Kate Jennings, MD

Clinical question: What is the current level of continuity of care, and what factors affect continuity of care in the hospital setting? Has this changed with increasing use of hospitalists and limits on residency duty hours?

Background: Outpatient continuity of care leads to lower costs, better quality of life, and less emergency room use. Recent changes in residency hours have increased hand-offs and decreased inpatient continuity, but to what extent is unknown.

Study design: Retrospective cohort of 5% of Medicare claims data (530,000 patients in all) from 1996 to 2006, including patients admitted for COPD, congestive heart failure, or pneumonia who were cared for by a general internist or family practitioner.

Setting: Nationwide in the U.S.

Synopsis: The authors defined patients as having a primary-care physician (PCP) if they had three billed visits with the PCP in the last year, hospitalists as those who derived at least 90% of their Medicare claims from inpatient billing, and other generalists as those who met criteria as a generalist but did not fit these categories. Inpatient continuity of care decreased to 59% of patients seeing a single physician in the hospital in 2006 from 71% in 1996, with an accompanying decrease in the length of stay of one full day. There were large variations by geographic region, population size, and hospital characteristics. Patients cared for by hospitalists had slightly better continuity of care than those cared for by nonhospitalist generalists, and those who were cared for by both hospitalists and nonhospitalist generalists had the worst continuity of care. Having a PCP was associated with increased discontinuity of care as an inpatient, perhaps because of individual members of a practice rounding on all of the practice’s inpatients.

Bottom line: Patients were 5% less likely per year between 1996 and 2006 to have a single physician be their primary caregiver in the hospital, but the rise of the hospitalist movement does not seem to be the cause.

Citation: Fletcher KE, Sharma G, Zhang D, Kuo YF, Goodwin JS. Trends in inpatient continuity of care for a cohort of Medicare patients 1996-2006. J Hosp Med. 2011;6:441-447.

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

Clinical question: What is the current level of continuity of care, and what factors affect continuity of care in the hospital setting? Has this changed with increasing use of hospitalists and limits on residency duty hours?

Background: Outpatient continuity of care leads to lower costs, better quality of life, and less emergency room use. Recent changes in residency hours have increased hand-offs and decreased inpatient continuity, but to what extent is unknown.

Study design: Retrospective cohort of 5% of Medicare claims data (530,000 patients in all) from 1996 to 2006, including patients admitted for COPD, congestive heart failure, or pneumonia who were cared for by a general internist or family practitioner.

Setting: Nationwide in the U.S.

Synopsis: The authors defined patients as having a primary-care physician (PCP) if they had three billed visits with the PCP in the last year, hospitalists as those who derived at least 90% of their Medicare claims from inpatient billing, and other generalists as those who met criteria as a generalist but did not fit these categories. Inpatient continuity of care decreased to 59% of patients seeing a single physician in the hospital in 2006 from 71% in 1996, with an accompanying decrease in the length of stay of one full day. There were large variations by geographic region, population size, and hospital characteristics. Patients cared for by hospitalists had slightly better continuity of care than those cared for by nonhospitalist generalists, and those who were cared for by both hospitalists and nonhospitalist generalists had the worst continuity of care. Having a PCP was associated with increased discontinuity of care as an inpatient, perhaps because of individual members of a practice rounding on all of the practice’s inpatients.

Bottom line: Patients were 5% less likely per year between 1996 and 2006 to have a single physician be their primary caregiver in the hospital, but the rise of the hospitalist movement does not seem to be the cause.

Citation: Fletcher KE, Sharma G, Zhang D, Kuo YF, Goodwin JS. Trends in inpatient continuity of care for a cohort of Medicare patients 1996-2006. J Hosp Med. 2011;6:441-447.

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

Clinical question: What is the current level of continuity of care, and what factors affect continuity of care in the hospital setting? Has this changed with increasing use of hospitalists and limits on residency duty hours?

Background: Outpatient continuity of care leads to lower costs, better quality of life, and less emergency room use. Recent changes in residency hours have increased hand-offs and decreased inpatient continuity, but to what extent is unknown.

Study design: Retrospective cohort of 5% of Medicare claims data (530,000 patients in all) from 1996 to 2006, including patients admitted for COPD, congestive heart failure, or pneumonia who were cared for by a general internist or family practitioner.

Setting: Nationwide in the U.S.

Synopsis: The authors defined patients as having a primary-care physician (PCP) if they had three billed visits with the PCP in the last year, hospitalists as those who derived at least 90% of their Medicare claims from inpatient billing, and other generalists as those who met criteria as a generalist but did not fit these categories. Inpatient continuity of care decreased to 59% of patients seeing a single physician in the hospital in 2006 from 71% in 1996, with an accompanying decrease in the length of stay of one full day. There were large variations by geographic region, population size, and hospital characteristics. Patients cared for by hospitalists had slightly better continuity of care than those cared for by nonhospitalist generalists, and those who were cared for by both hospitalists and nonhospitalist generalists had the worst continuity of care. Having a PCP was associated with increased discontinuity of care as an inpatient, perhaps because of individual members of a practice rounding on all of the practice’s inpatients.

Bottom line: Patients were 5% less likely per year between 1996 and 2006 to have a single physician be their primary caregiver in the hospital, but the rise of the hospitalist movement does not seem to be the cause.

Citation: Fletcher KE, Sharma G, Zhang D, Kuo YF, Goodwin JS. Trends in inpatient continuity of care for a cohort of Medicare patients 1996-2006. J Hosp Med. 2011;6:441-447.

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

In This Edition

Literature At A Glance

A guide to this month’s studies

1. Higher loading dose of clopidogrel in STEMI
2. Early vs. late surgery following hip fracture
3. Beta-blockers in chronic kidney disease
4. Long-term azithromycin in COPD
5. CT screening for lung cancer
6. Timing of parenteral nutrition in the ICU
7. Intrapleural management of empyema with DNase and t-PA
8. Effect of weekend elective admissions on hospital flow
9. Expectations and outcomes of medical comanagement

Higher-Dose Clopidogrel Improves Outcomes at 30 Days in STEMI Patients

Clinical question: Does a 600-mg loading dose of clopidogrel given before percutaneous coronary intervention (PCI) in patients with an ST-segment elevation myocardial infarction (STEMI) provide more protection from thrombotic complications at 30 days than a 300-mg dose?

Background: Clopidogrel at 600 mg is active more quickly (two hours versus 12 hours) and inhibits platelets more completely than does a 300-mg dose, but it has never been tested prospectively in patients undergoing percutaneous intervention with a STEMI.

Study design: Prospective, multicenter, randomized controlled trial.

Setting: Five hospitals in Italy, Belgium, Serbia, and Hungary.

Synopsis: Two-hundred-one patients with STEMI were randomized to either 600 mg or 300 mg of clopidogrel, given an average of 30 minutes before initial PCI, as well as other standard treatment for STEMI. The primary outcome was “infarct size,” judged as the area under the curve (AUC) of serial creatine kinase-MB (CK-MB) and troponin measurements. At 30 days, patients who received the 600-mg dose of clopidogrel had lower AUCs of cardiac biomarkers, statistically significant (though clinically small) improvement in left ventricle ejection fraction at discharge, lower incidence of severely impaired post-PCI thrombolitis in myocardial infarction (TIMI) flow, and fewer “major cardiovascular events,” driven mainly by a reduction in revascularizations. Measurement of biomarkers to calculate infarct size can be confounded by hypertrophy, and the trial was underpowered for cardiovascular events. However, there was no increase in bleeding events.

Bottom line: In patients with STEMI, treatment with a higher loading dose of clopidogrel before PCI reduces revascularizations and might decrease infarct size without an increase in adverse events.

Citation: Patti G, Bárczi G, Orlic D, et al. Outcome comparison of 600- and 300-mg loading doses of clopidogrel in patients undergoing primary percutaneous coronary intervention for ST-segment elevation myocardial infarction: results from the ARMYDA-6 MI (Antiplatelet therapy for Reduction of MYocardial Damage during Angioplasty-Myocardial Infarction) randomized study. J Am Coll Cardiol. 2011;58:1592-1599.

Medical Comorbidities Explain Most of Excess Risk Seen in Patients with Delayed Hip Fracture Repair

Clinical question: Could the increased mortality found with delays in hip fracture surgery be confounded by the premorbid functional status and medical comorbidities of patients whose surgery is more likely to be delayed?

Background: Guidelines recommend operating on patients with hip fracture within 24 hours, but the supporting evidence has not adjusted for underlying medical comorbidities, which could delay surgery and contribute to poor outcomes, making delays look harmful.

Study design: Prospective cohort, single-center design.

Setting: University hospital in Spain.

Synopsis: The study included 2,250 consecutive elderly patients admitted to the hospital for hip fracture who had their functional status and medical comorbidities assessed at enrollment. If their surgery was delayed beyond 24 hours, the reason was sought. Medical outcomes assessed daily while in hospital were delirium, pneumonia, heart failure, urinary tract infection, and new pressure sores, while the dose of pain medication, surgical complications, and in-hospital mortality were also compiled. No post-discharge data were available.

The median time to surgery was 72 hours. Patients with more medical comorbidities and poorer preoperative functional status had longer delays to operation, most commonly due to interrupting antiplatelet treatment or need for preoperative “echocardiography or other examinations.” When these medical factors were included in logistic regression analysis, the increased mortality seen with delays of surgery in the cohort was no longer statistically significant, suggesting the underlying comorbidities of these patients, rather than the delay to surgery itself, explained the increased mortality.

Bottom line: Delaying hip fracture surgery is less important than preoperative functional status and medical comorbidity in contributing to poor outcomes.

Citation: Vidán MT, Sánchez E, Gracia Y, Marañón E, Vaquero J, Serra JA. Causes and effects of surgical delay in patients with hip fracture: a cohort study. Ann Intern Med. 2011;155:226-233.

Beta-Blockers Decrease All-Cause and Cardiovascular Mortality in Patients with Chronic Kidney Disease

Clinical question: Are beta-blockers as effective in patients on dialysis and with end-stage chronic kidney disease (CKD) as they are in patients without severe renal disease?

Background: Patients with CKD have been largely excluded from trials of beta-blocker therapy even though they have high rates of cardiovascular disease and might be most likely to benefit. However, patients on dialysis might be predisposed to adverse complications of beta-blocker therapy, including hypotension.

Study design: Meta-analysis of eight trials of beta-blockade versus placebo (six heart failure trials) or ACE-I (two non-heart-failure trials) that included post-hoc analyses of CKD patients.

Setting: Varied, usually multinational RCTs.

Synopsis: The six congestive heart failure (CHF) studies were not designed to evaluate patients with CKD, and the two non-CHF studies were intended to evaluate progression of CKD, not cardiac outcomes. Thus, this is a meta-analysis of post-hoc CKD subgroups included in these trials. Compared with placebo, beta-blockers reduced all-cause and cardiovascular mortality without significant heterogeneity between trials. The magnitude of the effect was similar in CKD and non-CKD patients. Patients with CKD treated with beta-blockers were at increased risk of bradycardia and hypotension, but this did not lead to increased discontinuation of the drug. Only 114 dialysis patients were included in one of the eight trials (7,000 patients overall) and no outcomes were assessed.

Bottom line: Beta-blockers lower all-cause and cardiovascular mortality in patients with CKD similarly to patients without kidney disease but are associated with an increased risk for hypotension and bradycardia. Their effect in dialysis patients is unknown.

Citation: Badve SV, Roberts MA, Hawley CM, et al. Effects of beta-adrenergic antagonists in patients with chronic kidney disease: a systematic review and meta-analysis. J Am Coll Cardiol. 2011;58:1152-1161.

Daily Long-Term Azithromycin in COPD Patients Decreases Frequency of Exacerbations

Clinical question: Does long-term treatment with azithromycin decrease COPD exacerbations and improve quality of life with an acceptable risk profile?

Background: Patients with acute exacerbations of COPD have increased risks of death and more rapid decline in lung function. Macrolide antibiotics might decrease exacerbations via antibacterial and anti-inflammatory effects. Small studies of long-term treatment with macrolides in COPD have had conflicting results.

Study design: Prospective, multicenter, parallel-group, placebo-controlled trial.

Setting: Twelve U.S. academic health centers.

Synopsis: Eligible patients were those older than 40 with pulmonary function test-proven obstructive disease, a significant smoking history, a use of supplemental oxygen, or an oral glucocorticoid treatment in the previous year, and those who had previously come to healthcare attention for a COPD exacerbation. They were randomized to daily azithromycin (250 mg) or placebo. The primary outcome was time to first COPD exacerbation at one year. The investigators used deep nasopharyngeal swabs to evaluate the development of microbial resistance, and all patients were regularly screened for hearing loss. Time to first exacerbation and frequency of exacerbations decreased significantly with azithromycin treatment; the number needed to treat to prevent a COPD exacerbation was 2.86. More than half of nasopharyngeal samples had evidence of macrolide resistance at enrollment, and this increased to more than 80% at one year. Audiometric testing did show a decrease in hearing in those treated with azithromycin, but even those who did not stop their medication had recovery on subsequent testing.

Bottom line: Daily azithromycin decreases COPD exacerbations in those with a history of previous exacerbation at the cost of increased macrolide resistance and possible reversible decrements in hearing. The high degree of macrolide resistance at enrollment suggests non-antibacterial mechanisms might be responsible.

Citation: Albert RK, Connett J, Bailey WC, et al. Azithromycin for prevention of exacerbations of COPD. N Engl J Med. 2011;365:689-698.

Serial Low-Dose CT Scans Decrease Lung Cancer Mortality

Clinical question: Does annual screening with low-dose CT scans reduce mortality from lung cancer in current or former heavy smokers?

Background: Lung cancer is still the top cause of cancer death in the United States, and its prevalence is increasing in the developing world. Prior trials of screening chest radiography and sputum samples have not shown a decrease in lung cancer mortality, but CT scans could help identify lung cancer at an earlier, more treatable stage.

Study design: Prospective, multicenter, randomized trial of low-dose (roughly one-fifth the radiation of traditional CT) annual CT versus annual chest radiography.

Setting: Thirty-three U.S. medical institutions.

Synopsis: The study enrolled 53,454 current or prior smokers with a history of smoking more than 30 packs a year who were randomly assigned to undergo three annual CT scans or chest X-rays. Any CT scans that showed a noncalcified nodule greater than 4 mm and chest X-rays that showed a noncalcified nodule or mass of any size were classified as “positive,” but follow-up was left to the discretion of the treating physicians. Participants were followed for 6.5 years on average.

Of those screened with CT, 40% had a positive screening test at some point, more than three times the rate of chest radiography, and 96.4% of these were false-positives, which was similar to chest radiography. Adverse events as a result of workup of eventual false positives were uncommon, occurring in around 1% of those who did not have lung cancer and in 11% of those who did. Screening reduced lung cancer mortality by 20% compared with chest radiography, with a number needed to screen to prevent one lung cancer death of 320. Cost-effectiveness and effects of radiation were not assessed.

Bottom line: Serial screening with low-dose CT reduces lung cancer mortality at the cost of a high rate of false-positives. Questions remain about cost-effectiveness and radiation exposure.

Citation: National Lung Screening Trial Research Team, Aberle DR, Adams AM, Berg CD, et al. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med. 2011;365:395-409.

One-Week Delay in Starting Parenteral Nutrition in ICU Patients Is Associated with Better Outcomes

Clinical question: In patients admitted to the ICU who are at risk for malnutrition, does supplementing enteral feeding with parenteral feeding on the day of admission improve outcomes when compared with supplementation starting on Day 8?

Background: Patients with critical illness are at risk for malnutrition, which may lead to worsened outcomes. Many are unable to tolerate enteral feeding. However, adding parenteral nutrition has risks, including overfeeding and hyperglycemia.

Study design: Prospective, multicenter, randomized controlled trial. Neither patients/families nor ICU physicians were blinded.

Setting: Two university hospitals in Belgium.

Synopsis: The study included 4,640 patients admitted to the ICU who were “at risk” for malnutrition (assessed using a Nutritional Risk Score of >3). They were randomized to early initiation of parenteral nutrition on Day 1 or late initiation of parenteral nutrition on Day 8, in both cases only if needed to supplement enteral feeding. Both groups received intensive IV insulin protocols to keep blood glucose at 80 mg/dL to 110 mg/dL. The study populations were well-matched, with similar APACHE II scores. Of the total patients, 42% were emergency admissions to the ICU (more than half of which had sepsis), while scheduled cardiac surgery patients made up the majority of the rest of the study participants.

Patients with late initiation of total parenteral nutrition (TPN) were more likely to be discharged alive from the ICU within eight days despite having increased risk of hypoglycemia and more elevation of inflammatory markers. They had shorter ICU and hospital stays, fewer days on the ventilator, fewer infectious complications, were less likely to develop renal failure, and had lower overall costs. The unblinded management of these patients raises concern in interpreting infectious outcomes because study investigators could have been biased to look for infection more often in the early initiation group.

Bottom line: In ICU patients at risk for malnutrition, delaying initiation of TPN to supplement enteral feeding shortens ICU stay and reduces infectious complications when compared with early initiation of TPN.

Citation: Casaer MP, Mesotten D, Hermans G, et al. Early versus late parenteral nutrition in critically ill adults. N Engl J Med. 2011;365:506-17.

Intrapleural Instillation of Combined DNase and t-PA Improve Outcomes in Patients with Empyema

Clinical question: Does the intrapleural administration of a fibrinolytic, a recombinant DNase, or a combination of the two improve outcomes in patients with pleural infections compared with placebo?

Background: Pleural-based infections confer high morbidity and mortality, especially in the one-third of cases in which chest tube drainage fails. Observational data suggest intrapleural administration of fibrinolytics can improve drainage; however, a large randomized trial (MIST1) failed to show improvement with streptokinase.

Study design: Randomized, double-blind, double-dummy, factorial, multicenter trial.

Setting: University hospitals in the U.K.

Synopsis: A total of 193 patients with clinical evidence of an infection as well as laboratory evidence of pleural infection were randomized in factorial design to placebo, DNase alone, t-PA alone, and DNase plus t-PA as a twice-daily, hourlong instillation into the pleural space on hospital days 1 through 3. The combination group had improvements in the size of the pleural effusion compared with placebo, the primary endpoint. This group also was less likely to be referred for surgery, had shorter hospital stays, had less fever, and had lower inflammatory markers by hospital Day 7. Neither of the single-agent groups were better than placebo. The confidence intervals for all outcomes besides radiographic size of the effusion were large, as the trial had limited power for secondary endpoints.

Bottom line: In patients with empyema, the addition of twice-daily instillations of DNase and tPA can improve drainage and decrease risk of treatment failure when combined with chest tube drainage and antibiotics.

Citation: Rahman NM, Maskell NA, West A, et al. Intrapleural use of tissue plasminogen activator and DNase in pleural infection. N Engl J Med. 2011;365:518-526.

Smoothing Admissions over the Week Improves Patient Flow

Clinical question: What is the difference between weekday and weekend hospital occupancy, and what might the effect be of redistributing, or “smoothing,” elective admissions across the week?

Background: Efforts to improve patient flow have largely focused on reducing the average length of stay (ALOS). As the ALOS decreases, though, further reductions have limited yield in improving patient flow and could occur at the expense of patient safety. Smoothing admissions is a recognized but underutilized tool to address patient flow and hospital overcrowding.

Study design: Retrospective cohort.

Setting: Thirty-nine freestanding, tertiary-care children’s hospitals in the U.S.

Synopsis: Hospital occupancy ranged from 70.9% to 108.1% for weekdays, and from 65.7% to 94.4% on weekends. The mean difference between weekday and weekend occupancy was 8.2%. Using a mathematical model to redistribute admissions from peak days to nonpeak days (within a one-week time frame), the investigators found that percent occupancy on average dropped by 6.6%—the same number of patients over the same time interval, but with lower average occupancy. And, while not projected by these authors, the correlate of lower average occupancy would be reduced waits and delays for admission and within-hospital services. Only 2.6%, or about 7.5 admissions per week per hospital, would need to be rescheduled in order to realize this gain.

Bottom line: Where feasible, reshuffling elective admissions to “smooth” demand across the week is associated with improved hospital flow.

Citation: Fieldston ES, Hall M, Shah SS, et al. Addressing inpatient crowding by smoothing occupancy at children’s hospitals. J Hosp Med. 2011;6:466-473.

Variability in Preferences, Experiences, and Expectations for Hospitalist Roles in Comanagement

Clinical question: Do provider expectations and experiences in comanagement differ from the traditional medical consultation model?

Background: Hospitalists participate in traditional medical consultation and, increasingly, a variety of comanagement with surgical and various medical specialists. It is uncertain what preferences and expectations on either side of the comanagement relationship might be. Learning more might lead to a better conceptual understanding and working definition of inpatient comanagement.

Study design: Baseline and follow-up surveys.

Setting: Large single-site academic medical center, hospitalist-hepatologist comanagement service.

Synopsis: All 41 respondents completed the baseline survey, with 32 respondents completing a total of 177 follow-up surveys. Respondents included hospitalists, hepatologists, nonphysician providers (NPPs), and GI fellows. Hospitalists were more likely than hepatologists to want to be informed about every patient care decision (83% versus 17%). The majority of respondents reported understanding their roles. At baseline, nearly all agreed that “comanagement tends to improve patient care.” However, at follow-up, fewer NPP and fewer GI fellows reported that patients on their service had received “better care than they would without comanagement.” At follow-up, fewer hospitalists than at baseline reported that there was one “physician leader to direct the overall management of the patient’s hospital course.”

Bottom line: Hospitalists were more likely than hepatologists, within a comanagement relationship, to want to be informed about every decision. Most expected—but fewer actually later reported—that the comanagement model improved patient care.

Citation: Hinami K, Whelan CT, Konetzka RT, and Meltzer DO. Provider expectations and experiences of comanagement. J Hosp Med. 2011;6:401-404.

In This Edition

Literature At A Glance

A guide to this month’s studies

1. Higher loading dose of clopidogrel in STEMI
2. Early vs. late surgery following hip fracture
3. Beta-blockers in chronic kidney disease
4. Long-term azithromycin in COPD
5. CT screening for lung cancer
6. Timing of parenteral nutrition in the ICU
7. Intrapleural management of empyema with DNase and t-PA
8. Effect of weekend elective admissions on hospital flow
9. Expectations and outcomes of medical comanagement

Higher-Dose Clopidogrel Improves Outcomes at 30 Days in STEMI Patients

Clinical question: Does a 600-mg loading dose of clopidogrel given before percutaneous coronary intervention (PCI) in patients with an ST-segment elevation myocardial infarction (STEMI) provide more protection from thrombotic complications at 30 days than a 300-mg dose?

Background: Clopidogrel at 600 mg is active more quickly (two hours versus 12 hours) and inhibits platelets more completely than does a 300-mg dose, but it has never been tested prospectively in patients undergoing percutaneous intervention with a STEMI.

Study design: Prospective, multicenter, randomized controlled trial.

Setting: Five hospitals in Italy, Belgium, Serbia, and Hungary.

Synopsis: Two-hundred-one patients with STEMI were randomized to either 600 mg or 300 mg of clopidogrel, given an average of 30 minutes before initial PCI, as well as other standard treatment for STEMI. The primary outcome was “infarct size,” judged as the area under the curve (AUC) of serial creatine kinase-MB (CK-MB) and troponin measurements. At 30 days, patients who received the 600-mg dose of clopidogrel had lower AUCs of cardiac biomarkers, statistically significant (though clinically small) improvement in left ventricle ejection fraction at discharge, lower incidence of severely impaired post-PCI thrombolitis in myocardial infarction (TIMI) flow, and fewer “major cardiovascular events,” driven mainly by a reduction in revascularizations. Measurement of biomarkers to calculate infarct size can be confounded by hypertrophy, and the trial was underpowered for cardiovascular events. However, there was no increase in bleeding events.

Bottom line: In patients with STEMI, treatment with a higher loading dose of clopidogrel before PCI reduces revascularizations and might decrease infarct size without an increase in adverse events.

Citation: Patti G, Bárczi G, Orlic D, et al. Outcome comparison of 600- and 300-mg loading doses of clopidogrel in patients undergoing primary percutaneous coronary intervention for ST-segment elevation myocardial infarction: results from the ARMYDA-6 MI (Antiplatelet therapy for Reduction of MYocardial Damage during Angioplasty-Myocardial Infarction) randomized study. J Am Coll Cardiol. 2011;58:1592-1599.

Medical Comorbidities Explain Most of Excess Risk Seen in Patients with Delayed Hip Fracture Repair

Clinical question: Could the increased mortality found with delays in hip fracture surgery be confounded by the premorbid functional status and medical comorbidities of patients whose surgery is more likely to be delayed?

Background: Guidelines recommend operating on patients with hip fracture within 24 hours, but the supporting evidence has not adjusted for underlying medical comorbidities, which could delay surgery and contribute to poor outcomes, making delays look harmful.

Study design: Prospective cohort, single-center design.

Setting: University hospital in Spain.

Synopsis: The study included 2,250 consecutive elderly patients admitted to the hospital for hip fracture who had their functional status and medical comorbidities assessed at enrollment. If their surgery was delayed beyond 24 hours, the reason was sought. Medical outcomes assessed daily while in hospital were delirium, pneumonia, heart failure, urinary tract infection, and new pressure sores, while the dose of pain medication, surgical complications, and in-hospital mortality were also compiled. No post-discharge data were available.

The median time to surgery was 72 hours. Patients with more medical comorbidities and poorer preoperative functional status had longer delays to operation, most commonly due to interrupting antiplatelet treatment or need for preoperative “echocardiography or other examinations.” When these medical factors were included in logistic regression analysis, the increased mortality seen with delays of surgery in the cohort was no longer statistically significant, suggesting the underlying comorbidities of these patients, rather than the delay to surgery itself, explained the increased mortality.

Bottom line: Delaying hip fracture surgery is less important than preoperative functional status and medical comorbidity in contributing to poor outcomes.

Citation: Vidán MT, Sánchez E, Gracia Y, Marañón E, Vaquero J, Serra JA. Causes and effects of surgical delay in patients with hip fracture: a cohort study. Ann Intern Med. 2011;155:226-233.

Beta-Blockers Decrease All-Cause and Cardiovascular Mortality in Patients with Chronic Kidney Disease

Clinical question: Are beta-blockers as effective in patients on dialysis and with end-stage chronic kidney disease (CKD) as they are in patients without severe renal disease?

Background: Patients with CKD have been largely excluded from trials of beta-blocker therapy even though they have high rates of cardiovascular disease and might be most likely to benefit. However, patients on dialysis might be predisposed to adverse complications of beta-blocker therapy, including hypotension.

Study design: Meta-analysis of eight trials of beta-blockade versus placebo (six heart failure trials) or ACE-I (two non-heart-failure trials) that included post-hoc analyses of CKD patients.

Setting: Varied, usually multinational RCTs.

Synopsis: The six congestive heart failure (CHF) studies were not designed to evaluate patients with CKD, and the two non-CHF studies were intended to evaluate progression of CKD, not cardiac outcomes. Thus, this is a meta-analysis of post-hoc CKD subgroups included in these trials. Compared with placebo, beta-blockers reduced all-cause and cardiovascular mortality without significant heterogeneity between trials. The magnitude of the effect was similar in CKD and non-CKD patients. Patients with CKD treated with beta-blockers were at increased risk of bradycardia and hypotension, but this did not lead to increased discontinuation of the drug. Only 114 dialysis patients were included in one of the eight trials (7,000 patients overall) and no outcomes were assessed.

Bottom line: Beta-blockers lower all-cause and cardiovascular mortality in patients with CKD similarly to patients without kidney disease but are associated with an increased risk for hypotension and bradycardia. Their effect in dialysis patients is unknown.

Citation: Badve SV, Roberts MA, Hawley CM, et al. Effects of beta-adrenergic antagonists in patients with chronic kidney disease: a systematic review and meta-analysis. J Am Coll Cardiol. 2011;58:1152-1161.

Daily Long-Term Azithromycin in COPD Patients Decreases Frequency of Exacerbations

Clinical question: Does long-term treatment with azithromycin decrease COPD exacerbations and improve quality of life with an acceptable risk profile?

Background: Patients with acute exacerbations of COPD have increased risks of death and more rapid decline in lung function. Macrolide antibiotics might decrease exacerbations via antibacterial and anti-inflammatory effects. Small studies of long-term treatment with macrolides in COPD have had conflicting results.

Study design: Prospective, multicenter, parallel-group, placebo-controlled trial.

Setting: Twelve U.S. academic health centers.

Synopsis: Eligible patients were those older than 40 with pulmonary function test-proven obstructive disease, a significant smoking history, a use of supplemental oxygen, or an oral glucocorticoid treatment in the previous year, and those who had previously come to healthcare attention for a COPD exacerbation. They were randomized to daily azithromycin (250 mg) or placebo. The primary outcome was time to first COPD exacerbation at one year. The investigators used deep nasopharyngeal swabs to evaluate the development of microbial resistance, and all patients were regularly screened for hearing loss. Time to first exacerbation and frequency of exacerbations decreased significantly with azithromycin treatment; the number needed to treat to prevent a COPD exacerbation was 2.86. More than half of nasopharyngeal samples had evidence of macrolide resistance at enrollment, and this increased to more than 80% at one year. Audiometric testing did show a decrease in hearing in those treated with azithromycin, but even those who did not stop their medication had recovery on subsequent testing.

Bottom line: Daily azithromycin decreases COPD exacerbations in those with a history of previous exacerbation at the cost of increased macrolide resistance and possible reversible decrements in hearing. The high degree of macrolide resistance at enrollment suggests non-antibacterial mechanisms might be responsible.

Citation: Albert RK, Connett J, Bailey WC, et al. Azithromycin for prevention of exacerbations of COPD. N Engl J Med. 2011;365:689-698.

Serial Low-Dose CT Scans Decrease Lung Cancer Mortality

Clinical question: Does annual screening with low-dose CT scans reduce mortality from lung cancer in current or former heavy smokers?

Background: Lung cancer is still the top cause of cancer death in the United States, and its prevalence is increasing in the developing world. Prior trials of screening chest radiography and sputum samples have not shown a decrease in lung cancer mortality, but CT scans could help identify lung cancer at an earlier, more treatable stage.

Study design: Prospective, multicenter, randomized trial of low-dose (roughly one-fifth the radiation of traditional CT) annual CT versus annual chest radiography.

Setting: Thirty-three U.S. medical institutions.

Synopsis: The study enrolled 53,454 current or prior smokers with a history of smoking more than 30 packs a year who were randomly assigned to undergo three annual CT scans or chest X-rays. Any CT scans that showed a noncalcified nodule greater than 4 mm and chest X-rays that showed a noncalcified nodule or mass of any size were classified as “positive,” but follow-up was left to the discretion of the treating physicians. Participants were followed for 6.5 years on average.

Of those screened with CT, 40% had a positive screening test at some point, more than three times the rate of chest radiography, and 96.4% of these were false-positives, which was similar to chest radiography. Adverse events as a result of workup of eventual false positives were uncommon, occurring in around 1% of those who did not have lung cancer and in 11% of those who did. Screening reduced lung cancer mortality by 20% compared with chest radiography, with a number needed to screen to prevent one lung cancer death of 320. Cost-effectiveness and effects of radiation were not assessed.

Bottom line: Serial screening with low-dose CT reduces lung cancer mortality at the cost of a high rate of false-positives. Questions remain about cost-effectiveness and radiation exposure.

Citation: National Lung Screening Trial Research Team, Aberle DR, Adams AM, Berg CD, et al. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med. 2011;365:395-409.

One-Week Delay in Starting Parenteral Nutrition in ICU Patients Is Associated with Better Outcomes

Clinical question: In patients admitted to the ICU who are at risk for malnutrition, does supplementing enteral feeding with parenteral feeding on the day of admission improve outcomes when compared with supplementation starting on Day 8?

Background: Patients with critical illness are at risk for malnutrition, which may lead to worsened outcomes. Many are unable to tolerate enteral feeding. However, adding parenteral nutrition has risks, including overfeeding and hyperglycemia.

Study design: Prospective, multicenter, randomized controlled trial. Neither patients/families nor ICU physicians were blinded.

Setting: Two university hospitals in Belgium.

Synopsis: The study included 4,640 patients admitted to the ICU who were “at risk” for malnutrition (assessed using a Nutritional Risk Score of >3). They were randomized to early initiation of parenteral nutrition on Day 1 or late initiation of parenteral nutrition on Day 8, in both cases only if needed to supplement enteral feeding. Both groups received intensive IV insulin protocols to keep blood glucose at 80 mg/dL to 110 mg/dL. The study populations were well-matched, with similar APACHE II scores. Of the total patients, 42% were emergency admissions to the ICU (more than half of which had sepsis), while scheduled cardiac surgery patients made up the majority of the rest of the study participants.

Patients with late initiation of total parenteral nutrition (TPN) were more likely to be discharged alive from the ICU within eight days despite having increased risk of hypoglycemia and more elevation of inflammatory markers. They had shorter ICU and hospital stays, fewer days on the ventilator, fewer infectious complications, were less likely to develop renal failure, and had lower overall costs. The unblinded management of these patients raises concern in interpreting infectious outcomes because study investigators could have been biased to look for infection more often in the early initiation group.

Bottom line: In ICU patients at risk for malnutrition, delaying initiation of TPN to supplement enteral feeding shortens ICU stay and reduces infectious complications when compared with early initiation of TPN.

Citation: Casaer MP, Mesotten D, Hermans G, et al. Early versus late parenteral nutrition in critically ill adults. N Engl J Med. 2011;365:506-17.

Intrapleural Instillation of Combined DNase and t-PA Improve Outcomes in Patients with Empyema

Clinical question: Does the intrapleural administration of a fibrinolytic, a recombinant DNase, or a combination of the two improve outcomes in patients with pleural infections compared with placebo?

Background: Pleural-based infections confer high morbidity and mortality, especially in the one-third of cases in which chest tube drainage fails. Observational data suggest intrapleural administration of fibrinolytics can improve drainage; however, a large randomized trial (MIST1) failed to show improvement with streptokinase.

Study design: Randomized, double-blind, double-dummy, factorial, multicenter trial.

Setting: University hospitals in the U.K.

Synopsis: A total of 193 patients with clinical evidence of an infection as well as laboratory evidence of pleural infection were randomized in factorial design to placebo, DNase alone, t-PA alone, and DNase plus t-PA as a twice-daily, hourlong instillation into the pleural space on hospital days 1 through 3. The combination group had improvements in the size of the pleural effusion compared with placebo, the primary endpoint. This group also was less likely to be referred for surgery, had shorter hospital stays, had less fever, and had lower inflammatory markers by hospital Day 7. Neither of the single-agent groups were better than placebo. The confidence intervals for all outcomes besides radiographic size of the effusion were large, as the trial had limited power for secondary endpoints.

Bottom line: In patients with empyema, the addition of twice-daily instillations of DNase and tPA can improve drainage and decrease risk of treatment failure when combined with chest tube drainage and antibiotics.

Citation: Rahman NM, Maskell NA, West A, et al. Intrapleural use of tissue plasminogen activator and DNase in pleural infection. N Engl J Med. 2011;365:518-526.

Smoothing Admissions over the Week Improves Patient Flow

Clinical question: What is the difference between weekday and weekend hospital occupancy, and what might the effect be of redistributing, or “smoothing,” elective admissions across the week?

Background: Efforts to improve patient flow have largely focused on reducing the average length of stay (ALOS). As the ALOS decreases, though, further reductions have limited yield in improving patient flow and could occur at the expense of patient safety. Smoothing admissions is a recognized but underutilized tool to address patient flow and hospital overcrowding.

Study design: Retrospective cohort.

Setting: Thirty-nine freestanding, tertiary-care children’s hospitals in the U.S.

Synopsis: Hospital occupancy ranged from 70.9% to 108.1% for weekdays, and from 65.7% to 94.4% on weekends. The mean difference between weekday and weekend occupancy was 8.2%. Using a mathematical model to redistribute admissions from peak days to nonpeak days (within a one-week time frame), the investigators found that percent occupancy on average dropped by 6.6%—the same number of patients over the same time interval, but with lower average occupancy. And, while not projected by these authors, the correlate of lower average occupancy would be reduced waits and delays for admission and within-hospital services. Only 2.6%, or about 7.5 admissions per week per hospital, would need to be rescheduled in order to realize this gain.

Bottom line: Where feasible, reshuffling elective admissions to “smooth” demand across the week is associated with improved hospital flow.

Citation: Fieldston ES, Hall M, Shah SS, et al. Addressing inpatient crowding by smoothing occupancy at children’s hospitals. J Hosp Med. 2011;6:466-473.

Variability in Preferences, Experiences, and Expectations for Hospitalist Roles in Comanagement

Clinical question: Do provider expectations and experiences in comanagement differ from the traditional medical consultation model?

Background: Hospitalists participate in traditional medical consultation and, increasingly, a variety of comanagement with surgical and various medical specialists. It is uncertain what preferences and expectations on either side of the comanagement relationship might be. Learning more might lead to a better conceptual understanding and working definition of inpatient comanagement.

Study design: Baseline and follow-up surveys.

Setting: Large single-site academic medical center, hospitalist-hepatologist comanagement service.

Synopsis: All 41 respondents completed the baseline survey, with 32 respondents completing a total of 177 follow-up surveys. Respondents included hospitalists, hepatologists, nonphysician providers (NPPs), and GI fellows. Hospitalists were more likely than hepatologists to want to be informed about every patient care decision (83% versus 17%). The majority of respondents reported understanding their roles. At baseline, nearly all agreed that “comanagement tends to improve patient care.” However, at follow-up, fewer NPP and fewer GI fellows reported that patients on their service had received “better care than they would without comanagement.” At follow-up, fewer hospitalists than at baseline reported that there was one “physician leader to direct the overall management of the patient’s hospital course.”

Bottom line: Hospitalists were more likely than hepatologists, within a comanagement relationship, to want to be informed about every decision. Most expected—but fewer actually later reported—that the comanagement model improved patient care.

Citation: Hinami K, Whelan CT, Konetzka RT, and Meltzer DO. Provider expectations and experiences of comanagement. J Hosp Med. 2011;6:401-404.

In This Edition

Literature At A Glance

A guide to this month’s studies

1. Higher loading dose of clopidogrel in STEMI
2. Early vs. late surgery following hip fracture
3. Beta-blockers in chronic kidney disease
4. Long-term azithromycin in COPD
5. CT screening for lung cancer
6. Timing of parenteral nutrition in the ICU
7. Intrapleural management of empyema with DNase and t-PA
8. Effect of weekend elective admissions on hospital flow
9. Expectations and outcomes of medical comanagement

Higher-Dose Clopidogrel Improves Outcomes at 30 Days in STEMI Patients

Clinical question: Does a 600-mg loading dose of clopidogrel given before percutaneous coronary intervention (PCI) in patients with an ST-segment elevation myocardial infarction (STEMI) provide more protection from thrombotic complications at 30 days than a 300-mg dose?

Background: Clopidogrel at 600 mg is active more quickly (two hours versus 12 hours) and inhibits platelets more completely than does a 300-mg dose, but it has never been tested prospectively in patients undergoing percutaneous intervention with a STEMI.

Study design: Prospective, multicenter, randomized controlled trial.

Setting: Five hospitals in Italy, Belgium, Serbia, and Hungary.

Synopsis: Two-hundred-one patients with STEMI were randomized to either 600 mg or 300 mg of clopidogrel, given an average of 30 minutes before initial PCI, as well as other standard treatment for STEMI. The primary outcome was “infarct size,” judged as the area under the curve (AUC) of serial creatine kinase-MB (CK-MB) and troponin measurements. At 30 days, patients who received the 600-mg dose of clopidogrel had lower AUCs of cardiac biomarkers, statistically significant (though clinically small) improvement in left ventricle ejection fraction at discharge, lower incidence of severely impaired post-PCI thrombolitis in myocardial infarction (TIMI) flow, and fewer “major cardiovascular events,” driven mainly by a reduction in revascularizations. Measurement of biomarkers to calculate infarct size can be confounded by hypertrophy, and the trial was underpowered for cardiovascular events. However, there was no increase in bleeding events.

Bottom line: In patients with STEMI, treatment with a higher loading dose of clopidogrel before PCI reduces revascularizations and might decrease infarct size without an increase in adverse events.

Citation: Patti G, Bárczi G, Orlic D, et al. Outcome comparison of 600- and 300-mg loading doses of clopidogrel in patients undergoing primary percutaneous coronary intervention for ST-segment elevation myocardial infarction: results from the ARMYDA-6 MI (Antiplatelet therapy for Reduction of MYocardial Damage during Angioplasty-Myocardial Infarction) randomized study. J Am Coll Cardiol. 2011;58:1592-1599.

Medical Comorbidities Explain Most of Excess Risk Seen in Patients with Delayed Hip Fracture Repair

Clinical question: Could the increased mortality found with delays in hip fracture surgery be confounded by the premorbid functional status and medical comorbidities of patients whose surgery is more likely to be delayed?

Background: Guidelines recommend operating on patients with hip fracture within 24 hours, but the supporting evidence has not adjusted for underlying medical comorbidities, which could delay surgery and contribute to poor outcomes, making delays look harmful.

Study design: Prospective cohort, single-center design.

Setting: University hospital in Spain.

Synopsis: The study included 2,250 consecutive elderly patients admitted to the hospital for hip fracture who had their functional status and medical comorbidities assessed at enrollment. If their surgery was delayed beyond 24 hours, the reason was sought. Medical outcomes assessed daily while in hospital were delirium, pneumonia, heart failure, urinary tract infection, and new pressure sores, while the dose of pain medication, surgical complications, and in-hospital mortality were also compiled. No post-discharge data were available.

The median time to surgery was 72 hours. Patients with more medical comorbidities and poorer preoperative functional status had longer delays to operation, most commonly due to interrupting antiplatelet treatment or need for preoperative “echocardiography or other examinations.” When these medical factors were included in logistic regression analysis, the increased mortality seen with delays of surgery in the cohort was no longer statistically significant, suggesting the underlying comorbidities of these patients, rather than the delay to surgery itself, explained the increased mortality.

Bottom line: Delaying hip fracture surgery is less important than preoperative functional status and medical comorbidity in contributing to poor outcomes.

Citation: Vidán MT, Sánchez E, Gracia Y, Marañón E, Vaquero J, Serra JA. Causes and effects of surgical delay in patients with hip fracture: a cohort study. Ann Intern Med. 2011;155:226-233.

Beta-Blockers Decrease All-Cause and Cardiovascular Mortality in Patients with Chronic Kidney Disease

Clinical question: Are beta-blockers as effective in patients on dialysis and with end-stage chronic kidney disease (CKD) as they are in patients without severe renal disease?

Background: Patients with CKD have been largely excluded from trials of beta-blocker therapy even though they have high rates of cardiovascular disease and might be most likely to benefit. However, patients on dialysis might be predisposed to adverse complications of beta-blocker therapy, including hypotension.

Study design: Meta-analysis of eight trials of beta-blockade versus placebo (six heart failure trials) or ACE-I (two non-heart-failure trials) that included post-hoc analyses of CKD patients.

Setting: Varied, usually multinational RCTs.

Synopsis: The six congestive heart failure (CHF) studies were not designed to evaluate patients with CKD, and the two non-CHF studies were intended to evaluate progression of CKD, not cardiac outcomes. Thus, this is a meta-analysis of post-hoc CKD subgroups included in these trials. Compared with placebo, beta-blockers reduced all-cause and cardiovascular mortality without significant heterogeneity between trials. The magnitude of the effect was similar in CKD and non-CKD patients. Patients with CKD treated with beta-blockers were at increased risk of bradycardia and hypotension, but this did not lead to increased discontinuation of the drug. Only 114 dialysis patients were included in one of the eight trials (7,000 patients overall) and no outcomes were assessed.

Bottom line: Beta-blockers lower all-cause and cardiovascular mortality in patients with CKD similarly to patients without kidney disease but are associated with an increased risk for hypotension and bradycardia. Their effect in dialysis patients is unknown.

Citation: Badve SV, Roberts MA, Hawley CM, et al. Effects of beta-adrenergic antagonists in patients with chronic kidney disease: a systematic review and meta-analysis. J Am Coll Cardiol. 2011;58:1152-1161.

Daily Long-Term Azithromycin in COPD Patients Decreases Frequency of Exacerbations

Clinical question: Does long-term treatment with azithromycin decrease COPD exacerbations and improve quality of life with an acceptable risk profile?

Background: Patients with acute exacerbations of COPD have increased risks of death and more rapid decline in lung function. Macrolide antibiotics might decrease exacerbations via antibacterial and anti-inflammatory effects. Small studies of long-term treatment with macrolides in COPD have had conflicting results.

Study design: Prospective, multicenter, parallel-group, placebo-controlled trial.

Setting: Twelve U.S. academic health centers.

Synopsis: Eligible patients were those older than 40 with pulmonary function test-proven obstructive disease, a significant smoking history, a use of supplemental oxygen, or an oral glucocorticoid treatment in the previous year, and those who had previously come to healthcare attention for a COPD exacerbation. They were randomized to daily azithromycin (250 mg) or placebo. The primary outcome was time to first COPD exacerbation at one year. The investigators used deep nasopharyngeal swabs to evaluate the development of microbial resistance, and all patients were regularly screened for hearing loss. Time to first exacerbation and frequency of exacerbations decreased significantly with azithromycin treatment; the number needed to treat to prevent a COPD exacerbation was 2.86. More than half of nasopharyngeal samples had evidence of macrolide resistance at enrollment, and this increased to more than 80% at one year. Audiometric testing did show a decrease in hearing in those treated with azithromycin, but even those who did not stop their medication had recovery on subsequent testing.

Bottom line: Daily azithromycin decreases COPD exacerbations in those with a history of previous exacerbation at the cost of increased macrolide resistance and possible reversible decrements in hearing. The high degree of macrolide resistance at enrollment suggests non-antibacterial mechanisms might be responsible.

Citation: Albert RK, Connett J, Bailey WC, et al. Azithromycin for prevention of exacerbations of COPD. N Engl J Med. 2011;365:689-698.

Serial Low-Dose CT Scans Decrease Lung Cancer Mortality

Clinical question: Does annual screening with low-dose CT scans reduce mortality from lung cancer in current or former heavy smokers?

Background: Lung cancer is still the top cause of cancer death in the United States, and its prevalence is increasing in the developing world. Prior trials of screening chest radiography and sputum samples have not shown a decrease in lung cancer mortality, but CT scans could help identify lung cancer at an earlier, more treatable stage.

Study design: Prospective, multicenter, randomized trial of low-dose (roughly one-fifth the radiation of traditional CT) annual CT versus annual chest radiography.

Setting: Thirty-three U.S. medical institutions.

Synopsis: The study enrolled 53,454 current or prior smokers with a history of smoking more than 30 packs a year who were randomly assigned to undergo three annual CT scans or chest X-rays. Any CT scans that showed a noncalcified nodule greater than 4 mm and chest X-rays that showed a noncalcified nodule or mass of any size were classified as “positive,” but follow-up was left to the discretion of the treating physicians. Participants were followed for 6.5 years on average.

Of those screened with CT, 40% had a positive screening test at some point, more than three times the rate of chest radiography, and 96.4% of these were false-positives, which was similar to chest radiography. Adverse events as a result of workup of eventual false positives were uncommon, occurring in around 1% of those who did not have lung cancer and in 11% of those who did. Screening reduced lung cancer mortality by 20% compared with chest radiography, with a number needed to screen to prevent one lung cancer death of 320. Cost-effectiveness and effects of radiation were not assessed.

Bottom line: Serial screening with low-dose CT reduces lung cancer mortality at the cost of a high rate of false-positives. Questions remain about cost-effectiveness and radiation exposure.

Citation: National Lung Screening Trial Research Team, Aberle DR, Adams AM, Berg CD, et al. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med. 2011;365:395-409.

One-Week Delay in Starting Parenteral Nutrition in ICU Patients Is Associated with Better Outcomes

Clinical question: In patients admitted to the ICU who are at risk for malnutrition, does supplementing enteral feeding with parenteral feeding on the day of admission improve outcomes when compared with supplementation starting on Day 8?

Background: Patients with critical illness are at risk for malnutrition, which may lead to worsened outcomes. Many are unable to tolerate enteral feeding. However, adding parenteral nutrition has risks, including overfeeding and hyperglycemia.

Study design: Prospective, multicenter, randomized controlled trial. Neither patients/families nor ICU physicians were blinded.

Setting: Two university hospitals in Belgium.

Synopsis: The study included 4,640 patients admitted to the ICU who were “at risk” for malnutrition (assessed using a Nutritional Risk Score of >3). They were randomized to early initiation of parenteral nutrition on Day 1 or late initiation of parenteral nutrition on Day 8, in both cases only if needed to supplement enteral feeding. Both groups received intensive IV insulin protocols to keep blood glucose at 80 mg/dL to 110 mg/dL. The study populations were well-matched, with similar APACHE II scores. Of the total patients, 42% were emergency admissions to the ICU (more than half of which had sepsis), while scheduled cardiac surgery patients made up the majority of the rest of the study participants.

Patients with late initiation of total parenteral nutrition (TPN) were more likely to be discharged alive from the ICU within eight days despite having increased risk of hypoglycemia and more elevation of inflammatory markers. They had shorter ICU and hospital stays, fewer days on the ventilator, fewer infectious complications, were less likely to develop renal failure, and had lower overall costs. The unblinded management of these patients raises concern in interpreting infectious outcomes because study investigators could have been biased to look for infection more often in the early initiation group.

Bottom line: In ICU patients at risk for malnutrition, delaying initiation of TPN to supplement enteral feeding shortens ICU stay and reduces infectious complications when compared with early initiation of TPN.

Citation: Casaer MP, Mesotten D, Hermans G, et al. Early versus late parenteral nutrition in critically ill adults. N Engl J Med. 2011;365:506-17.

Intrapleural Instillation of Combined DNase and t-PA Improve Outcomes in Patients with Empyema

Clinical question: Does the intrapleural administration of a fibrinolytic, a recombinant DNase, or a combination of the two improve outcomes in patients with pleural infections compared with placebo?

Background: Pleural-based infections confer high morbidity and mortality, especially in the one-third of cases in which chest tube drainage fails. Observational data suggest intrapleural administration of fibrinolytics can improve drainage; however, a large randomized trial (MIST1) failed to show improvement with streptokinase.

Study design: Randomized, double-blind, double-dummy, factorial, multicenter trial.

Setting: University hospitals in the U.K.

Synopsis: A total of 193 patients with clinical evidence of an infection as well as laboratory evidence of pleural infection were randomized in factorial design to placebo, DNase alone, t-PA alone, and DNase plus t-PA as a twice-daily, hourlong instillation into the pleural space on hospital days 1 through 3. The combination group had improvements in the size of the pleural effusion compared with placebo, the primary endpoint. This group also was less likely to be referred for surgery, had shorter hospital stays, had less fever, and had lower inflammatory markers by hospital Day 7. Neither of the single-agent groups were better than placebo. The confidence intervals for all outcomes besides radiographic size of the effusion were large, as the trial had limited power for secondary endpoints.

Bottom line: In patients with empyema, the addition of twice-daily instillations of DNase and tPA can improve drainage and decrease risk of treatment failure when combined with chest tube drainage and antibiotics.

Citation: Rahman NM, Maskell NA, West A, et al. Intrapleural use of tissue plasminogen activator and DNase in pleural infection. N Engl J Med. 2011;365:518-526.

Smoothing Admissions over the Week Improves Patient Flow

Clinical question: What is the difference between weekday and weekend hospital occupancy, and what might the effect be of redistributing, or “smoothing,” elective admissions across the week?

Background: Efforts to improve patient flow have largely focused on reducing the average length of stay (ALOS). As the ALOS decreases, though, further reductions have limited yield in improving patient flow and could occur at the expense of patient safety. Smoothing admissions is a recognized but underutilized tool to address patient flow and hospital overcrowding.

Study design: Retrospective cohort.

Setting: Thirty-nine freestanding, tertiary-care children’s hospitals in the U.S.

Synopsis: Hospital occupancy ranged from 70.9% to 108.1% for weekdays, and from 65.7% to 94.4% on weekends. The mean difference between weekday and weekend occupancy was 8.2%. Using a mathematical model to redistribute admissions from peak days to nonpeak days (within a one-week time frame), the investigators found that percent occupancy on average dropped by 6.6%—the same number of patients over the same time interval, but with lower average occupancy. And, while not projected by these authors, the correlate of lower average occupancy would be reduced waits and delays for admission and within-hospital services. Only 2.6%, or about 7.5 admissions per week per hospital, would need to be rescheduled in order to realize this gain.

Bottom line: Where feasible, reshuffling elective admissions to “smooth” demand across the week is associated with improved hospital flow.

Citation: Fieldston ES, Hall M, Shah SS, et al. Addressing inpatient crowding by smoothing occupancy at children’s hospitals. J Hosp Med. 2011;6:466-473.

Variability in Preferences, Experiences, and Expectations for Hospitalist Roles in Comanagement

Clinical question: Do provider expectations and experiences in comanagement differ from the traditional medical consultation model?

Background: Hospitalists participate in traditional medical consultation and, increasingly, a variety of comanagement with surgical and various medical specialists. It is uncertain what preferences and expectations on either side of the comanagement relationship might be. Learning more might lead to a better conceptual understanding and working definition of inpatient comanagement.

Study design: Baseline and follow-up surveys.

Setting: Large single-site academic medical center, hospitalist-hepatologist comanagement service.

Synopsis: All 41 respondents completed the baseline survey, with 32 respondents completing a total of 177 follow-up surveys. Respondents included hospitalists, hepatologists, nonphysician providers (NPPs), and GI fellows. Hospitalists were more likely than hepatologists to want to be informed about every patient care decision (83% versus 17%). The majority of respondents reported understanding their roles. At baseline, nearly all agreed that “comanagement tends to improve patient care.” However, at follow-up, fewer NPP and fewer GI fellows reported that patients on their service had received “better care than they would without comanagement.” At follow-up, fewer hospitalists than at baseline reported that there was one “physician leader to direct the overall management of the patient’s hospital course.”

Bottom line: Hospitalists were more likely than hepatologists, within a comanagement relationship, to want to be informed about every decision. Most expected—but fewer actually later reported—that the comanagement model improved patient care.

Citation: Hinami K, Whelan CT, Konetzka RT, and Meltzer DO. Provider expectations and experiences of comanagement. J Hosp Med. 2011;6:401-404.