion of patients living at home five years after stroke.

Citation: Fjaertoft H, Rohweder G, Indredavik B. Stroke unit care combined with early supported discharge improves 5-year outcome: a randomized controlled trial. Stroke. 2011;42:1707-1711.

Criteria May Help Identify Patients at Risk for Infective Endocarditis

Clinical question: Which patients with Staphylococcus aureus bacteremia benefit the most from transesophageal echocardiography?

Background: Infective endocarditis is a serious complication of S. aureus bacteremia (SAB), occurring in 5% to 17% of patients with documented SAB. It has been recommended to perform transesophageal echocardiography (TEE) in all patients with SAB. Large variation exists in rates of TEE, and identifying patients at low risk for endocarditis may help with more appropriate utilization of this test.

Study design: Retrospective cohort analysis.

Setting: Two university-based German tertiary hospitals (INSTINCT cohort) and one North American university-based hospital from October 1994 to December 2009 (SABG cohort).

Synopsis: A total of 736 cases of nosocomial SAB were analyzed. Age, source of infection, and 30-day and 90-day case fatality rates were similar between the two cohorts. Patients were followed during the index hospitalization and for three months after discharge.

Patients with infective endocarditis were more likely to have prolonged bacteremia; a permanent intracardiac device, such as a pacemaker or a heart valve; be recipients of hemodialysis; and have osteomyelitis. Of the 83 patients who did not fulfill any of the prediction criteria, no cases of infective endocarditis were found.

Bottom line: A set of simple criteria may help identify patients with nosocomial SAB who are at risk for infective endocarditis. The subset of patients who do not meet any of these criteria may not need diagnostic evaluation with TEE.

Citation: Kaasch, AJ, Fowler VG Jr., Rieg S, et al. Use of a simple criteria set for guiding echocardiography in nosocomial Staphylococcus aureus bacteremia. Clin Infect Dis. 2011;53:1-9.

In This Edition

Literature at a Glance

A guide to this month’s studies

• Risk factors for iatrogenic pneumothorax
• Residency acceptance and use of pharmaceutical industry funding
• Early cholecystectomy outcomes for gallstone pancreatitis
• Use of microbial DNA in sepsis
• Adding rifampicin to vancomycin in MRSA pneumonia
• Rate and outcomes of culture-negative severe sepsis
• Rates of surgical comanagement in U.S. hospitals
• Probiotics and rates of ventilator-associated pneumonia

Ultrasound Guidance and Operator Experience Decrease Risk of Pneumothorax Following Thoracentesis

Clinical question: How often does pneumothorax happen following thoracentesis, and what factors are associated with increased risk of this complication?

Background: Procedural complications are an important source of adverse events in the hospital. Iatrogenic pneumothorax after thoracentesis results in increased hospital length of stay, morbidity, and mortality. Large variation exists in reported pneumothorax rates, and little is known about procedure- and patient-specific factors associated with development of this complication.

Study design: Systematic review and meta-analysis.

Setting: Review of 24 MEDLINE-indexed studies from January 1966 to April 2009.

Synopsis: A total of 349 pneumothoraces were reported after 6,605 thoracenteses (overall incidence 6.0%). Chest-tube insertion was required in 34.1% of the cases. Risk for pneumothorax was significantly higher when larger needles or catheters were used compared with needles smaller than 20-gauge (odds ratio 2.5, 95% confidence interval [CI], 1.1-6.0) and after therapeutic thoracentesis compared with diagnostic procedures (OR 2.6, 95% CI, 1.8-3.8).

Procedures requiring two or more needle passes did not significantly increase pneumothorax risk (OR 2.5, 95% CI, 0.3-20.1). In contrast, pneumothorax rates were significantly lower when using ultrasound guidance (OR 0.3, 95% CI, 0.2-0.7) and with experienced operators (3.9% vs. 8.5%, P=0.04).

Examining patient risk factors, pneumothorax rates were similar regardless of effusion size and patient gender. Additionally, rates were similar among non-ICU inpatients, ICU inpatients, and outpatients. Data did show a trend toward increased risk of pneumothorax with mechanical ventilation (OR 4.0, 95% CI, 0.95-16.8), although no study directly compared rates in ICU patients with and without mechanical ventilation.

Bottom line: Ultrasound guidance is a modifiable factor that decreases the risk of post-thoracentesis pneumothorax. Pneumothorax rates are lower when performed by experienced clinicians, providing an important opportunity to reduce procedure-related complications by increasing direct trainee supervision.

Citation: Gordon CE, Feller-Kopman D, Balk EM, Smetana GW. Pneumothorax following thoracentesis: a systematic review and meta-analysis. Arch Intern Med. 2010;170(4):332-339.

Pharmaceutical Industry Support Is Common in U.S. Internal-Medicine Residency Programs

Clinical question: What are the current attitudes of program directors regarding pharmaceutical industry support of internal-medicine residency activities? What are the potential associations between program characteristics and acceptance of industry support?

Background: Increasing evidence suggests that interactions with the pharmaceutical industry influence physician attitudes and practices. Recently, the Association of American Medical Colleges (AAMC) proposed that academic medical centers prohibit the acceptance of all gifts and restrict access by pharmaceutical industry representatives.

Study design: Survey of U.S. internal-medicine residency program directors.

Setting: Web-based survey of residency program directors in 388 U.S. internal-medicine residency programs.

Synopsis: Of the 236 program directors responding to the survey, 132 (55.9%) reported accepting some kind of support from the pharmaceutical industry. Support was most commonly provided in the form of food for conferences (90.9%), educational materials (83.3%), office supplies (68.9%), and drug samples (57.6%).

When programs reported accepting pharmaceutical industry support, 67.9% cited a lack of other funding sources as the reason for acceptance. Only 22.7% of programs with a program director who thinks pharmaceutical support is unacceptable actually accepted industry support. The likelihood of accepting support was associated with location in the Southern U.S. and was inversely associated with the three-year rolling American Board of Internal Medicine (ABIM) pass rates (each 1% decrease in the pass rate was associated with a 21% increase in the odds of accepting pharmaceutical industry support).

Bottom line: While most program directors did not find pharmaceutical industry support desirable, more than half reported acceptance of such support, with most citing lack of other funding resources as the reason for acceptance.

Citation: Loertscher LL, Halvorsen AJ, Beasley BW, Holmboe ES, Kolars JC, McDonald FS. Pharmaceutical industry support and residency education: a survey of internal medicine program directors. Arch Intern Med. 2010;170(4):356-362.

Early Cholecystectomy Safely Decreases Hospital Stay in Patients with Mild Gallstone Pancreatitis

Clinical question: Can laparoscopic cholecystectomy performed within 48 hours of admission for mild gallstone pancreatitis reduce hospital length of stay without increasing perioperative complications?

Background: Although there is a clear consensus that patients who present with gallstone pancreatitis should undergo cholecystectomy to prevent recurrence, precise timing of surgery remains controversial.

Study design: Randomized prospective trial.

Setting: Harbor-UCLA Medical Center, a Los Angeles County public teaching hospital and Level I trauma center.

Synopsis: Patients were prospectively randomized to an early group and a control group. Inclusion criteria consisted of adults from the ages of 18 to 100 with mild gallstone pancreatitis and three or fewer Ranson criteria. The primary endpoint was length of hospital stay. The secondary endpoint was a composite of complications, including the need for conversion to open cholecystectomy, readmission within 30 days, bleeding requiring transfusion, bile duct injury, or wound infection.

The study was terminated after 50 patients, as there was a difference in the length of hospital stay with a predefined alpha level of 0.005. Patients in the early group were taken to the operating room at a mean of 35.1 hours after admission, compared with 77.8 hours in the control group. The overall length of hospital stay was shorter in the early group (mean 3.5 days, 95% CI, 2.7-4.3), compared with the control group (mean 5.8, 95% CI, 3.8-7.9). All cholecystectomies were completed laparoscopically, without conversion to open. No statistically significant difference existed in secondary endpoints (P=0.48, OR 1.66, 95% CI, 0.41-6.78).

Bottom line: Laparoscopic cholecystectomy performed within 48 hours of admission, irrespective of normalization of laboratory values or clinical progress, safely decreases the overall length of stay, compared with delaying laparoscopic cholecystectomy until laboratory values and clinical condition normalize.

Citation: Aboulian A, Chan T, Yaghoubian A, et al. Early cholecystectomy safely decreases hospital stay in patients with mild gallstone pancreatitis: a randomized prospective study. Ann Surg. 2010;251(4): 615-619.

Presence of Microbial DNA in Blood Correlates with Disease Severity

Clinical question: Is the presence of microbial DNA in the blood associated with disease severity in severe sepsis, and how does detection of this microbial DNA by polymerase chain reaction (PCR) compare with blood cultures (BC)?

Background: Inadequate antibiotic therapy is a strong and independent predictor of poor outcomes in sepsis. Diagnostic uncertainty regarding the causative micro-organism is compensated for by liberal use of broad-spectrum antibiotics. As a result, resistance to antibiotics is an increasing public-health problem.

Study design: Prospective multicenter controlled observational study.

Setting: Three ICUs in Germany and France.

Synopsis: From 2005 to 2007, 63 patients were enrolled in the control group and 142 patients were enrolled in the sepsis group. In control patients, blood cultures and specimens were drawn daily at a maximum of three days after admission. In the sepsis group, blood samples were obtained on the day severe sepsis was suspected. Consecutive samples for the next two days after study inclusion were taken.

Taking BC as the laboratory comparison method, the sensitivity of PCR to detect culture-positive bacteremia in sepsis was 0.80 with a specificity of 0.77. PCR detected 29 of 41 microorganisms (70.3%) found in the BC. The highest recovery rate was observed for gram-negative bacteria (78.6%), fungi (50.0%), and gram-positive bacteria (47.6%). PCR from septic patients correlated well with markers of host response (IL-6 and PCT) and disease severity (SOFA score), even when the BC remained negative.

The appropriateness of antimicrobial therapy based on culture-based methods was not recorded, so it’s impossible to conclude whether or not the PCR would have contributed to a more effective therapy.

Bottom line: Concordance between BC and PCR is moderate in septic patients. PCR-based pathogen detection correlated with disease severity even if the BC remained negative, suggesting that the presence of microbial DNA in the bloodstream is a clinically significant event.

Citation: Bloos F, Hinder F, Becker K, et al. A multicenter trial to compare blood culture with polymerase chain reaction in severe human sepsis. Intensive Care Med. 2010;36(2):241-247.

Adding Rifampicin to Vancomycin Improves Outcomes in MRSA Pneumonia

Clinical question: Does adding rifampicin to vancomycin improve outcomes in patients with hospital-acquired MRSA pneumonia?

Background: Hospital-acquired MRSA pneumonia has a mortality of more than 20%. Vancomycin penetrates the lung tissue poorly. The value of adding rifampicin, an antibiotic with broad-spectrum coverage and good tissue penetration, was investigated.

Study design: Randomized open-label trial.

Setting: Medical ICU patients at Ulsan College of Medicine, Asan Medical Center, South Korea.

Synopsis: Patients older than 18 years of age with clinical symptoms suggestive of nosocomial pneumonia were randomized to receive vancomycin alone (V) or vancomycin plus rifampicin (VR). Clinicians could add additional antibiotics for gram-negative coverage as needed.

Of the 183 patients screened, 93 met the inclusion criteria and were randomized in a 1:1 ratio. MRSA infection was microbiologically confirmed. Clinical cure rate in VR patients was significantly greater at day 14 compared with the V group (53.7% vs. 31.0%, P=0.047) based on a modified intention-to-treat model. The overall mortality at day 28 did not significantly differ between the groups (22.0% vs. 38.1%, P=0.15), although the 60-day mortality was lower in the VR group (26.8% vs. 50.0%, P=0.042). Mortality from MRSA pneumonia had a trend toward a decrease in the VR group (14.7% vs. 28.6%, P=0.18).

The trial was limited because it was a single-site study and lacked statistical power to assess certain outcomes. Additionally, treatment protocols were not compared with other antimicrobial therapies.

Bottom line: Vancomycin plus rifampicin improves MRSA pneumonia outcomes in ICU patients.

Citation: Jung YJ, Koh Y, Hong SB, et al. Effect of vancomycin plus rifampicin in the treatment of nosocomial MRSA pneumonia. Crit Care Med. 2010;38(1):175-180.

Severe Sepsis Syndromes Are Not Always Caused by Bacteremia

Clinical question: What are the common causes of clinical sepsis?

Background: When sepsis is defined by systemic inflammatory response syndrome (SIRS) criteria, the etiology is not always infectious. Rapid initiation of antimicrobial therapy for infectious SIRS is a priority, but it could result in treating a significant number of patients who are not bacteremic.

Study design: Prospective secondary analysis of a registry of patients created to evaluate an institutional standard-of-care protocol.

Setting: Urban, 850-bed, tertiary-care teaching institution in North Carolina.

Synopsis: ED cases meeting the criteria for severe sepsis underwent a secondary review that looked at the cause of the sepsis. Only 45% of patients identified as having severe sepsis were blood-culture-positive during that episode of care. The culture-positive group was more likely to have central lines, malignancies, or reside in a nursing home.

Of the subgroup of culture-negative patients, 52% had another infectious etiology, most commonly pneumonia. Other “noninfectious mimics,” including inflammatory colitis, myocardial infarction, and pulmonary embolism, were noted in 32% of patients in the subgroup, and the cause was not identified in 16% of the patients.

In-hospital mortality was higher in the culture-positive group than in the culture-negative group (25% vs. 4%, P=0.05). There was no evidence of harm in patients with culture-negative sepsis treated for a systemic infection.

Bottom line: Many patients with a clinical picture of severe sepsis will not have positive blood cultures or an infectious etiology.

Citation: Heffner AC, Horton JM, Marchick MR, Jones AE. Etiology of illness in patients with severe sepsis admitted to the hospital from the emergency department. Clin Infect Dis. 2010;50(6):814-820.

Comanagement of Surgical Inpatients by Hospitalists Is Rapidly Expanding

Clinical question: What is the prevalence and nature of comanagement of surgical patients by medicine physicians?

Background: Comanagement of surgical patients is a common clinical role for hospitalists, but the relationship is not well characterized in the literature in terms of numbers of patients or types of physicians involved in this practice.

Study design: Retrospective cohort.

Setting: Cross-section of hospitals from a Medicare database.

Synopsis: During the study period, 35.2% of patients were comanaged by a medicine physician—23.7% by a generalist and 14% by a subspecialist. Cardiothoracic surgery patients were more likely to be comanaged by a subspecialist, whereas all other patients were more likely to be comanaged by a generalist.

Although subspecialist comanagement actually declined during the study period, overall comanagement increased from 33.3% in 1996 to 40.8% in 2006. This increase is entirely attributable to the increase in comanagement by hospitalists. Most of this growth occurred with orthopedic patients.

Patient factors associated with comanagement include advanced age, emergency admissions, and increasing comorbidities. Teaching hospitals had less comanagement, while midsize, nonteaching, and for-profit hospitals had more comanagement.

Bottom line: Comanagement of surgical patients by medicine physicians is a common and growing clinical relationship. Hospitalists are responsible for increasing numbers of comanaged surgical patients.

Citation: Sharma G, Kuo YF, Freeman J, Zhang DD, Goodwin JS. Comanagement of hospitalized surgical patients by medicine physicians in the United States. Arch Intern Med. 2010;170(4):363-368.

Probiotics Might Decrease Risk of Ventilator-Associated Pneumonia

Clinical question: Does the administration of probiotics decrease the incidence of ventilator-associated pneumonia in critically ill patients?

Background: Ventilator-associated pneumonia (VAP) is a major nosocomial infection in ICUs. Probiotics are thought to decrease colonization and, therefore, infection with serious hospital-acquired pathogens.

Study design: Meta-analysis of five randomized controlled trials.

Setting: ICU patients on mechanical ventilation for at least 24 hours.

Synopsis: Five trials met the inclusion criteria of comparing probiotics to placebo in critically ill patients on mechanical ventilation and reporting the outcome of VAP. Administration of probiotics decreased the incidence of VAP (odds ratio 0.61, 95% CI, 0.41-0.91) and colonization of the respiratory tract with Pseudomonas aeruginosa (OR 0.35, 95% CI, 0.13-0.93).

Length of ICU stay was decreased in the probiotic arm, although this effect was not statistically significant in all analyses. Probiotics had no effect on such outcomes as ICU mortality, in-hospital mortality, or duration of mechanical ventilation.

Bottom line: Probiotics might be an effective strategy to reduce the risk of VAP, even if they do not appear to impact such outcomes as mortality.

Citation: Siempos II, Ntaidou TK, Falagas ME. Impact of the administration of probiotics on the incidence of ventilator-associated pneumonia: a meta-analysis of randomized controlled trials. Crit Care Med. 2010;38(3):954-962. TH

PEDIATRIC HM LITERATURE

By Mark Shen, MD

Renal Ultrasound Identifies Children with High-Grade Vesicoureteral Reflux

Reviewed by Pediatric Editor Mark Shen, MD, medical director of hospital medicine at Dell Children’s Medical Center, Austin, Texas.

Clinical question: What is the diagnostic accuracy of specific renal ultrasound (US) criterion for detection of vesicoureteral reflux (VUR)?

Background: Based on the paradigm that undetected and untreated VUR might lead to long-term complications, voiding cystography traditionally has been recommended for all young children with a first urinary tract infection (UTI). An increasing base of evidence suggests that antibiotic prophylaxis for low-grade VUR might be unnecessary. However, less-invasive methods of screening for high-grade reflux have not yet been identified.

Study design: Secondary analysis of data from a prior prospective study.

Setting: Nephrology department of a French teaching hospital.

Synopsis: One hundred seventeen children (0-16 years) with a UTI were included and underwent renal US and voiding cystography. Patients with a known uropathy or those who had received antibiotics within the past 48 hours were excluded. A generalized linear multilevel model was used to analyze the relationship between standardized renal US criterion and VUR.

Twenty-seven percent of children had VUR and 8% had high-grade VUR (grade ≥3). Pelvic, ureteral, and urinary tract dilatation were significantly associated with high-grade VUR. Ureteral dilatation offered the best combination of standardized criterion, sensitivity (75%), and specificity (88%).

Significant limitations of this study include the use of bag urine cultures and the lack of consensus-based US criterion for ureteral and pelvic dilatation. The authors appropriately caution that these renal US criteria do not identify all children with high-grade VUR and are merely one step toward an intermediate screening strategy for high-grade VUR in order to mitigate adverse effects of universal voiding cystography. Further validation of this work in a clinically representative population will be needed.

Bottom line: Ureteral dilatation accurately identifies children with high-grade VUR.

Citation: Leroy S, Vantalon S, Larakeb A, Ducou-Le-Pointe H, Bensman A. Vesicoureteral reflux in children with urinary tract infection: comparison of diagnostic accuracy of renal US criteria. Radiology. 2010;255(3):890-898.

In This Edition

Literature at a Glance

A guide to this month’s studies

• Risk factors for iatrogenic pneumothorax
• Residency acceptance and use of pharmaceutical industry funding
• Early cholecystectomy outcomes for gallstone pancreatitis
• Use of microbial DNA in sepsis
• Adding rifampicin to vancomycin in MRSA pneumonia
• Rate and outcomes of culture-negative severe sepsis
• Rates of surgical comanagement in U.S. hospitals
• Probiotics and rates of ventilator-associated pneumonia

Ultrasound Guidance and Operator Experience Decrease Risk of Pneumothorax Following Thoracentesis

Clinical question: How often does pneumothorax happen following thoracentesis, and what factors are associated with increased risk of this complication?

Background: Procedural complications are an important source of adverse events in the hospital. Iatrogenic pneumothorax after thoracentesis results in increased hospital length of stay, morbidity, and mortality. Large variation exists in reported pneumothorax rates, and little is known about procedure- and patient-specific factors associated with development of this complication.

Study design: Systematic review and meta-analysis.

Setting: Review of 24 MEDLINE-indexed studies from January 1966 to April 2009.

Synopsis: A total of 349 pneumothoraces were reported after 6,605 thoracenteses (overall incidence 6.0%). Chest-tube insertion was required in 34.1% of the cases. Risk for pneumothorax was significantly higher when larger needles or catheters were used compared with needles smaller than 20-gauge (odds ratio 2.5, 95% confidence interval [CI], 1.1-6.0) and after therapeutic thoracentesis compared with diagnostic procedures (OR 2.6, 95% CI, 1.8-3.8).

Procedures requiring two or more needle passes did not significantly increase pneumothorax risk (OR 2.5, 95% CI, 0.3-20.1). In contrast, pneumothorax rates were significantly lower when using ultrasound guidance (OR 0.3, 95% CI, 0.2-0.7) and with experienced operators (3.9% vs. 8.5%, P=0.04).

Examining patient risk factors, pneumothorax rates were similar regardless of effusion size and patient gender. Additionally, rates were similar among non-ICU inpatients, ICU inpatients, and outpatients. Data did show a trend toward increased risk of pneumothorax with mechanical ventilation (OR 4.0, 95% CI, 0.95-16.8), although no study directly compared rates in ICU patients with and without mechanical ventilation.

Bottom line: Ultrasound guidance is a modifiable factor that decreases the risk of post-thoracentesis pneumothorax. Pneumothorax rates are lower when performed by experienced clinicians, providing an important opportunity to reduce procedure-related complications by increasing direct trainee supervision.

Citation: Gordon CE, Feller-Kopman D, Balk EM, Smetana GW. Pneumothorax following thoracentesis: a systematic review and meta-analysis. Arch Intern Med. 2010;170(4):332-339.

Pharmaceutical Industry Support Is Common in U.S. Internal-Medicine Residency Programs

Clinical question: What are the current attitudes of program directors regarding pharmaceutical industry support of internal-medicine residency activities? What are the potential associations between program characteristics and acceptance of industry support?

Background: Increasing evidence suggests that interactions with the pharmaceutical industry influence physician attitudes and practices. Recently, the Association of American Medical Colleges (AAMC) proposed that academic medical centers prohibit the acceptance of all gifts and restrict access by pharmaceutical industry representatives.

Study design: Survey of U.S. internal-medicine residency program directors.

Setting: Web-based survey of residency program directors in 388 U.S. internal-medicine residency programs.

Synopsis: Of the 236 program directors responding to the survey, 132 (55.9%) reported accepting some kind of support from the pharmaceutical industry. Support was most commonly provided in the form of food for conferences (90.9%), educational materials (83.3%), office supplies (68.9%), and drug samples (57.6%).

When programs reported accepting pharmaceutical industry support, 67.9% cited a lack of other funding sources as the reason for acceptance. Only 22.7% of programs with a program director who thinks pharmaceutical support is unacceptable actually accepted industry support. The likelihood of accepting support was associated with location in the Southern U.S. and was inversely associated with the three-year rolling American Board of Internal Medicine (ABIM) pass rates (each 1% decrease in the pass rate was associated with a 21% increase in the odds of accepting pharmaceutical industry support).

Bottom line: While most program directors did not find pharmaceutical industry support desirable, more than half reported acceptance of such support, with most citing lack of other funding resources as the reason for acceptance.

Citation: Loertscher LL, Halvorsen AJ, Beasley BW, Holmboe ES, Kolars JC, McDonald FS. Pharmaceutical industry support and residency education: a survey of internal medicine program directors. Arch Intern Med. 2010;170(4):356-362.

Early Cholecystectomy Safely Decreases Hospital Stay in Patients with Mild Gallstone Pancreatitis

Clinical question: Can laparoscopic cholecystectomy performed within 48 hours of admission for mild gallstone pancreatitis reduce hospital length of stay without increasing perioperative complications?

Background: Although there is a clear consensus that patients who present with gallstone pancreatitis should undergo cholecystectomy to prevent recurrence, precise timing of surgery remains controversial.

Study design: Randomized prospective trial.

Setting: Harbor-UCLA Medical Center, a Los Angeles County public teaching hospital and Level I trauma center.

Synopsis: Patients were prospectively randomized to an early group and a control group. Inclusion criteria consisted of adults from the ages of 18 to 100 with mild gallstone pancreatitis and three or fewer Ranson criteria. The primary endpoint was length of hospital stay. The secondary endpoint was a composite of complications, including the need for conversion to open cholecystectomy, readmission within 30 days, bleeding requiring transfusion, bile duct injury, or wound infection.

The study was terminated after 50 patients, as there was a difference in the length of hospital stay with a predefined alpha level of 0.005. Patients in the early group were taken to the operating room at a mean of 35.1 hours after admission, compared with 77.8 hours in the control group. The overall length of hospital stay was shorter in the early group (mean 3.5 days, 95% CI, 2.7-4.3), compared with the control group (mean 5.8, 95% CI, 3.8-7.9). All cholecystectomies were completed laparoscopically, without conversion to open. No statistically significant difference existed in secondary endpoints (P=0.48, OR 1.66, 95% CI, 0.41-6.78).

Bottom line: Laparoscopic cholecystectomy performed within 48 hours of admission, irrespective of normalization of laboratory values or clinical progress, safely decreases the overall length of stay, compared with delaying laparoscopic cholecystectomy until laboratory values and clinical condition normalize.

Citation: Aboulian A, Chan T, Yaghoubian A, et al. Early cholecystectomy safely decreases hospital stay in patients with mild gallstone pancreatitis: a randomized prospective study. Ann Surg. 2010;251(4): 615-619.

Presence of Microbial DNA in Blood Correlates with Disease Severity

Clinical question: Is the presence of microbial DNA in the blood associated with disease severity in severe sepsis, and how does detection of this microbial DNA by polymerase chain reaction (PCR) compare with blood cultures (BC)?

Background: Inadequate antibiotic therapy is a strong and independent predictor of poor outcomes in sepsis. Diagnostic uncertainty regarding the causative micro-organism is compensated for by liberal use of broad-spectrum antibiotics. As a result, resistance to antibiotics is an increasing public-health problem.

Study design: Prospective multicenter controlled observational study.

Setting: Three ICUs in Germany and France.

Synopsis: From 2005 to 2007, 63 patients were enrolled in the control group and 142 patients were enrolled in the sepsis group. In control patients, blood cultures and specimens were drawn daily at a maximum of three days after admission. In the sepsis group, blood samples were obtained on the day severe sepsis was suspected. Consecutive samples for the next two days after study inclusion were taken.

Taking BC as the laboratory comparison method, the sensitivity of PCR to detect culture-positive bacteremia in sepsis was 0.80 with a specificity of 0.77. PCR detected 29 of 41 microorganisms (70.3%) found in the BC. The highest recovery rate was observed for gram-negative bacteria (78.6%), fungi (50.0%), and gram-positive bacteria (47.6%). PCR from septic patients correlated well with markers of host response (IL-6 and PCT) and disease severity (SOFA score), even when the BC remained negative.

The appropriateness of antimicrobial therapy based on culture-based methods was not recorded, so it’s impossible to conclude whether or not the PCR would have contributed to a more effective therapy.

Bottom line: Concordance between BC and PCR is moderate in septic patients. PCR-based pathogen detection correlated with disease severity even if the BC remained negative, suggesting that the presence of microbial DNA in the bloodstream is a clinically significant event.

Citation: Bloos F, Hinder F, Becker K, et al. A multicenter trial to compare blood culture with polymerase chain reaction in severe human sepsis. Intensive Care Med. 2010;36(2):241-247.

Adding Rifampicin to Vancomycin Improves Outcomes in MRSA Pneumonia

Clinical question: Does adding rifampicin to vancomycin improve outcomes in patients with hospital-acquired MRSA pneumonia?

Background: Hospital-acquired MRSA pneumonia has a mortality of more than 20%. Vancomycin penetrates the lung tissue poorly. The value of adding rifampicin, an antibiotic with broad-spectrum coverage and good tissue penetration, was investigated.

Study design: Randomized open-label trial.

Setting: Medical ICU patients at Ulsan College of Medicine, Asan Medical Center, South Korea.

Synopsis: Patients older than 18 years of age with clinical symptoms suggestive of nosocomial pneumonia were randomized to receive vancomycin alone (V) or vancomycin plus rifampicin (VR). Clinicians could add additional antibiotics for gram-negative coverage as needed.

Of the 183 patients screened, 93 met the inclusion criteria and were randomized in a 1:1 ratio. MRSA infection was microbiologically confirmed. Clinical cure rate in VR patients was significantly greater at day 14 compared with the V group (53.7% vs. 31.0%, P=0.047) based on a modified intention-to-treat model. The overall mortality at day 28 did not significantly differ between the groups (22.0% vs. 38.1%, P=0.15), although the 60-day mortality was lower in the VR group (26.8% vs. 50.0%, P=0.042). Mortality from MRSA pneumonia had a trend toward a decrease in the VR group (14.7% vs. 28.6%, P=0.18).

The trial was limited because it was a single-site study and lacked statistical power to assess certain outcomes. Additionally, treatment protocols were not compared with other antimicrobial therapies.

Bottom line: Vancomycin plus rifampicin improves MRSA pneumonia outcomes in ICU patients.

Citation: Jung YJ, Koh Y, Hong SB, et al. Effect of vancomycin plus rifampicin in the treatment of nosocomial MRSA pneumonia. Crit Care Med. 2010;38(1):175-180.

Severe Sepsis Syndromes Are Not Always Caused by Bacteremia

Clinical question: What are the common causes of clinical sepsis?

Background: When sepsis is defined by systemic inflammatory response syndrome (SIRS) criteria, the etiology is not always infectious. Rapid initiation of antimicrobial therapy for infectious SIRS is a priority, but it could result in treating a significant number of patients who are not bacteremic.

Study design: Prospective secondary analysis of a registry of patients created to evaluate an institutional standard-of-care protocol.

Setting: Urban, 850-bed, tertiary-care teaching institution in North Carolina.

Synopsis: ED cases meeting the criteria for severe sepsis underwent a secondary review that looked at the cause of the sepsis. Only 45% of patients identified as having severe sepsis were blood-culture-positive during that episode of care. The culture-positive group was more likely to have central lines, malignancies, or reside in a nursing home.

Of the subgroup of culture-negative patients, 52% had another infectious etiology, most commonly pneumonia. Other “noninfectious mimics,” including inflammatory colitis, myocardial infarction, and pulmonary embolism, were noted in 32% of patients in the subgroup, and the cause was not identified in 16% of the patients.

In-hospital mortality was higher in the culture-positive group than in the culture-negative group (25% vs. 4%, P=0.05). There was no evidence of harm in patients with culture-negative sepsis treated for a systemic infection.

Bottom line: Many patients with a clinical picture of severe sepsis will not have positive blood cultures or an infectious etiology.

Citation: Heffner AC, Horton JM, Marchick MR, Jones AE. Etiology of illness in patients with severe sepsis admitted to the hospital from the emergency department. Clin Infect Dis. 2010;50(6):814-820.

Comanagement of Surgical Inpatients by Hospitalists Is Rapidly Expanding

Clinical question: What is the prevalence and nature of comanagement of surgical patients by medicine physicians?

Background: Comanagement of surgical patients is a common clinical role for hospitalists, but the relationship is not well characterized in the literature in terms of numbers of patients or types of physicians involved in this practice.

Study design: Retrospective cohort.

Setting: Cross-section of hospitals from a Medicare database.

Synopsis: During the study period, 35.2% of patients were comanaged by a medicine physician—23.7% by a generalist and 14% by a subspecialist. Cardiothoracic surgery patients were more likely to be comanaged by a subspecialist, whereas all other patients were more likely to be comanaged by a generalist.

Although subspecialist comanagement actually declined during the study period, overall comanagement increased from 33.3% in 1996 to 40.8% in 2006. This increase is entirely attributable to the increase in comanagement by hospitalists. Most of this growth occurred with orthopedic patients.

Patient factors associated with comanagement include advanced age, emergency admissions, and increasing comorbidities. Teaching hospitals had less comanagement, while midsize, nonteaching, and for-profit hospitals had more comanagement.

Bottom line: Comanagement of surgical patients by medicine physicians is a common and growing clinical relationship. Hospitalists are responsible for increasing numbers of comanaged surgical patients.

Citation: Sharma G, Kuo YF, Freeman J, Zhang DD, Goodwin JS. Comanagement of hospitalized surgical patients by medicine physicians in the United States. Arch Intern Med. 2010;170(4):363-368.

Probiotics Might Decrease Risk of Ventilator-Associated Pneumonia

Clinical question: Does the administration of probiotics decrease the incidence of ventilator-associated pneumonia in critically ill patients?

Background: Ventilator-associated pneumonia (VAP) is a major nosocomial infection in ICUs. Probiotics are thought to decrease colonization and, therefore, infection with serious hospital-acquired pathogens.

Study design: Meta-analysis of five randomized controlled trials.

Setting: ICU patients on mechanical ventilation for at least 24 hours.

Synopsis: Five trials met the inclusion criteria of comparing probiotics to placebo in critically ill patients on mechanical ventilation and reporting the outcome of VAP. Administration of probiotics decreased the incidence of VAP (odds ratio 0.61, 95% CI, 0.41-0.91) and colonization of the respiratory tract with Pseudomonas aeruginosa (OR 0.35, 95% CI, 0.13-0.93).

Length of ICU stay was decreased in the probiotic arm, although this effect was not statistically significant in all analyses. Probiotics had no effect on such outcomes as ICU mortality, in-hospital mortality, or duration of mechanical ventilation.

Bottom line: Probiotics might be an effective strategy to reduce the risk of VAP, even if they do not appear to impact such outcomes as mortality.

Citation: Siempos II, Ntaidou TK, Falagas ME. Impact of the administration of probiotics on the incidence of ventilator-associated pneumonia: a meta-analysis of randomized controlled trials. Crit Care Med. 2010;38(3):954-962. TH

PEDIATRIC HM LITERATURE

By Mark Shen, MD

Renal Ultrasound Identifies Children with High-Grade Vesicoureteral Reflux

Reviewed by Pediatric Editor Mark Shen, MD, medical director of hospital medicine at Dell Children’s Medical Center, Austin, Texas.

Clinical question: What is the diagnostic accuracy of specific renal ultrasound (US) criterion for detection of vesicoureteral reflux (VUR)?

Background: Based on the paradigm that undetected and untreated VUR might lead to long-term complications, voiding cystography traditionally has been recommended for all young children with a first urinary tract infection (UTI). An increasing base of evidence suggests that antibiotic prophylaxis for low-grade VUR might be unnecessary. However, less-invasive methods of screening for high-grade reflux have not yet been identified.

Study design: Secondary analysis of data from a prior prospective study.

Setting: Nephrology department of a French teaching hospital.

Synopsis: One hundred seventeen children (0-16 years) with a UTI were included and underwent renal US and voiding cystography. Patients with a known uropathy or those who had received antibiotics within the past 48 hours were excluded. A generalized linear multilevel model was used to analyze the relationship between standardized renal US criterion and VUR.

Twenty-seven percent of children had VUR and 8% had high-grade VUR (grade ≥3). Pelvic, ureteral, and urinary tract dilatation were significantly associated with high-grade VUR. Ureteral dilatation offered the best combination of standardized criterion, sensitivity (75%), and specificity (88%).

Significant limitations of this study include the use of bag urine cultures and the lack of consensus-based US criterion for ureteral and pelvic dilatation. The authors appropriately caution that these renal US criteria do not identify all children with high-grade VUR and are merely one step toward an intermediate screening strategy for high-grade VUR in order to mitigate adverse effects of universal voiding cystography. Further validation of this work in a clinically representative population will be needed.

Bottom line: Ureteral dilatation accurately identifies children with high-grade VUR.

Citation: Leroy S, Vantalon S, Larakeb A, Ducou-Le-Pointe H, Bensman A. Vesicoureteral reflux in children with urinary tract infection: comparison of diagnostic accuracy of renal US criteria. Radiology. 2010;255(3):890-898.

In This Edition

Literature at a Glance

A guide to this month’s studies

• Risk factors for iatrogenic pneumothorax
• Residency acceptance and use of pharmaceutical industry funding
• Early cholecystectomy outcomes for gallstone pancreatitis
• Use of microbial DNA in sepsis
• Adding rifampicin to vancomycin in MRSA pneumonia
• Rate and outcomes of culture-negative severe sepsis
• Rates of surgical comanagement in U.S. hospitals
• Probiotics and rates of ventilator-associated pneumonia

Ultrasound Guidance and Operator Experience Decrease Risk of Pneumothorax Following Thoracentesis

Clinical question: How often does pneumothorax happen following thoracentesis, and what factors are associated with increased risk of this complication?

Background: Procedural complications are an important source of adverse events in the hospital. Iatrogenic pneumothorax after thoracentesis results in increased hospital length of stay, morbidity, and mortality. Large variation exists in reported pneumothorax rates, and little is known about procedure- and patient-specific factors associated with development of this complication.

Study design: Systematic review and meta-analysis.

Setting: Review of 24 MEDLINE-indexed studies from January 1966 to April 2009.

Synopsis: A total of 349 pneumothoraces were reported after 6,605 thoracenteses (overall incidence 6.0%). Chest-tube insertion was required in 34.1% of the cases. Risk for pneumothorax was significantly higher when larger needles or catheters were used compared with needles smaller than 20-gauge (odds ratio 2.5, 95% confidence interval [CI], 1.1-6.0) and after therapeutic thoracentesis compared with diagnostic procedures (OR 2.6, 95% CI, 1.8-3.8).

Procedures requiring two or more needle passes did not significantly increase pneumothorax risk (OR 2.5, 95% CI, 0.3-20.1). In contrast, pneumothorax rates were significantly lower when using ultrasound guidance (OR 0.3, 95% CI, 0.2-0.7) and with experienced operators (3.9% vs. 8.5%, P=0.04).

Examining patient risk factors, pneumothorax rates were similar regardless of effusion size and patient gender. Additionally, rates were similar among non-ICU inpatients, ICU inpatients, and outpatients. Data did show a trend toward increased risk of pneumothorax with mechanical ventilation (OR 4.0, 95% CI, 0.95-16.8), although no study directly compared rates in ICU patients with and without mechanical ventilation.

Bottom line: Ultrasound guidance is a modifiable factor that decreases the risk of post-thoracentesis pneumothorax. Pneumothorax rates are lower when performed by experienced clinicians, providing an important opportunity to reduce procedure-related complications by increasing direct trainee supervision.

Citation: Gordon CE, Feller-Kopman D, Balk EM, Smetana GW. Pneumothorax following thoracentesis: a systematic review and meta-analysis. Arch Intern Med. 2010;170(4):332-339.

Pharmaceutical Industry Support Is Common in U.S. Internal-Medicine Residency Programs

Clinical question: What are the current attitudes of program directors regarding pharmaceutical industry support of internal-medicine residency activities? What are the potential associations between program characteristics and acceptance of industry support?

Background: Increasing evidence suggests that interactions with the pharmaceutical industry influence physician attitudes and practices. Recently, the Association of American Medical Colleges (AAMC) proposed that academic medical centers prohibit the acceptance of all gifts and restrict access by pharmaceutical industry representatives.

Study design: Survey of U.S. internal-medicine residency program directors.

Setting: Web-based survey of residency program directors in 388 U.S. internal-medicine residency programs.

Synopsis: Of the 236 program directors responding to the survey, 132 (55.9%) reported accepting some kind of support from the pharmaceutical industry. Support was most commonly provided in the form of food for conferences (90.9%), educational materials (83.3%), office supplies (68.9%), and drug samples (57.6%).

When programs reported accepting pharmaceutical industry support, 67.9% cited a lack of other funding sources as the reason for acceptance. Only 22.7% of programs with a program director who thinks pharmaceutical support is unacceptable actually accepted industry support. The likelihood of accepting support was associated with location in the Southern U.S. and was inversely associated with the three-year rolling American Board of Internal Medicine (ABIM) pass rates (each 1% decrease in the pass rate was associated with a 21% increase in the odds of accepting pharmaceutical industry support).

Bottom line: While most program directors did not find pharmaceutical industry support desirable, more than half reported acceptance of such support, with most citing lack of other funding resources as the reason for acceptance.

Citation: Loertscher LL, Halvorsen AJ, Beasley BW, Holmboe ES, Kolars JC, McDonald FS. Pharmaceutical industry support and residency education: a survey of internal medicine program directors. Arch Intern Med. 2010;170(4):356-362.

Early Cholecystectomy Safely Decreases Hospital Stay in Patients with Mild Gallstone Pancreatitis

Clinical question: Can laparoscopic cholecystectomy performed within 48 hours of admission for mild gallstone pancreatitis reduce hospital length of stay without increasing perioperative complications?

Background: Although there is a clear consensus that patients who present with gallstone pancreatitis should undergo cholecystectomy to prevent recurrence, precise timing of surgery remains controversial.

Study design: Randomized prospective trial.

Setting: Harbor-UCLA Medical Center, a Los Angeles County public teaching hospital and Level I trauma center.

Synopsis: Patients were prospectively randomized to an early group and a control group. Inclusion criteria consisted of adults from the ages of 18 to 100 with mild gallstone pancreatitis and three or fewer Ranson criteria. The primary endpoint was length of hospital stay. The secondary endpoint was a composite of complications, including the need for conversion to open cholecystectomy, readmission within 30 days, bleeding requiring transfusion, bile duct injury, or wound infection.

The study was terminated after 50 patients, as there was a difference in the length of hospital stay with a predefined alpha level of 0.005. Patients in the early group were taken to the operating room at a mean of 35.1 hours after admission, compared with 77.8 hours in the control group. The overall length of hospital stay was shorter in the early group (mean 3.5 days, 95% CI, 2.7-4.3), compared with the control group (mean 5.8, 95% CI, 3.8-7.9). All cholecystectomies were completed laparoscopically, without conversion to open. No statistically significant difference existed in secondary endpoints (P=0.48, OR 1.66, 95% CI, 0.41-6.78).

Bottom line: Laparoscopic cholecystectomy performed within 48 hours of admission, irrespective of normalization of laboratory values or clinical progress, safely decreases the overall length of stay, compared with delaying laparoscopic cholecystectomy until laboratory values and clinical condition normalize.

Citation: Aboulian A, Chan T, Yaghoubian A, et al. Early cholecystectomy safely decreases hospital stay in patients with mild gallstone pancreatitis: a randomized prospective study. Ann Surg. 2010;251(4): 615-619.

Presence of Microbial DNA in Blood Correlates with Disease Severity

Clinical question: Is the presence of microbial DNA in the blood associated with disease severity in severe sepsis, and how does detection of this microbial DNA by polymerase chain reaction (PCR) compare with blood cultures (BC)?

Background: Inadequate antibiotic therapy is a strong and independent predictor of poor outcomes in sepsis. Diagnostic uncertainty regarding the causative micro-organism is compensated for by liberal use of broad-spectrum antibiotics. As a result, resistance to antibiotics is an increasing public-health problem.

Study design: Prospective multicenter controlled observational study.

Setting: Three ICUs in Germany and France.

Synopsis: From 2005 to 2007, 63 patients were enrolled in the control group and 142 patients were enrolled in the sepsis group. In control patients, blood cultures and specimens were drawn daily at a maximum of three days after admission. In the sepsis group, blood samples were obtained on the day severe sepsis was suspected. Consecutive samples for the next two days after study inclusion were taken.

Taking BC as the laboratory comparison method, the sensitivity of PCR to detect culture-positive bacteremia in sepsis was 0.80 with a specificity of 0.77. PCR detected 29 of 41 microorganisms (70.3%) found in the BC. The highest recovery rate was observed for gram-negative bacteria (78.6%), fungi (50.0%), and gram-positive bacteria (47.6%). PCR from septic patients correlated well with markers of host response (IL-6 and PCT) and disease severity (SOFA score), even when the BC remained negative.

The appropriateness of antimicrobial therapy based on culture-based methods was not recorded, so it’s impossible to conclude whether or not the PCR would have contributed to a more effective therapy.

Bottom line: Concordance between BC and PCR is moderate in septic patients. PCR-based pathogen detection correlated with disease severity even if the BC remained negative, suggesting that the presence of microbial DNA in the bloodstream is a clinically significant event.

Citation: Bloos F, Hinder F, Becker K, et al. A multicenter trial to compare blood culture with polymerase chain reaction in severe human sepsis. Intensive Care Med. 2010;36(2):241-247.

Adding Rifampicin to Vancomycin Improves Outcomes in MRSA Pneumonia

Clinical question: Does adding rifampicin to vancomycin improve outcomes in patients with hospital-acquired MRSA pneumonia?

Background: Hospital-acquired MRSA pneumonia has a mortality of more than 20%. Vancomycin penetrates the lung tissue poorly. The value of adding rifampicin, an antibiotic with broad-spectrum coverage and good tissue penetration, was investigated.

Study design: Randomized open-label trial.

Setting: Medical ICU patients at Ulsan College of Medicine, Asan Medical Center, South Korea.

Synopsis: Patients older than 18 years of age with clinical symptoms suggestive of nosocomial pneumonia were randomized to receive vancomycin alone (V) or vancomycin plus rifampicin (VR). Clinicians could add additional antibiotics for gram-negative coverage as needed.

Of the 183 patients screened, 93 met the inclusion criteria and were randomized in a 1:1 ratio. MRSA infection was microbiologically confirmed. Clinical cure rate in VR patients was significantly greater at day 14 compared with the V group (53.7% vs. 31.0%, P=0.047) based on a modified intention-to-treat model. The overall mortality at day 28 did not significantly differ between the groups (22.0% vs. 38.1%, P=0.15), although the 60-day mortality was lower in the VR group (26.8% vs. 50.0%, P=0.042). Mortality from MRSA pneumonia had a trend toward a decrease in the VR group (14.7% vs. 28.6%, P=0.18).

The trial was limited because it was a single-site study and lacked statistical power to assess certain outcomes. Additionally, treatment protocols were not compared with other antimicrobial therapies.

Bottom line: Vancomycin plus rifampicin improves MRSA pneumonia outcomes in ICU patients.

Citation: Jung YJ, Koh Y, Hong SB, et al. Effect of vancomycin plus rifampicin in the treatment of nosocomial MRSA pneumonia. Crit Care Med. 2010;38(1):175-180.

Severe Sepsis Syndromes Are Not Always Caused by Bacteremia

Clinical question: What are the common causes of clinical sepsis?

Background: When sepsis is defined by systemic inflammatory response syndrome (SIRS) criteria, the etiology is not always infectious. Rapid initiation of antimicrobial therapy for infectious SIRS is a priority, but it could result in treating a significant number of patients who are not bacteremic.

Study design: Prospective secondary analysis of a registry of patients created to evaluate an institutional standard-of-care protocol.

Setting: Urban, 850-bed, tertiary-care teaching institution in North Carolina.

Synopsis: ED cases meeting the criteria for severe sepsis underwent a secondary review that looked at the cause of the sepsis. Only 45% of patients identified as having severe sepsis were blood-culture-positive during that episode of care. The culture-positive group was more likely to have central lines, malignancies, or reside in a nursing home.

Of the subgroup of culture-negative patients, 52% had another infectious etiology, most commonly pneumonia. Other “noninfectious mimics,” including inflammatory colitis, myocardial infarction, and pulmonary embolism, were noted in 32% of patients in the subgroup, and the cause was not identified in 16% of the patients.

In-hospital mortality was higher in the culture-positive group than in the culture-negative group (25% vs. 4%, P=0.05). There was no evidence of harm in patients with culture-negative sepsis treated for a systemic infection.

Bottom line: Many patients with a clinical picture of severe sepsis will not have positive blood cultures or an infectious etiology.

Citation: Heffner AC, Horton JM, Marchick MR, Jones AE. Etiology of illness in patients with severe sepsis admitted to the hospital from the emergency department. Clin Infect Dis. 2010;50(6):814-820.

Comanagement of Surgical Inpatients by Hospitalists Is Rapidly Expanding

Clinical question: What is the prevalence and nature of comanagement of surgical patients by medicine physicians?

Background: Comanagement of surgical patients is a common clinical role for hospitalists, but the relationship is not well characterized in the literature in terms of numbers of patients or types of physicians involved in this practice.

Study design: Retrospective cohort.

Setting: Cross-section of hospitals from a Medicare database.

Synopsis: During the study period, 35.2% of patients were comanaged by a medicine physician—23.7% by a generalist and 14% by a subspecialist. Cardiothoracic surgery patients were more likely to be comanaged by a subspecialist, whereas all other patients were more likely to be comanaged by a generalist.

Although subspecialist comanagement actually declined during the study period, overall comanagement increased from 33.3% in 1996 to 40.8% in 2006. This increase is entirely attributable to the increase in comanagement by hospitalists. Most of this growth occurred with orthopedic patients.

Patient factors associated with comanagement include advanced age, emergency admissions, and increasing comorbidities. Teaching hospitals had less comanagement, while midsize, nonteaching, and for-profit hospitals had more comanagement.

Bottom line: Comanagement of surgical patients by medicine physicians is a common and growing clinical relationship. Hospitalists are responsible for increasing numbers of comanaged surgical patients.

Citation: Sharma G, Kuo YF, Freeman J, Zhang DD, Goodwin JS. Comanagement of hospitalized surgical patients by medicine physicians in the United States. Arch Intern Med. 2010;170(4):363-368.

Probiotics Might Decrease Risk of Ventilator-Associated Pneumonia

Clinical question: Does the administration of probiotics decrease the incidence of ventilator-associated pneumonia in critically ill patients?

Background: Ventilator-associated pneumonia (VAP) is a major nosocomial infection in ICUs. Probiotics are thought to decrease colonization and, therefore, infection with serious hospital-acquired pathogens.

Study design: Meta-analysis of five randomized controlled trials.

Setting: ICU patients on mechanical ventilation for at least 24 hours.

Synopsis: Five trials met the inclusion criteria of comparing probiotics to placebo in critically ill patients on mechanical ventilation and reporting the outcome of VAP. Administration of probiotics decreased the incidence of VAP (odds ratio 0.61, 95% CI, 0.41-0.91) and colonization of the respiratory tract with Pseudomonas aeruginosa (OR 0.35, 95% CI, 0.13-0.93).

Length of ICU stay was decreased in the probiotic arm, although this effect was not statistically significant in all analyses. Probiotics had no effect on such outcomes as ICU mortality, in-hospital mortality, or duration of mechanical ventilation.

Bottom line: Probiotics might be an effective strategy to reduce the risk of VAP, even if they do not appear to impact such outcomes as mortality.

Citation: Siempos II, Ntaidou TK, Falagas ME. Impact of the administration of probiotics on the incidence of ventilator-associated pneumonia: a meta-analysis of randomized controlled trials. Crit Care Med. 2010;38(3):954-962. TH

PEDIATRIC HM LITERATURE

By Mark Shen, MD

Renal Ultrasound Identifies Children with High-Grade Vesicoureteral Reflux

Reviewed by Pediatric Editor Mark Shen, MD, medical director of hospital medicine at Dell Children’s Medical Center, Austin, Texas.

Clinical question: What is the diagnostic accuracy of specific renal ultrasound (US) criterion for detection of vesicoureteral reflux (VUR)?

Background: Based on the paradigm that undetected and untreated VUR might lead to long-term complications, voiding cystography traditionally has been recommended for all young children with a first urinary tract infection (UTI). An increasing base of evidence suggests that antibiotic prophylaxis for low-grade VUR might be unnecessary. However, less-invasive methods of screening for high-grade reflux have not yet been identified.

Study design: Secondary analysis of data from a prior prospective study.

Setting: Nephrology department of a French teaching hospital.

Synopsis: One hundred seventeen children (0-16 years) with a UTI were included and underwent renal US and voiding cystography. Patients with a known uropathy or those who had received antibiotics within the past 48 hours were excluded. A generalized linear multilevel model was used to analyze the relationship between standardized renal US criterion and VUR.

Twenty-seven percent of children had VUR and 8% had high-grade VUR (grade ≥3). Pelvic, ureteral, and urinary tract dilatation were significantly associated with high-grade VUR. Ureteral dilatation offered the best combination of standardized criterion, sensitivity (75%), and specificity (88%).

Significant limitations of this study include the use of bag urine cultures and the lack of consensus-based US criterion for ureteral and pelvic dilatation. The authors appropriately caution that these renal US criteria do not identify all children with high-grade VUR and are merely one step toward an intermediate screening strategy for high-grade VUR in order to mitigate adverse effects of universal voiding cystography. Further validation of this work in a clinically representative population will be needed.

Bottom line: Ureteral dilatation accurately identifies children with high-grade VUR.

Citation: Leroy S, Vantalon S, Larakeb A, Ducou-Le-Pointe H, Bensman A. Vesicoureteral reflux in children with urinary tract infection: comparison of diagnostic accuracy of renal US criteria. Radiology. 2010;255(3):890-898.

Case

A 72-year-old female with multiple medical problems is admitted with a hip fracture. Surgery is scheduled in 48 hours. The patient’s home medications include aspirin, carbidopa/levodopa, celecoxib, clonidine, estradiol, ginkgo, lisinopril, NPH insulin, sulfasalazine, and prednisone 10 mg a day, which she has been taking for years. How should these and other medications be managed in the perioperative period?

PHOTODISC

Background

Perioperative management of chronic medications is a complex issue, as physicians are required to balance the beneficial and harmful effects of the individual drugs prescribed to their patients. On one hand, cessation of medications can result in decompensation of disease or withdrawal. On the other hand, continuation of drugs can alter metabolism of anesthetic agents, cause perioperative hemodynamic instability, or result in such post-operative complications as acute renal failure, bleeding, infection, and impaired wound healing.

Certain traits make it reasonable to continue medications during the perioperative period. A long elimination half-life or duration of action makes stopping some medications impractical as it takes four to five half-lives to completely clear the drug from the body; holding the drug for a few days around surgery will not appreciably affect its concentration. Stopping drugs that carry severe withdrawal symptoms can be impractical because of the need for lengthy tapers, which can delay surgery and result in decompensation of underlying disease.

Drugs with no significant interactions with anesthesia or risk of perioperative complications should be continued in order to avoid deterioration of the underlying disease. Conversely, drugs that interact with anesthesia or increase risk for complications should be stopped if this can be accomplished safely. Patient-specific factors should receive consideration, as the risk of complications has to be balanced against the danger of exacerbating the underlying disease.

Overview of the Data

The challenge in providing recommendations on perioperative medication management lies in a dearth of high-quality clinical trials. Thus, much of the information comes from case reports, expert opinion, and sound application of pharmacology.

Antiplatelet therapy: Nuances of perioperative antiplatelet therapy are beyond the scope of this review, but some general principles can be elucidated from the American College of Cardiology/American Heart Association (ACC/AHA) 2007 perioperative guidelines.1 Management of antiplatelet therapy should be done in conjunction with the surgical team, as cardiovascular risk has to be weighed against bleeding risk.

Aspirin therapy should be continued in all patients with a history of coronary artery disease (CAD), balloon angioplasty, or percutaneous coronary intervention (PCI), unless the risk of bleeding complications is felt to exceed the cardioprotective benefits—for example, in some neurosurgical patients.1

Clopidogrel therapy is crucial for prevention of in-stent thrombosis (IST) following PCI because patients who experience IST suffer catastrophic myocardial infarctions with high mortality. Ideally, surgery should be delayed to permit completion of clopidogrel therapy—30 to 45 days after implantation of a bare-metal stent and 365 days after a drug-eluting stent. If surgery has to be performed sooner, guidelines recommend operating on dual antiplatelet therapy with aspirin and clopidogrel.1 Again, this course of treatment has to be balanced against the risk of hemorrhagic complications from surgery.

Both aspirin and clopidogrel irreversibly inhibit platelet aggregation. The recovery of normal coagulation involves formation of new platelets, which necessitates cessation of therapy for seven to 10 days before surgery. Platelet inhibition begins within minutes of restarting aspirin and within hours of taking clopidogrel, although attaining peak clopidogrel effect takes three to seven days, unless a loading dose is used.

Cardiovascular Drugs

Beta-blockers in the perioperative setting are a focus of an ongoing debate beyond the scope of this review (see “What Pre-Operative Cardiac Evaluation of Patients Undergoing Intermediate-Risk Surgery Is Most Effective?,” February 2008, p. 26). Given the current evidence and the latest ACC/AHA guidelines, it is still reasonable to continue therapy in patients who are already taking them to avoid precipitating cardiovascular events by withdrawal. Patients with increased cardiac risk, demonstrated by a Revised Cardiac Risk Index (RCRI) score of ≥2 (see Table 1, p. 12), should be considered for beta-blocker therapy before surgery.1 In either case, the dose should be titrated to a heart rate <65 for optimal cardiac protection.1

Statins should be continued if the patient is taking them, especially because preoperative withdrawal has been associated with a 4.6-fold increase in troponin release and a 7.5-fold increased risk of myocardial infarction (MI) and cardiovascular death following major vascular surgery.2 Patients with increased cardiac risk— RCRI ≥1—can be considered for initiation of statin therapy before surgery, although the benefit of this intervention has not been examined in prospective studies.1

Amiodarone has an exceptionally long half-life of up to 142 days. It should be continued in the perioperative period.

Calcium channel blockers (CCBs) can be continued with no unpleasant perioperative hemodynamic effects.1 CCBs have potential cardioprotective benefits.

Clonidine withdrawal can result in severe rebound hypertension with reports of encephalopathy, stroke, and death. These effects are exacerbated by concomitant beta-blocker therapy. For this reason, if a patient is expected to be NPO for more than 12 hours, they should be converted to a clonidine patch 48-72 hours before surgery with concurrent tapering of the oral dose.3

Digoxin has a long half-life (up to 48 hours) and should be continued with monitoring of levels if there is a change in renal function.

Angiotensin converting enzyme inhibitors (ACEI) and angiotensin receptor blockers (ARBs) have been associated with a 50% increased risk of hypotension requiring vasopressors during induction of anesthesia.4 However, it is worth mentioning that this finding has not been corroborated in other studies. A large retrospective cohort of cardiothoracic surgical patients found a 28% increased risk of post-operative acute renal failure (ARF) with both drug classes, although another cardiothoracic report published the same year demonstrated a 50% reduction in risk with ACEIs.5,6 Although the evidence of harm is not unequivocal, perioperative blood-pressure control can be achieved with other drugs without hemodynamic or renal risk, such as CCBs, and in most cases ACEIs/ARBs should be stopped one day before surgery.

Diuretics carry a risk of volume depletion and electrolyte derangements, and should be stopped once a patient becomes NPO. Excess volume is managed with as-needed intravenous formulations.

click for large version

Drugs Acting on the Central Nervous System

The majority of central nervous system (CNS)-active drugs, including antiepileptics, antipsychotics, benzodiazepines, bupropion, gabapentin, lithium, mirtazapine, selective serotonin and norepinephrine reuptake inhibitors (SSRIs and SNRIs), tricyclic antidepressants (TCAs), and valproic acid, balance a low risk of perioperative complications against a significant potential for withdrawal and disease decompensation. Therefore, these medications should be continued.

Carbidopa/Levodopa should be continued because abrupt cessation can precipitate systemic withdrawal resembling serotonin syndrome and rapid deterioration of Parkinson’s symptoms.

Monoamine oxidase inhibitor (MAOI) therapy usually indicates refractory psychiatric illness, so these drugs should be continued to avoid decompensation. Importantly, MAOI-safe anesthesia without dextromethorphan, meperidine, epinephrine, or norepinephrine has to be used due to the risk of cardiovascular instability.7

Diabetic Drugs

Insulin therapy should be continued with adjustments. Glargine basal insulin has no peak and can be continued without changing the dose. Likewise, patients with insulin pumps can continue the usual basal rate. Short-acting insulin or such insulin mixes as 70/30 should be stopped four hours before surgery to avoid hypoglycemia. Intermediate-acting insulin (e.g., NPH) can be administered at half the usual dose the day of surgery with a perioperative 5% dextrose infusion. NPH should not be given the day of surgery if the dextrose infusion cannot be used.8

Incretins (exenatide, sitagliptin) rarely cause hypoglycemia in the absence of insulin and may be beneficial in controlling intraoperative hyperglycemia. Therefore, these medications can be continued.8

Thiazolidinediones (TZDs; pioglitazone, rosiglitazone) alter gene transcription with biological duration of action on the order of weeks and low risk of hypoglycemia, and should be continued.

Metformin carries an FDA black-box warning to discontinue therapy before any intravascular radiocontrast studies or surgical procedures due to the risk of severe lactic acidosis if renal failure develops. It should be stopped 24 hours before surgery and restarted at least 48-72 hours after. Normal renal function should be confirmed before restarting therapy.8

Sulfonylureas (glimepiride, glipizide, glyburide) carry a significant risk of hypoglycemia in a patient who is NPO; they should be stopped the night before surgery or before commencement of NPO status.

Hormones

Antithyroid drugs (methimazole, propylthiouracil) and levothyroxine should be continued, as they have no perioperative contraindications.

Oral contraceptives (OCPs), hormone replacement therapy (HRT), and raloxifene can increase the risk of DVT. The largest study on the topic was the HERS trial of postmenopausal women on estrogen/progesterone HRT. The authors reported a 4.9-fold increased risk of DVT for 90 days after surgery.9 Unfortunately, no information was provided on the types of surgery, or whether appropriate and consistent DVT prophylaxis was utilized. HERS authors also reported a 2.5-fold increased risk of DVT for at least 30 days after cessation of HRT.9

Given the data, it is reasonable to stop hormone therapy four weeks before surgery when prolonged immobilization is anticipated and patients are able to tolerate hormone withdrawal, especially if other DVT risk factors are present. If hormone therapy cannot be stopped, strong consideration should be given to higher-intensity DVT prophylaxis (e.g., chemoprophylaxis as opposed to mechanical measures) of longer duration—up to 28 days following general surgery and up to 35 days after orthopedic procedures.10

click for large version

Perioperative Corticosteroids

Corticosteroid therapy in excess of prednisone 5 mg/day or equivalent for more than five days in the 30 days preceding surgery might predispose patients to acute adrenal insufficiency in the perioperative period. Surgical procedures typically result in cortisol release of 50-150 mg/day, which returns to baseline within 48 hours.11 Therefore, the recommendation is to continue a patient’s baseline steroid dose and supplement it with stress-dose steroids tailored to the severity of operative stress (see Table 2, above).

Mineralocorticoid supplementation is not necessary, because endogenous production is not suppressed by corticosteroid therapy.11 Although a recent systematic review suggests that routine stress-dose steroids might not be indicated, high-quality prospective data are needed before abandoning this strategy due to complications of acute adrenal insufficiency compared to the risk of a brief corticosteroid burst.

Non-Steroidal Anti-Inflammatory Drugs (NSAIDs)

Nonselective cyclooxygenase (COX) inhibitors reversibly decrease platelet aggregation only while the drug is present in the circulation and should be stopped one to three days before surgery due to risk of bleeding.

Selective COX-2 inhibitors do not significantly alter platelet aggregation and can be continued for opioid-sparing perioperative pain control.

Both COX-2-selective and nonselective inhibitors should be held if there are concerns for impaired renal function.

Disease-Modifying Antirheumatic Drugs (DMARDs) and Biological Response Modifiers (BRMs)

Methotrexate increases the risk of wound infections and dehiscence. However, this is offset by a decreased risk of post-operative disease flares with continued use. It can be continued unless the patient has medical comorbidities, advanced age, or chronic therapy with more than 10 mg/day of prednisone, in which case the drug should be stopped two weeks before surgery.12

Azathioprine, leflunomide, and sulfasalazine are renally cleared with a risk of myelosuppression; all of these medications should be stopped. Long half-life of leflunomide necessitates stopping it two weeks before surgery; azathioprine and sulfasalazine can be stopped one day in advance. The drugs can be restarted three days after surgery, assuming stable renal function.13

Anti-TNF-α (adalimumab, etanercept, infliximab), IL1 antagonist (anakinra), and anti-CD20 (rituximab) agents should be stopped one week before surgery and resumed 1-2 weeks afterward, unless risk of complications from disease flareup outweighs the concern for wound infections and dehiscence.14

Herbal Medicines

It is estimated that as much as a third of the U.S. population uses herbal medicines. These substances can result in perioperative hemodynamic instability (ephedra, ginseng, ma huang), hypoglycemia (ginseng), immunosuppression (echinacea, when taken for more than eight weeks), abnormal bleeding (garlic, ginkgo, ginseng), and prolongation of anesthesia (kava, St. John’s wort, valerian). All of these herbal medicine should be stopped one to two weeks before surgery.15,16

Back to the Case

The patient’s Carbidopa/Levodopa should be continued. Celecoxib can be continued if her renal function in stable. If aspirin is taken for a history of coronary artery disease or percutaneous coronary intervention, it should be continued, if possible. Clonidine should be continued or changed to a patch if an extended NPO period is anticipated. Ginkgo, lisinopril, and sulfasalazine should be stopped.

Hospitalization does not provide the luxury of stopping estradiol in advance, so it might be continued with chemical DVT prophylaxis for up to 35 days after surgery. The patient should receive 50-75 mg of IV hydrocortisone during surgery and an additional 25 mg the following day, in addition to her usual prednisone 10 mg/day. She can either receive half her usual NPH dose the morning of surgery with a 5% dextrose infusion in the operating room, or the NPH should be held altogether.

Bottom Line

Perioperative medication use should be tailored to each patient, balancing the risks and benefits of individual drugs. High-quality trials are needed to provide more robust clinical guidelines. TH

Dr. Levin is a hospitalist at the University of Colorado Denver.

References

1. Fleisher LA, Beckman JA, Brown KA, et al. ACC/AHA 2007 Guidelines on Perioperative Cardiovascular Evaluation and Care for Noncardiac Surgery: Executive Summary: A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the 2002 Guidelines on Perioperative Cardiovascular Evaluation for Noncardiac Surgery): Developed in Collaboration With the American Society of Echocardiography, American Society of Nuclear Cardiology, Heart Rhythm Society, Society of Cardiovascular Anesthesiologists, Society for Cardiovascular Angiography and Interventions, Society for Vascular Medicine and Biology, and Society for Vascular Surgery. Circulation. 2007;116(17):1971-1996.
2. Schouten O, Hoeks SE, Welten GM, et al. Effect of statin withdrawal on frequency of cardiac events after vascular surgery. Am J Cardiol. 2007;100(2):316-320.
3. Spell NO III. Stopping and restarting medications in the perioperative period. Med Clin North Am. 2001;85(5):1117-1128.
4. Rosenman DJ, McDonald FS, Ebbert JO, Erwin PJ, LaBella M, Montori VM. Clinical consequences of withholding versus administering renin-angiotensin-aldosterone system antagonists in the preoperative period. J Hosp Med. 2008;3(4):319-325.
5. Arora P, Rajagopalam S, Ranjan R, et al. Preoperative use of angiotensin-converting enzyme inhibitors/ angiotensin receptor blockers is associated with increased risk for acute kidney injury after cardiovascular surgery. Clin J Am Soc Nephrol. 2008;3(5):1266-1273.
6. Benedetto U, Sciarretta S, Roscitano A, et al. Preoperative Angiotensin-converting enzyme inhibitors and acute kidney injury after coronary artery bypass grafting. Ann Thorac Surg. 2008;86(4):1160-1165.
7. Pass SE, Simpson RW. Discontinuation and reinstitution of medications during the perioperative period. Am J Health Syst Pharm. 2004;61(9):899-914.
8. Kohl BA, Schwartz S. Surgery in the patient with endocrine dysfunction. Med Clin North Am. 2009;93(5):1031-1047.
9. Grady D, Wenger NK, Herrington D, et al. Postmenopausal hormone therapy increases risk for venous thromboembolic disease: The Heart and Estrogen/progestin Replacement Study. Ann Intern Med. 2000;132(9):689-696.
10. Hirsh J, Guyatt G, Albers GW, Harrington R, Schünemann HJ; American College of Chest Physicians. Executive summary: American College of Chest Physicians evidence-based clinical practice guidelines (8th edition). Chest. 2008;133(6 Suppl):71S-109S.
11. Axelrod L. Perioperative management of patients treated with glucocorticoids. Endocrinol Metab Clin North Am. 2003;32(2):367-383.
12. Marik PE, Varon J. Requirement of perioperative stress doses of corticosteroids: a systematic review of the literature. Arch Surg. 2008;143(12):1222-1226.
13. Rosandich PA, Kelley JT III, Conn DL. Perioperative management of patients with rheumatoid arthritis in the era of biologic response modifiers. Curr Opin Rheumatol. 2004;16(3):192-198.
14. Saag KG, Teng GG, Patkar NM, et al. American College of Rheumatology 2008 recommendations for the use of nonbiologic and biologic disease-modifying antirheumatic drugs in rheumatoid arthritis. Arthritis Rheum. 2008;59(6):762-784.
15. Ang-Lee MK, Moss J, Yuan CS. Herbal medicines and perioperative care. JAMA. 2001;286(2):208-216.
16. Hodges PJ, Kam PC. The peri-operative implications of herbal medicines. Anaesthesia. 2002;57(9):889-899.

Case

A 72-year-old female with multiple medical problems is admitted with a hip fracture. Surgery is scheduled in 48 hours. The patient’s home medications include aspirin, carbidopa/levodopa, celecoxib, clonidine, estradiol, ginkgo, lisinopril, NPH insulin, sulfasalazine, and prednisone 10 mg a day, which she has been taking for years. How should these and other medications be managed in the perioperative period?

PHOTODISC

Background

Perioperative management of chronic medications is a complex issue, as physicians are required to balance the beneficial and harmful effects of the individual drugs prescribed to their patients. On one hand, cessation of medications can result in decompensation of disease or withdrawal. On the other hand, continuation of drugs can alter metabolism of anesthetic agents, cause perioperative hemodynamic instability, or result in such post-operative complications as acute renal failure, bleeding, infection, and impaired wound healing.

Certain traits make it reasonable to continue medications during the perioperative period. A long elimination half-life or duration of action makes stopping some medications impractical as it takes four to five half-lives to completely clear the drug from the body; holding the drug for a few days around surgery will not appreciably affect its concentration. Stopping drugs that carry severe withdrawal symptoms can be impractical because of the need for lengthy tapers, which can delay surgery and result in decompensation of underlying disease.

Drugs with no significant interactions with anesthesia or risk of perioperative complications should be continued in order to avoid deterioration of the underlying disease. Conversely, drugs that interact with anesthesia or increase risk for complications should be stopped if this can be accomplished safely. Patient-specific factors should receive consideration, as the risk of complications has to be balanced against the danger of exacerbating the underlying disease.

Overview of the Data

The challenge in providing recommendations on perioperative medication management lies in a dearth of high-quality clinical trials. Thus, much of the information comes from case reports, expert opinion, and sound application of pharmacology.

Antiplatelet therapy: Nuances of perioperative antiplatelet therapy are beyond the scope of this review, but some general principles can be elucidated from the American College of Cardiology/American Heart Association (ACC/AHA) 2007 perioperative guidelines.1 Management of antiplatelet therapy should be done in conjunction with the surgical team, as cardiovascular risk has to be weighed against bleeding risk.

Aspirin therapy should be continued in all patients with a history of coronary artery disease (CAD), balloon angioplasty, or percutaneous coronary intervention (PCI), unless the risk of bleeding complications is felt to exceed the cardioprotective benefits—for example, in some neurosurgical patients.1

Clopidogrel therapy is crucial for prevention of in-stent thrombosis (IST) following PCI because patients who experience IST suffer catastrophic myocardial infarctions with high mortality. Ideally, surgery should be delayed to permit completion of clopidogrel therapy—30 to 45 days after implantation of a bare-metal stent and 365 days after a drug-eluting stent. If surgery has to be performed sooner, guidelines recommend operating on dual antiplatelet therapy with aspirin and clopidogrel.1 Again, this course of treatment has to be balanced against the risk of hemorrhagic complications from surgery.

Both aspirin and clopidogrel irreversibly inhibit platelet aggregation. The recovery of normal coagulation involves formation of new platelets, which necessitates cessation of therapy for seven to 10 days before surgery. Platelet inhibition begins within minutes of restarting aspirin and within hours of taking clopidogrel, although attaining peak clopidogrel effect takes three to seven days, unless a loading dose is used.

Cardiovascular Drugs

Beta-blockers in the perioperative setting are a focus of an ongoing debate beyond the scope of this review (see “What Pre-Operative Cardiac Evaluation of Patients Undergoing Intermediate-Risk Surgery Is Most Effective?,” February 2008, p. 26). Given the current evidence and the latest ACC/AHA guidelines, it is still reasonable to continue therapy in patients who are already taking them to avoid precipitating cardiovascular events by withdrawal. Patients with increased cardiac risk, demonstrated by a Revised Cardiac Risk Index (RCRI) score of ≥2 (see Table 1, p. 12), should be considered for beta-blocker therapy before surgery.1 In either case, the dose should be titrated to a heart rate <65 for optimal cardiac protection.1

Statins should be continued if the patient is taking them, especially because preoperative withdrawal has been associated with a 4.6-fold increase in troponin release and a 7.5-fold increased risk of myocardial infarction (MI) and cardiovascular death following major vascular surgery.2 Patients with increased cardiac risk— RCRI ≥1—can be considered for initiation of statin therapy before surgery, although the benefit of this intervention has not been examined in prospective studies.1

Amiodarone has an exceptionally long half-life of up to 142 days. It should be continued in the perioperative period.

Calcium channel blockers (CCBs) can be continued with no unpleasant perioperative hemodynamic effects.1 CCBs have potential cardioprotective benefits.

Clonidine withdrawal can result in severe rebound hypertension with reports of encephalopathy, stroke, and death. These effects are exacerbated by concomitant beta-blocker therapy. For this reason, if a patient is expected to be NPO for more than 12 hours, they should be converted to a clonidine patch 48-72 hours before surgery with concurrent tapering of the oral dose.3

Digoxin has a long half-life (up to 48 hours) and should be continued with monitoring of levels if there is a change in renal function.

Angiotensin converting enzyme inhibitors (ACEI) and angiotensin receptor blockers (ARBs) have been associated with a 50% increased risk of hypotension requiring vasopressors during induction of anesthesia.4 However, it is worth mentioning that this finding has not been corroborated in other studies. A large retrospective cohort of cardiothoracic surgical patients found a 28% increased risk of post-operative acute renal failure (ARF) with both drug classes, although another cardiothoracic report published the same year demonstrated a 50% reduction in risk with ACEIs.5,6 Although the evidence of harm is not unequivocal, perioperative blood-pressure control can be achieved with other drugs without hemodynamic or renal risk, such as CCBs, and in most cases ACEIs/ARBs should be stopped one day before surgery.

Diuretics carry a risk of volume depletion and electrolyte derangements, and should be stopped once a patient becomes NPO. Excess volume is managed with as-needed intravenous formulations.

click for large version

Drugs Acting on the Central Nervous System

The majority of central nervous system (CNS)-active drugs, including antiepileptics, antipsychotics, benzodiazepines, bupropion, gabapentin, lithium, mirtazapine, selective serotonin and norepinephrine reuptake inhibitors (SSRIs and SNRIs), tricyclic antidepressants (TCAs), and valproic acid, balance a low risk of perioperative complications against a significant potential for withdrawal and disease decompensation. Therefore, these medications should be continued.

Carbidopa/Levodopa should be continued because abrupt cessation can precipitate systemic withdrawal resembling serotonin syndrome and rapid deterioration of Parkinson’s symptoms.

Monoamine oxidase inhibitor (MAOI) therapy usually indicates refractory psychiatric illness, so these drugs should be continued to avoid decompensation. Importantly, MAOI-safe anesthesia without dextromethorphan, meperidine, epinephrine, or norepinephrine has to be used due to the risk of cardiovascular instability.7

Diabetic Drugs

Insulin therapy should be continued with adjustments. Glargine basal insulin has no peak and can be continued without changing the dose. Likewise, patients with insulin pumps can continue the usual basal rate. Short-acting insulin or such insulin mixes as 70/30 should be stopped four hours before surgery to avoid hypoglycemia. Intermediate-acting insulin (e.g., NPH) can be administered at half the usual dose the day of surgery with a perioperative 5% dextrose infusion. NPH should not be given the day of surgery if the dextrose infusion cannot be used.8

Incretins (exenatide, sitagliptin) rarely cause hypoglycemia in the absence of insulin and may be beneficial in controlling intraoperative hyperglycemia. Therefore, these medications can be continued.8

Thiazolidinediones (TZDs; pioglitazone, rosiglitazone) alter gene transcription with biological duration of action on the order of weeks and low risk of hypoglycemia, and should be continued.

Metformin carries an FDA black-box warning to discontinue therapy before any intravascular radiocontrast studies or surgical procedures due to the risk of severe lactic acidosis if renal failure develops. It should be stopped 24 hours before surgery and restarted at least 48-72 hours after. Normal renal function should be confirmed before restarting therapy.8

Sulfonylureas (glimepiride, glipizide, glyburide) carry a significant risk of hypoglycemia in a patient who is NPO; they should be stopped the night before surgery or before commencement of NPO status.

Hormones

Antithyroid drugs (methimazole, propylthiouracil) and levothyroxine should be continued, as they have no perioperative contraindications.

Oral contraceptives (OCPs), hormone replacement therapy (HRT), and raloxifene can increase the risk of DVT. The largest study on the topic was the HERS trial of postmenopausal women on estrogen/progesterone HRT. The authors reported a 4.9-fold increased risk of DVT for 90 days after surgery.9 Unfortunately, no information was provided on the types of surgery, or whether appropriate and consistent DVT prophylaxis was utilized. HERS authors also reported a 2.5-fold increased risk of DVT for at least 30 days after cessation of HRT.9

Given the data, it is reasonable to stop hormone therapy four weeks before surgery when prolonged immobilization is anticipated and patients are able to tolerate hormone withdrawal, especially if other DVT risk factors are present. If hormone therapy cannot be stopped, strong consideration should be given to higher-intensity DVT prophylaxis (e.g., chemoprophylaxis as opposed to mechanical measures) of longer duration—up to 28 days following general surgery and up to 35 days after orthopedic procedures.10

click for large version

Perioperative Corticosteroids

Corticosteroid therapy in excess of prednisone 5 mg/day or equivalent for more than five days in the 30 days preceding surgery might predispose patients to acute adrenal insufficiency in the perioperative period. Surgical procedures typically result in cortisol release of 50-150 mg/day, which returns to baseline within 48 hours.11 Therefore, the recommendation is to continue a patient’s baseline steroid dose and supplement it with stress-dose steroids tailored to the severity of operative stress (see Table 2, above).

Mineralocorticoid supplementation is not necessary, because endogenous production is not suppressed by corticosteroid therapy.11 Although a recent systematic review suggests that routine stress-dose steroids might not be indicated, high-quality prospective data are needed before abandoning this strategy due to complications of acute adrenal insufficiency compared to the risk of a brief corticosteroid burst.

Non-Steroidal Anti-Inflammatory Drugs (NSAIDs)

Nonselective cyclooxygenase (COX) inhibitors reversibly decrease platelet aggregation only while the drug is present in the circulation and should be stopped one to three days before surgery due to risk of bleeding.

Selective COX-2 inhibitors do not significantly alter platelet aggregation and can be continued for opioid-sparing perioperative pain control.

Both COX-2-selective and nonselective inhibitors should be held if there are concerns for impaired renal function.

Disease-Modifying Antirheumatic Drugs (DMARDs) and Biological Response Modifiers (BRMs)

Methotrexate increases the risk of wound infections and dehiscence. However, this is offset by a decreased risk of post-operative disease flares with continued use. It can be continued unless the patient has medical comorbidities, advanced age, or chronic therapy with more than 10 mg/day of prednisone, in which case the drug should be stopped two weeks before surgery.12

Azathioprine, leflunomide, and sulfasalazine are renally cleared with a risk of myelosuppression; all of these medications should be stopped. Long half-life of leflunomide necessitates stopping it two weeks before surgery; azathioprine and sulfasalazine can be stopped one day in advance. The drugs can be restarted three days after surgery, assuming stable renal function.13

Anti-TNF-α (adalimumab, etanercept, infliximab), IL1 antagonist (anakinra), and anti-CD20 (rituximab) agents should be stopped one week before surgery and resumed 1-2 weeks afterward, unless risk of complications from disease flareup outweighs the concern for wound infections and dehiscence.14

Herbal Medicines

It is estimated that as much as a third of the U.S. population uses herbal medicines. These substances can result in perioperative hemodynamic instability (ephedra, ginseng, ma huang), hypoglycemia (ginseng), immunosuppression (echinacea, when taken for more than eight weeks), abnormal bleeding (garlic, ginkgo, ginseng), and prolongation of anesthesia (kava, St. John’s wort, valerian). All of these herbal medicine should be stopped one to two weeks before surgery.15,16

Back to the Case

The patient’s Carbidopa/Levodopa should be continued. Celecoxib can be continued if her renal function in stable. If aspirin is taken for a history of coronary artery disease or percutaneous coronary intervention, it should be continued, if possible. Clonidine should be continued or changed to a patch if an extended NPO period is anticipated. Ginkgo, lisinopril, and sulfasalazine should be stopped.

Hospitalization does not provide the luxury of stopping estradiol in advance, so it might be continued with chemical DVT prophylaxis for up to 35 days after surgery. The patient should receive 50-75 mg of IV hydrocortisone during surgery and an additional 25 mg the following day, in addition to her usual prednisone 10 mg/day. She can either receive half her usual NPH dose the morning of surgery with a 5% dextrose infusion in the operating room, or the NPH should be held altogether.

Bottom Line

Perioperative medication use should be tailored to each patient, balancing the risks and benefits of individual drugs. High-quality trials are needed to provide more robust clinical guidelines. TH

Dr. Levin is a hospitalist at the University of Colorado Denver.

References

1. Fleisher LA, Beckman JA, Brown KA, et al. ACC/AHA 2007 Guidelines on Perioperative Cardiovascular Evaluation and Care for Noncardiac Surgery: Executive Summary: A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the 2002 Guidelines on Perioperative Cardiovascular Evaluation for Noncardiac Surgery): Developed in Collaboration With the American Society of Echocardiography, American Society of Nuclear Cardiology, Heart Rhythm Society, Society of Cardiovascular Anesthesiologists, Society for Cardiovascular Angiography and Interventions, Society for Vascular Medicine and Biology, and Society for Vascular Surgery. Circulation. 2007;116(17):1971-1996.
2. Schouten O, Hoeks SE, Welten GM, et al. Effect of statin withdrawal on frequency of cardiac events after vascular surgery. Am J Cardiol. 2007;100(2):316-320.
3. Spell NO III. Stopping and restarting medications in the perioperative period. Med Clin North Am. 2001;85(5):1117-1128.
4. Rosenman DJ, McDonald FS, Ebbert JO, Erwin PJ, LaBella M, Montori VM. Clinical consequences of withholding versus administering renin-angiotensin-aldosterone system antagonists in the preoperative period. J Hosp Med. 2008;3(4):319-325.
5. Arora P, Rajagopalam S, Ranjan R, et al. Preoperative use of angiotensin-converting enzyme inhibitors/ angiotensin receptor blockers is associated with increased risk for acute kidney injury after cardiovascular surgery. Clin J Am Soc Nephrol. 2008;3(5):1266-1273.
6. Benedetto U, Sciarretta S, Roscitano A, et al. Preoperative Angiotensin-converting enzyme inhibitors and acute kidney injury after coronary artery bypass grafting. Ann Thorac Surg. 2008;86(4):1160-1165.
7. Pass SE, Simpson RW. Discontinuation and reinstitution of medications during the perioperative period. Am J Health Syst Pharm. 2004;61(9):899-914.
8. Kohl BA, Schwartz S. Surgery in the patient with endocrine dysfunction. Med Clin North Am. 2009;93(5):1031-1047.
9. Grady D, Wenger NK, Herrington D, et al. Postmenopausal hormone therapy increases risk for venous thromboembolic disease: The Heart and Estrogen/progestin Replacement Study. Ann Intern Med. 2000;132(9):689-696.
10. Hirsh J, Guyatt G, Albers GW, Harrington R, Schünemann HJ; American College of Chest Physicians. Executive summary: American College of Chest Physicians evidence-based clinical practice guidelines (8th edition). Chest. 2008;133(6 Suppl):71S-109S.
11. Axelrod L. Perioperative management of patients treated with glucocorticoids. Endocrinol Metab Clin North Am. 2003;32(2):367-383.
12. Marik PE, Varon J. Requirement of perioperative stress doses of corticosteroids: a systematic review of the literature. Arch Surg. 2008;143(12):1222-1226.
13. Rosandich PA, Kelley JT III, Conn DL. Perioperative management of patients with rheumatoid arthritis in the era of biologic response modifiers. Curr Opin Rheumatol. 2004;16(3):192-198.
14. Saag KG, Teng GG, Patkar NM, et al. American College of Rheumatology 2008 recommendations for the use of nonbiologic and biologic disease-modifying antirheumatic drugs in rheumatoid arthritis. Arthritis Rheum. 2008;59(6):762-784.
15. Ang-Lee MK, Moss J, Yuan CS. Herbal medicines and perioperative care. JAMA. 2001;286(2):208-216.
16. Hodges PJ, Kam PC. The peri-operative implications of herbal medicines. Anaesthesia. 2002;57(9):889-899.

Case

A 72-year-old female with multiple medical problems is admitted with a hip fracture. Surgery is scheduled in 48 hours. The patient’s home medications include aspirin, carbidopa/levodopa, celecoxib, clonidine, estradiol, ginkgo, lisinopril, NPH insulin, sulfasalazine, and prednisone 10 mg a day, which she has been taking for years. How should these and other medications be managed in the perioperative period?

PHOTODISC

Background

Perioperative management of chronic medications is a complex issue, as physicians are required to balance the beneficial and harmful effects of the individual drugs prescribed to their patients. On one hand, cessation of medications can result in decompensation of disease or withdrawal. On the other hand, continuation of drugs can alter metabolism of anesthetic agents, cause perioperative hemodynamic instability, or result in such post-operative complications as acute renal failure, bleeding, infection, and impaired wound healing.

Certain traits make it reasonable to continue medications during the perioperative period. A long elimination half-life or duration of action makes stopping some medications impractical as it takes four to five half-lives to completely clear the drug from the body; holding the drug for a few days around surgery will not appreciably affect its concentration. Stopping drugs that carry severe withdrawal symptoms can be impractical because of the need for lengthy tapers, which can delay surgery and result in decompensation of underlying disease.

Drugs with no significant interactions with anesthesia or risk of perioperative complications should be continued in order to avoid deterioration of the underlying disease. Conversely, drugs that interact with anesthesia or increase risk for complications should be stopped if this can be accomplished safely. Patient-specific factors should receive consideration, as the risk of complications has to be balanced against the danger of exacerbating the underlying disease.

Overview of the Data

The challenge in providing recommendations on perioperative medication management lies in a dearth of high-quality clinical trials. Thus, much of the information comes from case reports, expert opinion, and sound application of pharmacology.

Antiplatelet therapy: Nuances of perioperative antiplatelet therapy are beyond the scope of this review, but some general principles can be elucidated from the American College of Cardiology/American Heart Association (ACC/AHA) 2007 perioperative guidelines.1 Management of antiplatelet therapy should be done in conjunction with the surgical team, as cardiovascular risk has to be weighed against bleeding risk.

Aspirin therapy should be continued in all patients with a history of coronary artery disease (CAD), balloon angioplasty, or percutaneous coronary intervention (PCI), unless the risk of bleeding complications is felt to exceed the cardioprotective benefits—for example, in some neurosurgical patients.1

Clopidogrel therapy is crucial for prevention of in-stent thrombosis (IST) following PCI because patients who experience IST suffer catastrophic myocardial infarctions with high mortality. Ideally, surgery should be delayed to permit completion of clopidogrel therapy—30 to 45 days after implantation of a bare-metal stent and 365 days after a drug-eluting stent. If surgery has to be performed sooner, guidelines recommend operating on dual antiplatelet therapy with aspirin and clopidogrel.1 Again, this course of treatment has to be balanced against the risk of hemorrhagic complications from surgery.

Both aspirin and clopidogrel irreversibly inhibit platelet aggregation. The recovery of normal coagulation involves formation of new platelets, which necessitates cessation of therapy for seven to 10 days before surgery. Platelet inhibition begins within minutes of restarting aspirin and within hours of taking clopidogrel, although attaining peak clopidogrel effect takes three to seven days, unless a loading dose is used.

Cardiovascular Drugs

Beta-blockers in the perioperative setting are a focus of an ongoing debate beyond the scope of this review (see “What Pre-Operative Cardiac Evaluation of Patients Undergoing Intermediate-Risk Surgery Is Most Effective?,” February 2008, p. 26). Given the current evidence and the latest ACC/AHA guidelines, it is still reasonable to continue therapy in patients who are already taking them to avoid precipitating cardiovascular events by withdrawal. Patients with increased cardiac risk, demonstrated by a Revised Cardiac Risk Index (RCRI) score of ≥2 (see Table 1, p. 12), should be considered for beta-blocker therapy before surgery.1 In either case, the dose should be titrated to a heart rate <65 for optimal cardiac protection.1

Statins should be continued if the patient is taking them, especially because preoperative withdrawal has been associated with a 4.6-fold increase in troponin release and a 7.5-fold increased risk of myocardial infarction (MI) and cardiovascular death following major vascular surgery.2 Patients with increased cardiac risk— RCRI ≥1—can be considered for initiation of statin therapy before surgery, although the benefit of this intervention has not been examined in prospective studies.1

Amiodarone has an exceptionally long half-life of up to 142 days. It should be continued in the perioperative period.

Calcium channel blockers (CCBs) can be continued with no unpleasant perioperative hemodynamic effects.1 CCBs have potential cardioprotective benefits.

Clonidine withdrawal can result in severe rebound hypertension with reports of encephalopathy, stroke, and death. These effects are exacerbated by concomitant beta-blocker therapy. For this reason, if a patient is expected to be NPO for more than 12 hours, they should be converted to a clonidine patch 48-72 hours before surgery with concurrent tapering of the oral dose.3

Digoxin has a long half-life (up to 48 hours) and should be continued with monitoring of levels if there is a change in renal function.

Angiotensin converting enzyme inhibitors (ACEI) and angiotensin receptor blockers (ARBs) have been associated with a 50% increased risk of hypotension requiring vasopressors during induction of anesthesia.4 However, it is worth mentioning that this finding has not been corroborated in other studies. A large retrospective cohort of cardiothoracic surgical patients found a 28% increased risk of post-operative acute renal failure (ARF) with both drug classes, although another cardiothoracic report published the same year demonstrated a 50% reduction in risk with ACEIs.5,6 Although the evidence of harm is not unequivocal, perioperative blood-pressure control can be achieved with other drugs without hemodynamic or renal risk, such as CCBs, and in most cases ACEIs/ARBs should be stopped one day before surgery.

Diuretics carry a risk of volume depletion and electrolyte derangements, and should be stopped once a patient becomes NPO. Excess volume is managed with as-needed intravenous formulations.

click for large version

Drugs Acting on the Central Nervous System

The majority of central nervous system (CNS)-active drugs, including antiepileptics, antipsychotics, benzodiazepines, bupropion, gabapentin, lithium, mirtazapine, selective serotonin and norepinephrine reuptake inhibitors (SSRIs and SNRIs), tricyclic antidepressants (TCAs), and valproic acid, balance a low risk of perioperative complications against a significant potential for withdrawal and disease decompensation. Therefore, these medications should be continued.

Carbidopa/Levodopa should be continued because abrupt cessation can precipitate systemic withdrawal resembling serotonin syndrome and rapid deterioration of Parkinson’s symptoms.

Monoamine oxidase inhibitor (MAOI) therapy usually indicates refractory psychiatric illness, so these drugs should be continued to avoid decompensation. Importantly, MAOI-safe anesthesia without dextromethorphan, meperidine, epinephrine, or norepinephrine has to be used due to the risk of cardiovascular instability.7

Diabetic Drugs

Insulin therapy should be continued with adjustments. Glargine basal insulin has no peak and can be continued without changing the dose. Likewise, patients with insulin pumps can continue the usual basal rate. Short-acting insulin or such insulin mixes as 70/30 should be stopped four hours before surgery to avoid hypoglycemia. Intermediate-acting insulin (e.g., NPH) can be administered at half the usual dose the day of surgery with a perioperative 5% dextrose infusion. NPH should not be given the day of surgery if the dextrose infusion cannot be used.8

Incretins (exenatide, sitagliptin) rarely cause hypoglycemia in the absence of insulin and may be beneficial in controlling intraoperative hyperglycemia. Therefore, these medications can be continued.8

Thiazolidinediones (TZDs; pioglitazone, rosiglitazone) alter gene transcription with biological duration of action on the order of weeks and low risk of hypoglycemia, and should be continued.

Metformin carries an FDA black-box warning to discontinue therapy before any intravascular radiocontrast studies or surgical procedures due to the risk of severe lactic acidosis if renal failure develops. It should be stopped 24 hours before surgery and restarted at least 48-72 hours after. Normal renal function should be confirmed before restarting therapy.8

Sulfonylureas (glimepiride, glipizide, glyburide) carry a significant risk of hypoglycemia in a patient who is NPO; they should be stopped the night before surgery or before commencement of NPO status.

Hormones

Antithyroid drugs (methimazole, propylthiouracil) and levothyroxine should be continued, as they have no perioperative contraindications.

Oral contraceptives (OCPs), hormone replacement therapy (HRT), and raloxifene can increase the risk of DVT. The largest study on the topic was the HERS trial of postmenopausal women on estrogen/progesterone HRT. The authors reported a 4.9-fold increased risk of DVT for 90 days after surgery.9 Unfortunately, no information was provided on the types of surgery, or whether appropriate and consistent DVT prophylaxis was utilized. HERS authors also reported a 2.5-fold increased risk of DVT for at least 30 days after cessation of HRT.9

Given the data, it is reasonable to stop hormone therapy four weeks before surgery when prolonged immobilization is anticipated and patients are able to tolerate hormone withdrawal, especially if other DVT risk factors are present. If hormone therapy cannot be stopped, strong consideration should be given to higher-intensity DVT prophylaxis (e.g., chemoprophylaxis as opposed to mechanical measures) of longer duration—up to 28 days following general surgery and up to 35 days after orthopedic procedures.10

click for large version

Perioperative Corticosteroids

Corticosteroid therapy in excess of prednisone 5 mg/day or equivalent for more than five days in the 30 days preceding surgery might predispose patients to acute adrenal insufficiency in the perioperative period. Surgical procedures typically result in cortisol release of 50-150 mg/day, which returns to baseline within 48 hours.11 Therefore, the recommendation is to continue a patient’s baseline steroid dose and supplement it with stress-dose steroids tailored to the severity of operative stress (see Table 2, above).

Mineralocorticoid supplementation is not necessary, because endogenous production is not suppressed by corticosteroid therapy.11 Although a recent systematic review suggests that routine stress-dose steroids might not be indicated, high-quality prospective data are needed before abandoning this strategy due to complications of acute adrenal insufficiency compared to the risk of a brief corticosteroid burst.

Non-Steroidal Anti-Inflammatory Drugs (NSAIDs)

Nonselective cyclooxygenase (COX) inhibitors reversibly decrease platelet aggregation only while the drug is present in the circulation and should be stopped one to three days before surgery due to risk of bleeding.

Selective COX-2 inhibitors do not significantly alter platelet aggregation and can be continued for opioid-sparing perioperative pain control.

Both COX-2-selective and nonselective inhibitors should be held if there are concerns for impaired renal function.

Disease-Modifying Antirheumatic Drugs (DMARDs) and Biological Response Modifiers (BRMs)

Methotrexate increases the risk of wound infections and dehiscence. However, this is offset by a decreased risk of post-operative disease flares with continued use. It can be continued unless the patient has medical comorbidities, advanced age, or chronic therapy with more than 10 mg/day of prednisone, in which case the drug should be stopped two weeks before surgery.12

Azathioprine, leflunomide, and sulfasalazine are renally cleared with a risk of myelosuppression; all of these medications should be stopped. Long half-life of leflunomide necessitates stopping it two weeks before surgery; azathioprine and sulfasalazine can be stopped one day in advance. The drugs can be restarted three days after surgery, assuming stable renal function.13

Anti-TNF-α (adalimumab, etanercept, infliximab), IL1 antagonist (anakinra), and anti-CD20 (rituximab) agents should be stopped one week before surgery and resumed 1-2 weeks afterward, unless risk of complications from disease flareup outweighs the concern for wound infections and dehiscence.14

Herbal Medicines

It is estimated that as much as a third of the U.S. population uses herbal medicines. These substances can result in perioperative hemodynamic instability (ephedra, ginseng, ma huang), hypoglycemia (ginseng), immunosuppression (echinacea, when taken for more than eight weeks), abnormal bleeding (garlic, ginkgo, ginseng), and prolongation of anesthesia (kava, St. John’s wort, valerian). All of these herbal medicine should be stopped one to two weeks before surgery.15,16

Back to the Case

The patient’s Carbidopa/Levodopa should be continued. Celecoxib can be continued if her renal function in stable. If aspirin is taken for a history of coronary artery disease or percutaneous coronary intervention, it should be continued, if possible. Clonidine should be continued or changed to a patch if an extended NPO period is anticipated. Ginkgo, lisinopril, and sulfasalazine should be stopped.

Hospitalization does not provide the luxury of stopping estradiol in advance, so it might be continued with chemical DVT prophylaxis for up to 35 days after surgery. The patient should receive 50-75 mg of IV hydrocortisone during surgery and an additional 25 mg the following day, in addition to her usual prednisone 10 mg/day. She can either receive half her usual NPH dose the morning of surgery with a 5% dextrose infusion in the operating room, or the NPH should be held altogether.

Bottom Line

Perioperative medication use should be tailored to each patient, balancing the risks and benefits of individual drugs. High-quality trials are needed to provide more robust clinical guidelines. TH

Dr. Levin is a hospitalist at the University of Colorado Denver.

References

1. Fleisher LA, Beckman JA, Brown KA, et al. ACC/AHA 2007 Guidelines on Perioperative Cardiovascular Evaluation and Care for Noncardiac Surgery: Executive Summary: A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the 2002 Guidelines on Perioperative Cardiovascular Evaluation for Noncardiac Surgery): Developed in Collaboration With the American Society of Echocardiography, American Society of Nuclear Cardiology, Heart Rhythm Society, Society of Cardiovascular Anesthesiologists, Society for Cardiovascular Angiography and Interventions, Society for Vascular Medicine and Biology, and Society for Vascular Surgery. Circulation. 2007;116(17):1971-1996.
2. Schouten O, Hoeks SE, Welten GM, et al. Effect of statin withdrawal on frequency of cardiac events after vascular surgery. Am J Cardiol. 2007;100(2):316-320.
3. Spell NO III. Stopping and restarting medications in the perioperative period. Med Clin North Am. 2001;85(5):1117-1128.
4. Rosenman DJ, McDonald FS, Ebbert JO, Erwin PJ, LaBella M, Montori VM. Clinical consequences of withholding versus administering renin-angiotensin-aldosterone system antagonists in the preoperative period. J Hosp Med. 2008;3(4):319-325.
5. Arora P, Rajagopalam S, Ranjan R, et al. Preoperative use of angiotensin-converting enzyme inhibitors/ angiotensin receptor blockers is associated with increased risk for acute kidney injury after cardiovascular surgery. Clin J Am Soc Nephrol. 2008;3(5):1266-1273.
6. Benedetto U, Sciarretta S, Roscitano A, et al. Preoperative Angiotensin-converting enzyme inhibitors and acute kidney injury after coronary artery bypass grafting. Ann Thorac Surg. 2008;86(4):1160-1165.
7. Pass SE, Simpson RW. Discontinuation and reinstitution of medications during the perioperative period. Am J Health Syst Pharm. 2004;61(9):899-914.
8. Kohl BA, Schwartz S. Surgery in the patient with endocrine dysfunction. Med Clin North Am. 2009;93(5):1031-1047.
9. Grady D, Wenger NK, Herrington D, et al. Postmenopausal hormone therapy increases risk for venous thromboembolic disease: The Heart and Estrogen/progestin Replacement Study. Ann Intern Med. 2000;132(9):689-696.
10. Hirsh J, Guyatt G, Albers GW, Harrington R, Schünemann HJ; American College of Chest Physicians. Executive summary: American College of Chest Physicians evidence-based clinical practice guidelines (8th edition). Chest. 2008;133(6 Suppl):71S-109S.
11. Axelrod L. Perioperative management of patients treated with glucocorticoids. Endocrinol Metab Clin North Am. 2003;32(2):367-383.
12. Marik PE, Varon J. Requirement of perioperative stress doses of corticosteroids: a systematic review of the literature. Arch Surg. 2008;143(12):1222-1226.
13. Rosandich PA, Kelley JT III, Conn DL. Perioperative management of patients with rheumatoid arthritis in the era of biologic response modifiers. Curr Opin Rheumatol. 2004;16(3):192-198.
14. Saag KG, Teng GG, Patkar NM, et al. American College of Rheumatology 2008 recommendations for the use of nonbiologic and biologic disease-modifying antirheumatic drugs in rheumatoid arthritis. Arthritis Rheum. 2008;59(6):762-784.
15. Ang-Lee MK, Moss J, Yuan CS. Herbal medicines and perioperative care. JAMA. 2001;286(2):208-216.
16. Hodges PJ, Kam PC. The peri-operative implications of herbal medicines. Anaesthesia. 2002;57(9):889-899.