Steven Deitelzweig, MD, MMM, SFHM

Clinical question: Is routine preoperative urine screening beneficial?

Background: The value of preoperative urine screening is unproven, except before urologic procedures. Furthermore, treatment of asymptomatic bacteriuria may lead to adverse events, including diarrhea, allergic reactions, and Clostridium difficile infection (CDI).

Study design: Retrospective chart review.

Setting: Patients who underwent cardiothoracic, orthopedic, and vascular surgeries at the Minneapolis Veterans Affairs Medical Center in 2010.

Synopsis: A total of 1,934 procedures were performed on 1,699 patients, most of which were orthopedics procedures (1,291 in 1,115 patients). A urine culture was obtained before 25% of procedures with significant variation by service (cardiothoracic, 85%; vascular, 48%; orthopedic, 4%). Bacteriuria was detected in 11% of urine cultures (54 of 489), but antimicrobial drugs were dispensed to just 16 patients.

To identify correlates of preoperative urine culture use, patients with and without urine cultures were compared. The rate of surgical-site infection was similar for both groups. Postoperative UTI was more frequent among patients with bacteriuria. Rates of diarrhea, allergy, and CDI did not differ. Paradoxically, patients treated for preoperative UTI were more likely to develop surgical-site infections (45% vs. 14%; P=0.03). Postoperative UTI was also more frequent among treated patients versus untreated patients (18% vs. 7%).

Bottom line: This is the largest study to assess outcomes for routine preoperative urine cultures. These findings demonstrate that preoperative screening for, and treatment of, asymptomatic bacteriuria should be avoided in patients undergoing nonurologic surgical procedures.

Citation: Drekonja DM, Zarmbinski B, Johnson JR. Preoperative urine culture at a veterans affairs medical center. JAMA Intern Med. 2013;173(1):71-72.

Visit our website for more physician reviews of recent HM-relevant literature.

Clinical question: Is routine preoperative urine screening beneficial?

Background: The value of preoperative urine screening is unproven, except before urologic procedures. Furthermore, treatment of asymptomatic bacteriuria may lead to adverse events, including diarrhea, allergic reactions, and Clostridium difficile infection (CDI).

Study design: Retrospective chart review.

Setting: Patients who underwent cardiothoracic, orthopedic, and vascular surgeries at the Minneapolis Veterans Affairs Medical Center in 2010.

Synopsis: A total of 1,934 procedures were performed on 1,699 patients, most of which were orthopedics procedures (1,291 in 1,115 patients). A urine culture was obtained before 25% of procedures with significant variation by service (cardiothoracic, 85%; vascular, 48%; orthopedic, 4%). Bacteriuria was detected in 11% of urine cultures (54 of 489), but antimicrobial drugs were dispensed to just 16 patients.

To identify correlates of preoperative urine culture use, patients with and without urine cultures were compared. The rate of surgical-site infection was similar for both groups. Postoperative UTI was more frequent among patients with bacteriuria. Rates of diarrhea, allergy, and CDI did not differ. Paradoxically, patients treated for preoperative UTI were more likely to develop surgical-site infections (45% vs. 14%; P=0.03). Postoperative UTI was also more frequent among treated patients versus untreated patients (18% vs. 7%).

Bottom line: This is the largest study to assess outcomes for routine preoperative urine cultures. These findings demonstrate that preoperative screening for, and treatment of, asymptomatic bacteriuria should be avoided in patients undergoing nonurologic surgical procedures.

Citation: Drekonja DM, Zarmbinski B, Johnson JR. Preoperative urine culture at a veterans affairs medical center. JAMA Intern Med. 2013;173(1):71-72.

Visit our website for more physician reviews of recent HM-relevant literature.

Clinical question: Is routine preoperative urine screening beneficial?

Background: The value of preoperative urine screening is unproven, except before urologic procedures. Furthermore, treatment of asymptomatic bacteriuria may lead to adverse events, including diarrhea, allergic reactions, and Clostridium difficile infection (CDI).

Study design: Retrospective chart review.

Setting: Patients who underwent cardiothoracic, orthopedic, and vascular surgeries at the Minneapolis Veterans Affairs Medical Center in 2010.

Synopsis: A total of 1,934 procedures were performed on 1,699 patients, most of which were orthopedics procedures (1,291 in 1,115 patients). A urine culture was obtained before 25% of procedures with significant variation by service (cardiothoracic, 85%; vascular, 48%; orthopedic, 4%). Bacteriuria was detected in 11% of urine cultures (54 of 489), but antimicrobial drugs were dispensed to just 16 patients.

To identify correlates of preoperative urine culture use, patients with and without urine cultures were compared. The rate of surgical-site infection was similar for both groups. Postoperative UTI was more frequent among patients with bacteriuria. Rates of diarrhea, allergy, and CDI did not differ. Paradoxically, patients treated for preoperative UTI were more likely to develop surgical-site infections (45% vs. 14%; P=0.03). Postoperative UTI was also more frequent among treated patients versus untreated patients (18% vs. 7%).

Bottom line: This is the largest study to assess outcomes for routine preoperative urine cultures. These findings demonstrate that preoperative screening for, and treatment of, asymptomatic bacteriuria should be avoided in patients undergoing nonurologic surgical procedures.

Citation: Drekonja DM, Zarmbinski B, Johnson JR. Preoperative urine culture at a veterans affairs medical center. JAMA Intern Med. 2013;173(1):71-72.

Visit our website for more physician reviews of recent HM-relevant literature.

In This Edition

Literature At A Glance

A guide to this month’s studies

1. Value of routine preoperative urine screening assessed
2. Impact of hospitalist-led intermediate care on patient survival
3. Risks of blood transfusion to treat upper GI bleeding
4. Low-dose steroids and increased mortality in sepsis
5. Reduced length of stay and hospital readmission rates
6. Restrictive blood transfusion strategies better for acute myocardial infarction
7. Trends in GI illnesses and their associated costs
8. Apixaban as a stand-alone anticoagulant in patients with VTE
9. Guidelines for upper endoscopy use in gastroesophageal reflux disease

Avoid Preoperative Urine Culture in Nonurologic Surgical Procedures

Clinical question: Is routine preoperative urine screening beneficial?

Background: The value of preoperative urine screening is unproven, except before urologic procedures. Furthermore, treatment of asymptomatic bacteriuria may lead to adverse events, including diarrhea, allergic reactions, and Clostridium difficile infection (CDI).

Study design: Retrospective chart review.

Setting: Patients who underwent cardiothoracic, orthopedic, and vascular surgeries at the Minneapolis Veterans Affairs Medical Center in 2010.

Synopsis: A total of 1,934 procedures were performed on 1,699 patients, most of which were orthopedics procedures (1,291 in 1,115 patients). A urine culture was obtained before 25% of procedures with significant variation by service (cardiothoracic, 85%; vascular, 48%; orthopedic, 4%). Bacteriuria was detected in 11% of urine cultures (54 of 489), but antimicrobial drugs were dispensed to just 16 patients.

To identify correlates of preoperative urine culture use, patients with and without urine cultures were compared. The rate of surgical-site infection was similar for both groups. Postoperative UTI was more frequent among patients with bacteriuria. Rates of diarrhea, allergy, and CDI did not differ. Paradoxically, patients treated for preoperative UTI were more likely to develop surgical-site infections (45% vs. 14%; P=0.03). Postoperative UTI was also more frequent among treated patients versus untreated patients (18% vs. 7%).

Bottom line: This is the largest study to assess outcomes for routine preoperative urine cultures. These findings demonstrate that preoperative screening for, and treatment of, asymptomatic bacteriuria should be avoided in patients undergoing nonurologic surgical procedures.

Citation: Drekonja DM, Zarmbinski B, Johnson JR. Preoperative urine culture at a veterans affairs medical center. JAMA Intern Med. 2013;173(1):71-72.

Intermediate Care Staffed by Hospitalists: Impact on Mortality, Comanagement, and Teaching

Clinical question: Does a hospitalist-led intermediate-care unit improve patient survival?

Background: Hospitalized patients are complex, and institutions often have to balance matching patient acuity to either an ICU or a regular ward. However, an intermediate-care setting might be an attractive strategy to provide rational care according to patient needs while expanding comanagement and teaching services.

Study design: Retrospective observational study.

Setting: Intermediate-care unit of a single academic hospital.

Synopsis: In-hospital mortality in this intermediate-care unit was 20.6%, whereas the expected mortality was 23.2% based on Simplified Acute Physiology Score II (SAPS II) score. The correlation between SAPS II predicted and observed death rates was accurate and statistically significant (P<0.001). Comanagement was performed with several medical and surgical teams, with an increase in perioperative comanagement of 22.7% (P=0.014). The number of training residents in the intermediate-care unit increased to 30.4% from 4.3% (P=0.002).

Bottom line: An intermediate-care unit led by hospitalists showed encouraging results in patient mortality, as well as comanagement and teaching opportunities.

Citation: Lucena JF, Alegre F, Rodil R, et al. Results of a retrospective observational study of intermediate care staffed by hospitalists: impact on mortality, co-management, and teaching. J Hosp Med. 2012;7(5):411-415.

Blood Transfusion Associated with Increased Risk of Rebleeding in Patients with Nonvariceal Upper GI Bleeding

Clinical question: Does more liberal use of blood transfusions in the setting of nonvariceal upper GI bleeding result in patient harm?

Background: Randomized controlled trials have demonstrated that a more liberal approach to blood transfusions for patients in the medical intensive-care unit results in higher mortality. However, the potential harmful effect of blood transfusions in the setting of GI bleeding has not been demonstrated.

Study design: Retrospective cohort study.

Setting: Canadian hospitals.

Synopsis: Based on a retrospective analysis of the Canadian Registry of patients with Upper Gastrointestinal Bleeding and Endoscopy (RUGBE), the authors determined there was a statistically significant association between patients who received a blood transfusion for the management of nonvariceal upper GI bleeding and the risk of rebleeding. The rate of rebleeding in patients who received a blood transfusion was 23.6% compared with 11.3% in patients who were not transfused (P<0.01). There was no statistically significant difference in mortality.

Although this was a reasonably large observational study that included 1,677 patients with nonvariceal upper GI bleeding, it is vulnerable to confounding. It suggests the need to further study potential harm of blood transfusion in the setting of GI bleeding, but it should not result in a change in clinical practice at this time.

Bottom line: Prospective randomized studies are needed to determine if there are harmful effects of blood transfusions in the setting of GI bleeding and to better define a threshold for transfusion.

Citation: Restellini S, Kherad O, Jairath V, Martel M, Barkun AN. Red blood cell transfusion is associated with increased rebleeding in patients with nonvariceal upper gastrointestinal bleeding. Aliment Pharmacol Ther. 2013;37:316-322.

Low-Dose Steroids in Sepsis Associated with Increase in Mortality

Clinical question: What is the role of steroids in the treatment of adult patients with sepsis?

Background: The Surviving Sepsis Campaign guidelines have previously recommended administering steroids to patients with septic shock not responsive to fluid resuscitation and who require vasopressors. However, prior randomized clinical trials studying the use of steroids in these settings have produced conflicting results.

Study design: Retrospective cohort study.

Setting: Two hundred fifty-two hospitals in North America, South America, and Europe.

Synopsis: The Surviving Sepsis Campaign management bundle has been shown to reduce mortality in patients with sepsis. However, it is not known which particular elements of the management bundle result in improved mortality. The Surviving Sepsis Campaign database included 17,847 patients who required vasopressor therapy after adequate fluid resuscitation. This subgroup was analyzed to see if there was a difference in mortality between patients who received low-dose steroids versus those who did not receive steroids. The mortality rate among those who received steroids was statistically higher (with odds ratio of 1.18 and P<0.001) compared with those who did not receive steroids. This finding adds to the body of evidence that calls into question the commonplace practice of administrating steroids to septic patients on vasopressor therapy.

The most recent campaign guidelines recommend the use of steroids in septic patients only if both adequate fluid resuscitation and vasopressor therapy are not able to restore hemodynamic stability.

Bottom line: Further studies are needed to better define the role of steroids in the treatment of sepsis.

Citation: Casserly B, Gerlach H, Phillips GS, et al. Low-dose steroids in adult septic shock: results of the Surviving Sepsis Campaign. Intensive Care Med. 2012;38:1946-1954.

Link Between Length of Stay and Readmission Rates

Clinical question: How has reducing length of stay affected hospital readmission rates?

Background: There are ongoing concerns that improving a hospital’s efficiency by reducing length of stay (LOS) could be associated with higher hospital readmission rates. However, no studies evaluating the relationship between LOS and readmission rates have been done using recent data.

Study design: Retrospective observational study.

Setting: All acute-care Veterans Affairs (VA) hospitals in the U.S.

Synopsis: A total of 4,124,907 index admissions were included in the final sample from all acute medical admissions in 129 acute-care VA hospitals from October 1996 to September 2010. The primary outcomes were the hospital LOS and the 30-day readmission rate. Index admissions for heart failure, chronic obstructive pulmonary disease (COPD), acute myocardial infarction (AMI), community-acquired pneumonia, and gastrointestinal hemorrhage were also analyzed separately.

The risk-adjusted analysis of LOS demonstrated significant reductions for all admissions over the 14-year period, to 3.98 days from 5.44 days, and for all of the individual conditions with reductions ranging from 1.40 days for gastrointestinal hemorrhage to 2.85 days for AMI. There were similar significant reductions in 30-day readmission rates for all admissions to 13.8% from 16.5% and within the individual conditions ranging from 0.9% in community-acquired pneumonia to 3.3% in COPD. These results show that the reductions in LOS did not increase the risk of readmissions. The major limitation of the study was that these data are only from a single healthcare system.

Bottom line: Data from VA hospitals show that reductions in LOS do not have adverse effects on 30-day readmission rates; instead, both LOS and readmission rates improved over the same time period.

Citation: Kaboli PJ, Go JT, Hockenberry J, et al. Associations between reduced hospital length of stay and 30-day readmission rate and mortality: 14-year experience in 129 Veterans Affairs hospitals. Ann Intern Med. 2012;157:837-845.

Better to Restrict Blood Transfusions in Acute Myocardial Infarction

Clinical question: Is a liberal or restrictive blood transfusion strategy better in patients with anemia and acute myocardial infarction?

Background: Patients with acute myocardial infarction (AMI) are often given therapies that can increase their risk for bleeding and anemia, and it is known that AMI patients have a worse prognosis if they have concomitant anemia. No clear consensus exists on the benefit or harm of blood transfusions in AMI patients.

Study design: Systematic review and meta-analysis.

Setting: Ten articles included in the qualitative and quantitative analyses out of 729 screened articles from Jan. 1, 1966, to March 31, 2012, using the search terms “transfusion,” “myocardial infarction,” and “mortality” in English language.

Synopsis: A total of 203,665 study participants were identified from the 10 studies (one randomized and nine observational) that met the inclusion and exclusion criteria. All-cause mortality was significantly higher in AMI patients who received a blood transfusion compared with those who did not (18.2% vs. 10.2%). However, this difference was not statistically significant in patients that had a STEMI or in patients with a baseline hematocrit less than 30%. A multivariate meta-regression with several covariates, excluding demographics, also showed that blood transfusion was associated with higher mortality and higher risk for subsequent myocardial infarction. There was significant heterogeneity in all results, but no single study was found as the source of the heterogeneity, and no significant publication bias was identified. The major limitations to this study are that there is a paucity of randomized trials available that pertain to this specific topic and the authors did not have patient-level covariates to include in their analyses.

Bottom line: There appears to be an increased risk of mortality and subsequent myocardial infarction in AMI patients who receive blood transfusions versus those who do not.

Citation: Chatterjee S, Wetterslev J, Sharma A, Lichstein E, Mukherjee D. Association of blood transfusion with increased mortality in myocardial infarction: a meta-analysis and diversity-adjusted study sequential analysis. JAMA Intern Med. 2013;173(2):132-139.

Trends in GI Illnesses and Their Associated Costs

Clinical question: What are the new trends in GI illnesses and their associated costs?

Background: The frequency of illnesses and their treatment costs have changed over the last decade. In order to help healthcare providers focus their attention on these new trends, a new compilation of data is needed.

Study design: Epidemiological analysis.

Setting: Various governmental and private databases representing outpatient clinics, hospitals, and death certificates from multiple regions of the U.S.

Synopsis: The analysis was blinded to patient identifiers but represented multiple regions of the U.S. Symptoms were abstracted from patient surveys, and the rest of the data were collected from record review. The most common reported symptoms were abdominal pain, followed by nausea, vomiting, diarrhea, constipation, and heartburn. The most common clinic diagnoses were reflux, abdominal pain, enteritis/dyspepsia, and constipation. The most common inpatient discharge primary diagnoses included acute pancreatitis, cholecystitis, and diverticulitis. Impressive increases were seen in the number of morbidly obese, C. diff, and fatty liver diagnoses.

Colon cancer was the most common GI malignancy and had the highest mortality. C. diff was the ninth-leading cause of GI-related deaths. All types of scopes (except endoscopic retrograde cholangiopancreatography) were performed more commonly now than in the past, with colonoscopy being the most common. The most common indication for an upper endoscopy was reflux, which was also the most common outpatient GI diagnosis.

Bottom line: Healthcare providers need to be aware of new GI illness trends and their associated costs.

Citation: Peery AF, Dellon ES, Lund J, et al. Burden of gastrointestinal disease in the United States: 2012 update. Gastroenterol. 2012;143:1179-1187.

Is Apixaban a Good Stand-Alone Anticoagulant for Extended Treatment in VTE Patients?

Clinical question: Is apixaban an option for the extended treatment of VTE in a simple, fixed-dose regimen?

Background: Apixaban is an oral factor Xa inhibitor that is administered in fixed doses without the need for laboratory monitoring. In the Apixaban after the Initial Management of Pulmonary Embolism and Deep Vein Thrombosis with First-Line Therapy—Extended Treatment (AMPLIFY-EXT) study, investigators compared the efficacy and safety of two doses of apixaban (2.5 mg and 5 mg) with those of placebo in patients with VTE who had completed six to 12 months of anticoagulation therapy and for whom treating physicians were uncertain about continuing therapy. Additional aims of the study were to determine whether the lower dose of apixaban was effective and whether it was associated with less bleeding than the higher dose, and to examine the effect of treatment on arterial thrombotic outcomes.

Study design: Randomized, double-blind study.

Synopsis: A total of 2,486 patients underwent randomization, 2,482 of whom were included in the intention-to-treat analyses. Symptomatic recurrent VTE or death from VTE occurred in 73 of the 829 patients (8.8%) who were receiving placebo, compared with 14 of the 840 patients (1.7%) who were receiving 2.5 mg of apixaban (a difference of 7.2 percentage points; 95% confidence interval [CI], 5.0 to 9.3) and 14 of the 813 patients (1.7%) who were receiving 5 mg of apixaban (a difference of 7.0 percentage points; 95% CI, 4.9 to 9.1) (P<0.001 for both comparisons). The rates of major bleeding were 0.5% in the placebo group, 0.2% in the 2.5-mg apixaban group, and 0.1% in the 5-mg apixaban group. The rates of clinically relevant nonmajor bleeding were 2.3% in the placebo group, 3.0% in the 2.5-mg apixaban group, and 4.2% in the 5-mg apixaban group. The rate of death from any cause was 1.7% in the placebo group, compared with 0.8% in the 2.5-mg apixaban group and 0.5% in the 5-mg apixaban group.

Bottom line: Apixaban is a safe and effective anticoagulant for extended anticoagulation in patients with VTE initially treated with six to 12 months of warfarin.

Citation: Agnelli GM, Buller HR, Cohen A, et al. Apixaban for extended treatment of venous thromboembolism. N Engl J Med. 2013;368(8):699-708.

ACP Guideline Review: Upper Endoscopy for Gastroesophageal Reflux Disease

Clinical question: What are the indications of upper endoscopy in the setting of gastroesophageal reflux disease (GERD)?

Background: GERD is a common condition. Upper endoscopy is widely available and routinely used for diagnosis and management of GERD and its complications. The indications for this procedure are not clearly defined. Overuse of upper endoscopy contributes to higher healthcare costs without improving patient outcomes.

Study design: Literature review and comparison of clinical guidelines from professional organizations by a team of general internists, gastroenterologists, and clinical epidemiologists. The document was not based on a formal systemic review but was intended to provide practical advice based on the best available evidence.

Synopsis: Best practice advice No. 1: Upper endoscopy is indicated in men and women with heartburn and alarm symptoms (dysphagia, bleeding, anemia, weight loss, and recurrent vomiting).

Best practice advice No. 2: Upper endoscopy is indicated in men and women with typical GERD symptoms that persist despite a therapeutic trial of four to eight weeks of twice-daily proton-pump inhibitor therapy, severe erosive esophagitis after a two-month course of proton-pump inhibitor therapy to assess healing and rule out Barrett esophagus, and history of esophageal stricture who have recurrent symptoms of dysphagia.

Best practice advice No. 3: Upper endoscopy might be indicated in men older than 50 with chronic GERD symptoms (symptoms for more than five years) and additional risk factors (nocturnal reflux symptoms, hiatal hernia, elevated body mass index, tobacco use, and intra-abdominal distribution of fat) to detect esophageal adenocarcinoma and Barrett esophagus. It might also be indicated for surveillance evaluation in men and women with a history of Barrett esophagus. In men and women with Barrett esophagus and no dysplasia, surveillance examinations should occur at intervals no more frequently than three to five years. More frequent intervals are indicated in patients with Barrett esophagus and dysplasia.

Bottom line: Use upper endoscopy selectively for patients with GERD.

Citation: Shaheen NJ, Weinberg DS, Denberg TD, et al. Upper endoscopy for gastroesophageal reflux disease: best practice advice from the Clinical Guidelines Committee of the American College of Physicians. Ann Intern Med. 2012;157(11):808-816.

In This Edition

Literature At A Glance

A guide to this month’s studies

1. Value of routine preoperative urine screening assessed
2. Impact of hospitalist-led intermediate care on patient survival
3. Risks of blood transfusion to treat upper GI bleeding
4. Low-dose steroids and increased mortality in sepsis
5. Reduced length of stay and hospital readmission rates
6. Restrictive blood transfusion strategies better for acute myocardial infarction
7. Trends in GI illnesses and their associated costs
8. Apixaban as a stand-alone anticoagulant in patients with VTE
9. Guidelines for upper endoscopy use in gastroesophageal reflux disease

Avoid Preoperative Urine Culture in Nonurologic Surgical Procedures

Clinical question: Is routine preoperative urine screening beneficial?

Background: The value of preoperative urine screening is unproven, except before urologic procedures. Furthermore, treatment of asymptomatic bacteriuria may lead to adverse events, including diarrhea, allergic reactions, and Clostridium difficile infection (CDI).

Study design: Retrospective chart review.

Setting: Patients who underwent cardiothoracic, orthopedic, and vascular surgeries at the Minneapolis Veterans Affairs Medical Center in 2010.

Synopsis: A total of 1,934 procedures were performed on 1,699 patients, most of which were orthopedics procedures (1,291 in 1,115 patients). A urine culture was obtained before 25% of procedures with significant variation by service (cardiothoracic, 85%; vascular, 48%; orthopedic, 4%). Bacteriuria was detected in 11% of urine cultures (54 of 489), but antimicrobial drugs were dispensed to just 16 patients.

To identify correlates of preoperative urine culture use, patients with and without urine cultures were compared. The rate of surgical-site infection was similar for both groups. Postoperative UTI was more frequent among patients with bacteriuria. Rates of diarrhea, allergy, and CDI did not differ. Paradoxically, patients treated for preoperative UTI were more likely to develop surgical-site infections (45% vs. 14%; P=0.03). Postoperative UTI was also more frequent among treated patients versus untreated patients (18% vs. 7%).

Bottom line: This is the largest study to assess outcomes for routine preoperative urine cultures. These findings demonstrate that preoperative screening for, and treatment of, asymptomatic bacteriuria should be avoided in patients undergoing nonurologic surgical procedures.

Citation: Drekonja DM, Zarmbinski B, Johnson JR. Preoperative urine culture at a veterans affairs medical center. JAMA Intern Med. 2013;173(1):71-72.

Intermediate Care Staffed by Hospitalists: Impact on Mortality, Comanagement, and Teaching

Clinical question: Does a hospitalist-led intermediate-care unit improve patient survival?

Background: Hospitalized patients are complex, and institutions often have to balance matching patient acuity to either an ICU or a regular ward. However, an intermediate-care setting might be an attractive strategy to provide rational care according to patient needs while expanding comanagement and teaching services.

Study design: Retrospective observational study.

Setting: Intermediate-care unit of a single academic hospital.

Synopsis: In-hospital mortality in this intermediate-care unit was 20.6%, whereas the expected mortality was 23.2% based on Simplified Acute Physiology Score II (SAPS II) score. The correlation between SAPS II predicted and observed death rates was accurate and statistically significant (P<0.001). Comanagement was performed with several medical and surgical teams, with an increase in perioperative comanagement of 22.7% (P=0.014). The number of training residents in the intermediate-care unit increased to 30.4% from 4.3% (P=0.002).

Bottom line: An intermediate-care unit led by hospitalists showed encouraging results in patient mortality, as well as comanagement and teaching opportunities.

Citation: Lucena JF, Alegre F, Rodil R, et al. Results of a retrospective observational study of intermediate care staffed by hospitalists: impact on mortality, co-management, and teaching. J Hosp Med. 2012;7(5):411-415.

Blood Transfusion Associated with Increased Risk of Rebleeding in Patients with Nonvariceal Upper GI Bleeding

Clinical question: Does more liberal use of blood transfusions in the setting of nonvariceal upper GI bleeding result in patient harm?

Background: Randomized controlled trials have demonstrated that a more liberal approach to blood transfusions for patients in the medical intensive-care unit results in higher mortality. However, the potential harmful effect of blood transfusions in the setting of GI bleeding has not been demonstrated.

Study design: Retrospective cohort study.

Setting: Canadian hospitals.

Synopsis: Based on a retrospective analysis of the Canadian Registry of patients with Upper Gastrointestinal Bleeding and Endoscopy (RUGBE), the authors determined there was a statistically significant association between patients who received a blood transfusion for the management of nonvariceal upper GI bleeding and the risk of rebleeding. The rate of rebleeding in patients who received a blood transfusion was 23.6% compared with 11.3% in patients who were not transfused (P<0.01). There was no statistically significant difference in mortality.

Although this was a reasonably large observational study that included 1,677 patients with nonvariceal upper GI bleeding, it is vulnerable to confounding. It suggests the need to further study potential harm of blood transfusion in the setting of GI bleeding, but it should not result in a change in clinical practice at this time.

Bottom line: Prospective randomized studies are needed to determine if there are harmful effects of blood transfusions in the setting of GI bleeding and to better define a threshold for transfusion.

Citation: Restellini S, Kherad O, Jairath V, Martel M, Barkun AN. Red blood cell transfusion is associated with increased rebleeding in patients with nonvariceal upper gastrointestinal bleeding. Aliment Pharmacol Ther. 2013;37:316-322.

Low-Dose Steroids in Sepsis Associated with Increase in Mortality

Clinical question: What is the role of steroids in the treatment of adult patients with sepsis?

Background: The Surviving Sepsis Campaign guidelines have previously recommended administering steroids to patients with septic shock not responsive to fluid resuscitation and who require vasopressors. However, prior randomized clinical trials studying the use of steroids in these settings have produced conflicting results.

Study design: Retrospective cohort study.

Setting: Two hundred fifty-two hospitals in North America, South America, and Europe.

Synopsis: The Surviving Sepsis Campaign management bundle has been shown to reduce mortality in patients with sepsis. However, it is not known which particular elements of the management bundle result in improved mortality. The Surviving Sepsis Campaign database included 17,847 patients who required vasopressor therapy after adequate fluid resuscitation. This subgroup was analyzed to see if there was a difference in mortality between patients who received low-dose steroids versus those who did not receive steroids. The mortality rate among those who received steroids was statistically higher (with odds ratio of 1.18 and P<0.001) compared with those who did not receive steroids. This finding adds to the body of evidence that calls into question the commonplace practice of administrating steroids to septic patients on vasopressor therapy.

The most recent campaign guidelines recommend the use of steroids in septic patients only if both adequate fluid resuscitation and vasopressor therapy are not able to restore hemodynamic stability.

Bottom line: Further studies are needed to better define the role of steroids in the treatment of sepsis.

Citation: Casserly B, Gerlach H, Phillips GS, et al. Low-dose steroids in adult septic shock: results of the Surviving Sepsis Campaign. Intensive Care Med. 2012;38:1946-1954.

Link Between Length of Stay and Readmission Rates

Clinical question: How has reducing length of stay affected hospital readmission rates?

Background: There are ongoing concerns that improving a hospital’s efficiency by reducing length of stay (LOS) could be associated with higher hospital readmission rates. However, no studies evaluating the relationship between LOS and readmission rates have been done using recent data.

Study design: Retrospective observational study.

Setting: All acute-care Veterans Affairs (VA) hospitals in the U.S.

Synopsis: A total of 4,124,907 index admissions were included in the final sample from all acute medical admissions in 129 acute-care VA hospitals from October 1996 to September 2010. The primary outcomes were the hospital LOS and the 30-day readmission rate. Index admissions for heart failure, chronic obstructive pulmonary disease (COPD), acute myocardial infarction (AMI), community-acquired pneumonia, and gastrointestinal hemorrhage were also analyzed separately.

The risk-adjusted analysis of LOS demonstrated significant reductions for all admissions over the 14-year period, to 3.98 days from 5.44 days, and for all of the individual conditions with reductions ranging from 1.40 days for gastrointestinal hemorrhage to 2.85 days for AMI. There were similar significant reductions in 30-day readmission rates for all admissions to 13.8% from 16.5% and within the individual conditions ranging from 0.9% in community-acquired pneumonia to 3.3% in COPD. These results show that the reductions in LOS did not increase the risk of readmissions. The major limitation of the study was that these data are only from a single healthcare system.

Bottom line: Data from VA hospitals show that reductions in LOS do not have adverse effects on 30-day readmission rates; instead, both LOS and readmission rates improved over the same time period.

Citation: Kaboli PJ, Go JT, Hockenberry J, et al. Associations between reduced hospital length of stay and 30-day readmission rate and mortality: 14-year experience in 129 Veterans Affairs hospitals. Ann Intern Med. 2012;157:837-845.

Better to Restrict Blood Transfusions in Acute Myocardial Infarction

Clinical question: Is a liberal or restrictive blood transfusion strategy better in patients with anemia and acute myocardial infarction?

Background: Patients with acute myocardial infarction (AMI) are often given therapies that can increase their risk for bleeding and anemia, and it is known that AMI patients have a worse prognosis if they have concomitant anemia. No clear consensus exists on the benefit or harm of blood transfusions in AMI patients.

Study design: Systematic review and meta-analysis.

Setting: Ten articles included in the qualitative and quantitative analyses out of 729 screened articles from Jan. 1, 1966, to March 31, 2012, using the search terms “transfusion,” “myocardial infarction,” and “mortality” in English language.

Synopsis: A total of 203,665 study participants were identified from the 10 studies (one randomized and nine observational) that met the inclusion and exclusion criteria. All-cause mortality was significantly higher in AMI patients who received a blood transfusion compared with those who did not (18.2% vs. 10.2%). However, this difference was not statistically significant in patients that had a STEMI or in patients with a baseline hematocrit less than 30%. A multivariate meta-regression with several covariates, excluding demographics, also showed that blood transfusion was associated with higher mortality and higher risk for subsequent myocardial infarction. There was significant heterogeneity in all results, but no single study was found as the source of the heterogeneity, and no significant publication bias was identified. The major limitations to this study are that there is a paucity of randomized trials available that pertain to this specific topic and the authors did not have patient-level covariates to include in their analyses.

Bottom line: There appears to be an increased risk of mortality and subsequent myocardial infarction in AMI patients who receive blood transfusions versus those who do not.

Citation: Chatterjee S, Wetterslev J, Sharma A, Lichstein E, Mukherjee D. Association of blood transfusion with increased mortality in myocardial infarction: a meta-analysis and diversity-adjusted study sequential analysis. JAMA Intern Med. 2013;173(2):132-139.

Trends in GI Illnesses and Their Associated Costs

Clinical question: What are the new trends in GI illnesses and their associated costs?

Background: The frequency of illnesses and their treatment costs have changed over the last decade. In order to help healthcare providers focus their attention on these new trends, a new compilation of data is needed.

Study design: Epidemiological analysis.

Setting: Various governmental and private databases representing outpatient clinics, hospitals, and death certificates from multiple regions of the U.S.

Synopsis: The analysis was blinded to patient identifiers but represented multiple regions of the U.S. Symptoms were abstracted from patient surveys, and the rest of the data were collected from record review. The most common reported symptoms were abdominal pain, followed by nausea, vomiting, diarrhea, constipation, and heartburn. The most common clinic diagnoses were reflux, abdominal pain, enteritis/dyspepsia, and constipation. The most common inpatient discharge primary diagnoses included acute pancreatitis, cholecystitis, and diverticulitis. Impressive increases were seen in the number of morbidly obese, C. diff, and fatty liver diagnoses.

Colon cancer was the most common GI malignancy and had the highest mortality. C. diff was the ninth-leading cause of GI-related deaths. All types of scopes (except endoscopic retrograde cholangiopancreatography) were performed more commonly now than in the past, with colonoscopy being the most common. The most common indication for an upper endoscopy was reflux, which was also the most common outpatient GI diagnosis.

Bottom line: Healthcare providers need to be aware of new GI illness trends and their associated costs.

Citation: Peery AF, Dellon ES, Lund J, et al. Burden of gastrointestinal disease in the United States: 2012 update. Gastroenterol. 2012;143:1179-1187.

Is Apixaban a Good Stand-Alone Anticoagulant for Extended Treatment in VTE Patients?

Clinical question: Is apixaban an option for the extended treatment of VTE in a simple, fixed-dose regimen?

Background: Apixaban is an oral factor Xa inhibitor that is administered in fixed doses without the need for laboratory monitoring. In the Apixaban after the Initial Management of Pulmonary Embolism and Deep Vein Thrombosis with First-Line Therapy—Extended Treatment (AMPLIFY-EXT) study, investigators compared the efficacy and safety of two doses of apixaban (2.5 mg and 5 mg) with those of placebo in patients with VTE who had completed six to 12 months of anticoagulation therapy and for whom treating physicians were uncertain about continuing therapy. Additional aims of the study were to determine whether the lower dose of apixaban was effective and whether it was associated with less bleeding than the higher dose, and to examine the effect of treatment on arterial thrombotic outcomes.

Study design: Randomized, double-blind study.

Synopsis: A total of 2,486 patients underwent randomization, 2,482 of whom were included in the intention-to-treat analyses. Symptomatic recurrent VTE or death from VTE occurred in 73 of the 829 patients (8.8%) who were receiving placebo, compared with 14 of the 840 patients (1.7%) who were receiving 2.5 mg of apixaban (a difference of 7.2 percentage points; 95% confidence interval [CI], 5.0 to 9.3) and 14 of the 813 patients (1.7%) who were receiving 5 mg of apixaban (a difference of 7.0 percentage points; 95% CI, 4.9 to 9.1) (P<0.001 for both comparisons). The rates of major bleeding were 0.5% in the placebo group, 0.2% in the 2.5-mg apixaban group, and 0.1% in the 5-mg apixaban group. The rates of clinically relevant nonmajor bleeding were 2.3% in the placebo group, 3.0% in the 2.5-mg apixaban group, and 4.2% in the 5-mg apixaban group. The rate of death from any cause was 1.7% in the placebo group, compared with 0.8% in the 2.5-mg apixaban group and 0.5% in the 5-mg apixaban group.

Bottom line: Apixaban is a safe and effective anticoagulant for extended anticoagulation in patients with VTE initially treated with six to 12 months of warfarin.

Citation: Agnelli GM, Buller HR, Cohen A, et al. Apixaban for extended treatment of venous thromboembolism. N Engl J Med. 2013;368(8):699-708.

ACP Guideline Review: Upper Endoscopy for Gastroesophageal Reflux Disease

Clinical question: What are the indications of upper endoscopy in the setting of gastroesophageal reflux disease (GERD)?

Background: GERD is a common condition. Upper endoscopy is widely available and routinely used for diagnosis and management of GERD and its complications. The indications for this procedure are not clearly defined. Overuse of upper endoscopy contributes to higher healthcare costs without improving patient outcomes.

Study design: Literature review and comparison of clinical guidelines from professional organizations by a team of general internists, gastroenterologists, and clinical epidemiologists. The document was not based on a formal systemic review but was intended to provide practical advice based on the best available evidence.

Synopsis: Best practice advice No. 1: Upper endoscopy is indicated in men and women with heartburn and alarm symptoms (dysphagia, bleeding, anemia, weight loss, and recurrent vomiting).

Best practice advice No. 2: Upper endoscopy is indicated in men and women with typical GERD symptoms that persist despite a therapeutic trial of four to eight weeks of twice-daily proton-pump inhibitor therapy, severe erosive esophagitis after a two-month course of proton-pump inhibitor therapy to assess healing and rule out Barrett esophagus, and history of esophageal stricture who have recurrent symptoms of dysphagia.

Best practice advice No. 3: Upper endoscopy might be indicated in men older than 50 with chronic GERD symptoms (symptoms for more than five years) and additional risk factors (nocturnal reflux symptoms, hiatal hernia, elevated body mass index, tobacco use, and intra-abdominal distribution of fat) to detect esophageal adenocarcinoma and Barrett esophagus. It might also be indicated for surveillance evaluation in men and women with a history of Barrett esophagus. In men and women with Barrett esophagus and no dysplasia, surveillance examinations should occur at intervals no more frequently than three to five years. More frequent intervals are indicated in patients with Barrett esophagus and dysplasia.

Bottom line: Use upper endoscopy selectively for patients with GERD.

Citation: Shaheen NJ, Weinberg DS, Denberg TD, et al. Upper endoscopy for gastroesophageal reflux disease: best practice advice from the Clinical Guidelines Committee of the American College of Physicians. Ann Intern Med. 2012;157(11):808-816.

In This Edition

Literature At A Glance

A guide to this month’s studies

1. Value of routine preoperative urine screening assessed
2. Impact of hospitalist-led intermediate care on patient survival
3. Risks of blood transfusion to treat upper GI bleeding
4. Low-dose steroids and increased mortality in sepsis
5. Reduced length of stay and hospital readmission rates
6. Restrictive blood transfusion strategies better for acute myocardial infarction
7. Trends in GI illnesses and their associated costs
8. Apixaban as a stand-alone anticoagulant in patients with VTE
9. Guidelines for upper endoscopy use in gastroesophageal reflux disease

Avoid Preoperative Urine Culture in Nonurologic Surgical Procedures

Clinical question: Is routine preoperative urine screening beneficial?

Background: The value of preoperative urine screening is unproven, except before urologic procedures. Furthermore, treatment of asymptomatic bacteriuria may lead to adverse events, including diarrhea, allergic reactions, and Clostridium difficile infection (CDI).

Study design: Retrospective chart review.

Setting: Patients who underwent cardiothoracic, orthopedic, and vascular surgeries at the Minneapolis Veterans Affairs Medical Center in 2010.

Synopsis: A total of 1,934 procedures were performed on 1,699 patients, most of which were orthopedics procedures (1,291 in 1,115 patients). A urine culture was obtained before 25% of procedures with significant variation by service (cardiothoracic, 85%; vascular, 48%; orthopedic, 4%). Bacteriuria was detected in 11% of urine cultures (54 of 489), but antimicrobial drugs were dispensed to just 16 patients.

To identify correlates of preoperative urine culture use, patients with and without urine cultures were compared. The rate of surgical-site infection was similar for both groups. Postoperative UTI was more frequent among patients with bacteriuria. Rates of diarrhea, allergy, and CDI did not differ. Paradoxically, patients treated for preoperative UTI were more likely to develop surgical-site infections (45% vs. 14%; P=0.03). Postoperative UTI was also more frequent among treated patients versus untreated patients (18% vs. 7%).

Bottom line: This is the largest study to assess outcomes for routine preoperative urine cultures. These findings demonstrate that preoperative screening for, and treatment of, asymptomatic bacteriuria should be avoided in patients undergoing nonurologic surgical procedures.

Citation: Drekonja DM, Zarmbinski B, Johnson JR. Preoperative urine culture at a veterans affairs medical center. JAMA Intern Med. 2013;173(1):71-72.

Intermediate Care Staffed by Hospitalists: Impact on Mortality, Comanagement, and Teaching

Clinical question: Does a hospitalist-led intermediate-care unit improve patient survival?

Background: Hospitalized patients are complex, and institutions often have to balance matching patient acuity to either an ICU or a regular ward. However, an intermediate-care setting might be an attractive strategy to provide rational care according to patient needs while expanding comanagement and teaching services.

Study design: Retrospective observational study.

Setting: Intermediate-care unit of a single academic hospital.

Synopsis: In-hospital mortality in this intermediate-care unit was 20.6%, whereas the expected mortality was 23.2% based on Simplified Acute Physiology Score II (SAPS II) score. The correlation between SAPS II predicted and observed death rates was accurate and statistically significant (P<0.001). Comanagement was performed with several medical and surgical teams, with an increase in perioperative comanagement of 22.7% (P=0.014). The number of training residents in the intermediate-care unit increased to 30.4% from 4.3% (P=0.002).

Bottom line: An intermediate-care unit led by hospitalists showed encouraging results in patient mortality, as well as comanagement and teaching opportunities.

Citation: Lucena JF, Alegre F, Rodil R, et al. Results of a retrospective observational study of intermediate care staffed by hospitalists: impact on mortality, co-management, and teaching. J Hosp Med. 2012;7(5):411-415.

Blood Transfusion Associated with Increased Risk of Rebleeding in Patients with Nonvariceal Upper GI Bleeding

Clinical question: Does more liberal use of blood transfusions in the setting of nonvariceal upper GI bleeding result in patient harm?

Background: Randomized controlled trials have demonstrated that a more liberal approach to blood transfusions for patients in the medical intensive-care unit results in higher mortality. However, the potential harmful effect of blood transfusions in the setting of GI bleeding has not been demonstrated.

Study design: Retrospective cohort study.

Setting: Canadian hospitals.

Synopsis: Based on a retrospective analysis of the Canadian Registry of patients with Upper Gastrointestinal Bleeding and Endoscopy (RUGBE), the authors determined there was a statistically significant association between patients who received a blood transfusion for the management of nonvariceal upper GI bleeding and the risk of rebleeding. The rate of rebleeding in patients who received a blood transfusion was 23.6% compared with 11.3% in patients who were not transfused (P<0.01). There was no statistically significant difference in mortality.

Although this was a reasonably large observational study that included 1,677 patients with nonvariceal upper GI bleeding, it is vulnerable to confounding. It suggests the need to further study potential harm of blood transfusion in the setting of GI bleeding, but it should not result in a change in clinical practice at this time.

Bottom line: Prospective randomized studies are needed to determine if there are harmful effects of blood transfusions in the setting of GI bleeding and to better define a threshold for transfusion.

Citation: Restellini S, Kherad O, Jairath V, Martel M, Barkun AN. Red blood cell transfusion is associated with increased rebleeding in patients with nonvariceal upper gastrointestinal bleeding. Aliment Pharmacol Ther. 2013;37:316-322.

Low-Dose Steroids in Sepsis Associated with Increase in Mortality

Clinical question: What is the role of steroids in the treatment of adult patients with sepsis?

Background: The Surviving Sepsis Campaign guidelines have previously recommended administering steroids to patients with septic shock not responsive to fluid resuscitation and who require vasopressors. However, prior randomized clinical trials studying the use of steroids in these settings have produced conflicting results.

Study design: Retrospective cohort study.

Setting: Two hundred fifty-two hospitals in North America, South America, and Europe.

Synopsis: The Surviving Sepsis Campaign management bundle has been shown to reduce mortality in patients with sepsis. However, it is not known which particular elements of the management bundle result in improved mortality. The Surviving Sepsis Campaign database included 17,847 patients who required vasopressor therapy after adequate fluid resuscitation. This subgroup was analyzed to see if there was a difference in mortality between patients who received low-dose steroids versus those who did not receive steroids. The mortality rate among those who received steroids was statistically higher (with odds ratio of 1.18 and P<0.001) compared with those who did not receive steroids. This finding adds to the body of evidence that calls into question the commonplace practice of administrating steroids to septic patients on vasopressor therapy.

The most recent campaign guidelines recommend the use of steroids in septic patients only if both adequate fluid resuscitation and vasopressor therapy are not able to restore hemodynamic stability.

Bottom line: Further studies are needed to better define the role of steroids in the treatment of sepsis.

Citation: Casserly B, Gerlach H, Phillips GS, et al. Low-dose steroids in adult septic shock: results of the Surviving Sepsis Campaign. Intensive Care Med. 2012;38:1946-1954.

Link Between Length of Stay and Readmission Rates

Clinical question: How has reducing length of stay affected hospital readmission rates?

Background: There are ongoing concerns that improving a hospital’s efficiency by reducing length of stay (LOS) could be associated with higher hospital readmission rates. However, no studies evaluating the relationship between LOS and readmission rates have been done using recent data.

Study design: Retrospective observational study.

Setting: All acute-care Veterans Affairs (VA) hospitals in the U.S.

Synopsis: A total of 4,124,907 index admissions were included in the final sample from all acute medical admissions in 129 acute-care VA hospitals from October 1996 to September 2010. The primary outcomes were the hospital LOS and the 30-day readmission rate. Index admissions for heart failure, chronic obstructive pulmonary disease (COPD), acute myocardial infarction (AMI), community-acquired pneumonia, and gastrointestinal hemorrhage were also analyzed separately.

The risk-adjusted analysis of LOS demonstrated significant reductions for all admissions over the 14-year period, to 3.98 days from 5.44 days, and for all of the individual conditions with reductions ranging from 1.40 days for gastrointestinal hemorrhage to 2.85 days for AMI. There were similar significant reductions in 30-day readmission rates for all admissions to 13.8% from 16.5% and within the individual conditions ranging from 0.9% in community-acquired pneumonia to 3.3% in COPD. These results show that the reductions in LOS did not increase the risk of readmissions. The major limitation of the study was that these data are only from a single healthcare system.

Bottom line: Data from VA hospitals show that reductions in LOS do not have adverse effects on 30-day readmission rates; instead, both LOS and readmission rates improved over the same time period.

Citation: Kaboli PJ, Go JT, Hockenberry J, et al. Associations between reduced hospital length of stay and 30-day readmission rate and mortality: 14-year experience in 129 Veterans Affairs hospitals. Ann Intern Med. 2012;157:837-845.

Better to Restrict Blood Transfusions in Acute Myocardial Infarction

Clinical question: Is a liberal or restrictive blood transfusion strategy better in patients with anemia and acute myocardial infarction?

Background: Patients with acute myocardial infarction (AMI) are often given therapies that can increase their risk for bleeding and anemia, and it is known that AMI patients have a worse prognosis if they have concomitant anemia. No clear consensus exists on the benefit or harm of blood transfusions in AMI patients.

Study design: Systematic review and meta-analysis.

Setting: Ten articles included in the qualitative and quantitative analyses out of 729 screened articles from Jan. 1, 1966, to March 31, 2012, using the search terms “transfusion,” “myocardial infarction,” and “mortality” in English language.

Synopsis: A total of 203,665 study participants were identified from the 10 studies (one randomized and nine observational) that met the inclusion and exclusion criteria. All-cause mortality was significantly higher in AMI patients who received a blood transfusion compared with those who did not (18.2% vs. 10.2%). However, this difference was not statistically significant in patients that had a STEMI or in patients with a baseline hematocrit less than 30%. A multivariate meta-regression with several covariates, excluding demographics, also showed that blood transfusion was associated with higher mortality and higher risk for subsequent myocardial infarction. There was significant heterogeneity in all results, but no single study was found as the source of the heterogeneity, and no significant publication bias was identified. The major limitations to this study are that there is a paucity of randomized trials available that pertain to this specific topic and the authors did not have patient-level covariates to include in their analyses.

Bottom line: There appears to be an increased risk of mortality and subsequent myocardial infarction in AMI patients who receive blood transfusions versus those who do not.

Citation: Chatterjee S, Wetterslev J, Sharma A, Lichstein E, Mukherjee D. Association of blood transfusion with increased mortality in myocardial infarction: a meta-analysis and diversity-adjusted study sequential analysis. JAMA Intern Med. 2013;173(2):132-139.

Trends in GI Illnesses and Their Associated Costs

Clinical question: What are the new trends in GI illnesses and their associated costs?

Background: The frequency of illnesses and their treatment costs have changed over the last decade. In order to help healthcare providers focus their attention on these new trends, a new compilation of data is needed.

Study design: Epidemiological analysis.

Setting: Various governmental and private databases representing outpatient clinics, hospitals, and death certificates from multiple regions of the U.S.

Synopsis: The analysis was blinded to patient identifiers but represented multiple regions of the U.S. Symptoms were abstracted from patient surveys, and the rest of the data were collected from record review. The most common reported symptoms were abdominal pain, followed by nausea, vomiting, diarrhea, constipation, and heartburn. The most common clinic diagnoses were reflux, abdominal pain, enteritis/dyspepsia, and constipation. The most common inpatient discharge primary diagnoses included acute pancreatitis, cholecystitis, and diverticulitis. Impressive increases were seen in the number of morbidly obese, C. diff, and fatty liver diagnoses.

Colon cancer was the most common GI malignancy and had the highest mortality. C. diff was the ninth-leading cause of GI-related deaths. All types of scopes (except endoscopic retrograde cholangiopancreatography) were performed more commonly now than in the past, with colonoscopy being the most common. The most common indication for an upper endoscopy was reflux, which was also the most common outpatient GI diagnosis.

Bottom line: Healthcare providers need to be aware of new GI illness trends and their associated costs.

Citation: Peery AF, Dellon ES, Lund J, et al. Burden of gastrointestinal disease in the United States: 2012 update. Gastroenterol. 2012;143:1179-1187.

Is Apixaban a Good Stand-Alone Anticoagulant for Extended Treatment in VTE Patients?

Clinical question: Is apixaban an option for the extended treatment of VTE in a simple, fixed-dose regimen?

Background: Apixaban is an oral factor Xa inhibitor that is administered in fixed doses without the need for laboratory monitoring. In the Apixaban after the Initial Management of Pulmonary Embolism and Deep Vein Thrombosis with First-Line Therapy—Extended Treatment (AMPLIFY-EXT) study, investigators compared the efficacy and safety of two doses of apixaban (2.5 mg and 5 mg) with those of placebo in patients with VTE who had completed six to 12 months of anticoagulation therapy and for whom treating physicians were uncertain about continuing therapy. Additional aims of the study were to determine whether the lower dose of apixaban was effective and whether it was associated with less bleeding than the higher dose, and to examine the effect of treatment on arterial thrombotic outcomes.

Study design: Randomized, double-blind study.

Synopsis: A total of 2,486 patients underwent randomization, 2,482 of whom were included in the intention-to-treat analyses. Symptomatic recurrent VTE or death from VTE occurred in 73 of the 829 patients (8.8%) who were receiving placebo, compared with 14 of the 840 patients (1.7%) who were receiving 2.5 mg of apixaban (a difference of 7.2 percentage points; 95% confidence interval [CI], 5.0 to 9.3) and 14 of the 813 patients (1.7%) who were receiving 5 mg of apixaban (a difference of 7.0 percentage points; 95% CI, 4.9 to 9.1) (P<0.001 for both comparisons). The rates of major bleeding were 0.5% in the placebo group, 0.2% in the 2.5-mg apixaban group, and 0.1% in the 5-mg apixaban group. The rates of clinically relevant nonmajor bleeding were 2.3% in the placebo group, 3.0% in the 2.5-mg apixaban group, and 4.2% in the 5-mg apixaban group. The rate of death from any cause was 1.7% in the placebo group, compared with 0.8% in the 2.5-mg apixaban group and 0.5% in the 5-mg apixaban group.

Bottom line: Apixaban is a safe and effective anticoagulant for extended anticoagulation in patients with VTE initially treated with six to 12 months of warfarin.

Citation: Agnelli GM, Buller HR, Cohen A, et al. Apixaban for extended treatment of venous thromboembolism. N Engl J Med. 2013;368(8):699-708.

ACP Guideline Review: Upper Endoscopy for Gastroesophageal Reflux Disease

Clinical question: What are the indications of upper endoscopy in the setting of gastroesophageal reflux disease (GERD)?

Background: GERD is a common condition. Upper endoscopy is widely available and routinely used for diagnosis and management of GERD and its complications. The indications for this procedure are not clearly defined. Overuse of upper endoscopy contributes to higher healthcare costs without improving patient outcomes.

Study design: Literature review and comparison of clinical guidelines from professional organizations by a team of general internists, gastroenterologists, and clinical epidemiologists. The document was not based on a formal systemic review but was intended to provide practical advice based on the best available evidence.

Synopsis: Best practice advice No. 1: Upper endoscopy is indicated in men and women with heartburn and alarm symptoms (dysphagia, bleeding, anemia, weight loss, and recurrent vomiting).

Best practice advice No. 2: Upper endoscopy is indicated in men and women with typical GERD symptoms that persist despite a therapeutic trial of four to eight weeks of twice-daily proton-pump inhibitor therapy, severe erosive esophagitis after a two-month course of proton-pump inhibitor therapy to assess healing and rule out Barrett esophagus, and history of esophageal stricture who have recurrent symptoms of dysphagia.

Best practice advice No. 3: Upper endoscopy might be indicated in men older than 50 with chronic GERD symptoms (symptoms for more than five years) and additional risk factors (nocturnal reflux symptoms, hiatal hernia, elevated body mass index, tobacco use, and intra-abdominal distribution of fat) to detect esophageal adenocarcinoma and Barrett esophagus. It might also be indicated for surveillance evaluation in men and women with a history of Barrett esophagus. In men and women with Barrett esophagus and no dysplasia, surveillance examinations should occur at intervals no more frequently than three to five years. More frequent intervals are indicated in patients with Barrett esophagus and dysplasia.

Bottom line: Use upper endoscopy selectively for patients with GERD.

Citation: Shaheen NJ, Weinberg DS, Denberg TD, et al. Upper endoscopy for gastroesophageal reflux disease: best practice advice from the Clinical Guidelines Committee of the American College of Physicians. Ann Intern Med. 2012;157(11):808-816.

Clinical Question: Is chest-compression-only bystander cardiopulmonary resuscitation (CPR) as effective as conventional CPR with rescue breathing for out-of-hospital cardiac arrest?

Background: Out-of-hospital cardiac arrest is a major public health problem, affecting approximately 300,000 people in the U.S. annually. Overall survival is generally less than 10% among those in whom resuscitation is attempted. Bystander CPR significantly improves outcomes but generally is performed in less than 30% of cases.

Study design: Prospective, observational cohort study.

Setting: Communities throughout the state of Arizona.

Synopsis: State officials undertook a multifaceted public service campaign to educate, inform, and encourage the use of compression-only CPR.

Over a period of four years, 5,272 adult out-of-hospital cardiac arrests were documented in Arizona. A total of 4,415 of these incidents met all of the inclusion criteria for analysis; 2,900 received no CPR, 666 received conventional CPR, and 849 received compression-only CPR. Rates of survival to hospital discharge were 5.2% for the no-CPR group, 7.8% for conventional CPR, and 13.3% for compression-only CPR.

Additionally, the successful public service campaign increased the use of lay rescuer conventional CPR by 10% and compression-only CPR by 56%.

Bottom line: Layperson, compression-only CPR was associated with an overall 6.1% increased survival compared with conventional CPR and no bystander CPR.

Citation: Bobrow BJ, Spaite DW, Berg RA, et al. Chest-compression-only CPR by lay rescuers and survival from out-of-hospital cardiac arrest. JAMA. 2010;304(13):1447-1454.

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

Clinical Question: Is chest-compression-only bystander cardiopulmonary resuscitation (CPR) as effective as conventional CPR with rescue breathing for out-of-hospital cardiac arrest?

Background: Out-of-hospital cardiac arrest is a major public health problem, affecting approximately 300,000 people in the U.S. annually. Overall survival is generally less than 10% among those in whom resuscitation is attempted. Bystander CPR significantly improves outcomes but generally is performed in less than 30% of cases.

Study design: Prospective, observational cohort study.

Setting: Communities throughout the state of Arizona.

Synopsis: State officials undertook a multifaceted public service campaign to educate, inform, and encourage the use of compression-only CPR.

Over a period of four years, 5,272 adult out-of-hospital cardiac arrests were documented in Arizona. A total of 4,415 of these incidents met all of the inclusion criteria for analysis; 2,900 received no CPR, 666 received conventional CPR, and 849 received compression-only CPR. Rates of survival to hospital discharge were 5.2% for the no-CPR group, 7.8% for conventional CPR, and 13.3% for compression-only CPR.

Additionally, the successful public service campaign increased the use of lay rescuer conventional CPR by 10% and compression-only CPR by 56%.

Bottom line: Layperson, compression-only CPR was associated with an overall 6.1% increased survival compared with conventional CPR and no bystander CPR.

Citation: Bobrow BJ, Spaite DW, Berg RA, et al. Chest-compression-only CPR by lay rescuers and survival from out-of-hospital cardiac arrest. JAMA. 2010;304(13):1447-1454.

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

Clinical Question: Is chest-compression-only bystander cardiopulmonary resuscitation (CPR) as effective as conventional CPR with rescue breathing for out-of-hospital cardiac arrest?

Background: Out-of-hospital cardiac arrest is a major public health problem, affecting approximately 300,000 people in the U.S. annually. Overall survival is generally less than 10% among those in whom resuscitation is attempted. Bystander CPR significantly improves outcomes but generally is performed in less than 30% of cases.

Study design: Prospective, observational cohort study.

Setting: Communities throughout the state of Arizona.

Synopsis: State officials undertook a multifaceted public service campaign to educate, inform, and encourage the use of compression-only CPR.

Over a period of four years, 5,272 adult out-of-hospital cardiac arrests were documented in Arizona. A total of 4,415 of these incidents met all of the inclusion criteria for analysis; 2,900 received no CPR, 666 received conventional CPR, and 849 received compression-only CPR. Rates of survival to hospital discharge were 5.2% for the no-CPR group, 7.8% for conventional CPR, and 13.3% for compression-only CPR.

Additionally, the successful public service campaign increased the use of lay rescuer conventional CPR by 10% and compression-only CPR by 56%.

Bottom line: Layperson, compression-only CPR was associated with an overall 6.1% increased survival compared with conventional CPR and no bystander CPR.

Citation: Bobrow BJ, Spaite DW, Berg RA, et al. Chest-compression-only CPR by lay rescuers and survival from out-of-hospital cardiac arrest. JAMA. 2010;304(13):1447-1454.

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

In This Edition

Literature at a Glance

A guide to this month’s studies

• Risk of adverse events with opioid use
• Drug of choice for outpatient treatment of cellulitis
• Preventing hospital falls
• Post-hospital outcomes based on status of PCP follow-up
• LOS, mortality, and readmission based on insurance
• Antiplatelets added to warfarin for atrial fibrillation.
• Cognitive effects of severe sepsis
• Effect of preoperative furosemide use

ED Visits Are Higher among Recipients of Chronic Opioid Therapy

Clinical question: Is there an association between the use of prescription opioids and adverse outcomes?

Background: Chronic opioid therapy is a common strategy for managing chronic, noncancer pain. There has been an increase in overdose deaths and ED visits (EDV) involving the use of prescription opioids.

Study design: Retrospective study from claims records.

Setting: Population in the Health Core Integrated Research Database, containing large, commercial insurance plans in 14 states, and Arkansas Medicaid.

Synopsis: Patients 18 and older without cancer diagnoses who used prescription opioids for at least 90 continuous days within a six-month period from 2000 to 2005 were examined for risk factors for EDVs and alcohol- or drug-related encounters (ADEs) in the 12 months following 90 days or more of prescribed opioids.

Patients with diagnoses of headache, back pain, and pre-existing substance-use disorders had significantly higher EDVs and ADEs. Opioid dose at morphine-equivalent doses over 120 mg per day doubled the risk of ADEs. The use of short-acting Schedule II opioids was associated with EDVs (relative risk, 1.09-1.74). The use of long-acting Schedule II opioids was strongly associated with ADEs (relative risk, 1.64-4.00).

Bottom line: In adults with noncancer pain prescribed opioids for 90 days or more, short-acting Schedule II opioid use was associated with an increased number of EDVs, and long-acting opioid use was associated with an increased number of ADEs. Minimizing Schedule II opioid prescription in these higher-risk patients might be prudent to increase patient safety.

Citation: Braden JB, Russo J, Fan MI, et al. Emergency department visits among recipients of chronic opioid therapy. Arch Intern Med. 2010; 170(16):1425-1432.

Empiric Outpatient Therapy with Trimethoprim-Sulfamethoxazole or Clindamycin Is Preferred for Cellulitis

Clinical question: What is the best empiric outpatient oral antibiotic treatment for cellulitis in areas with a high prevalence of community-associated MRSA infections?

Background: The increasing rates of community-associated MRSA skin and soft-tissue infections have raised concerns that such beta-lactams as cephalexin and other semisynthetic penicillins are not appropriate for empiric outpatient therapy for cellulitis.

Study design: Three-year, retrospective cohort study.

Setting: A teaching clinic of a tertiary-care medical center in Hawaii.

Synopsis: More than 540 patients with cellulitis were identified from January 2005 to December 2007. Of these, 139 patients were excluded for reasons such as hospitalization, surgical intervention, etc. In the final cohort of 405 patients, the three most commonly prescribed oral antibiotics were cephalexin (44%), trimethoprim-sulfamethoxazole (38%), and clindamycin (10%). Other antibiotics accounted for the remaining 8%.

MRSA was recovered in 62% of positive culture specimens. The success rate of trimethoprim-sulfamethoxazole was 91% vs. 74% in the cephalexin group (P<0.001). Clindamycin success rates were higher than those of cephalexin in patients who had subsequently confirmed MRSA infections (P=0.01) and moderately severe cellulitis (P=0.03) and were obese (P=0.04).

Bottom line: Antibiotics with activity against community-acquired MRSA (e.g. trimethroprim-sulfamethoxazole and clinidamycin) are the preferred empiric outpatient therapy for cellulitis in areas with a high prevalence of community-acquired MRSA.

Citation: Khawcharoenporn T, Tice A. Empiric outpatient therapy with trimethoprim-sulfamethoxazole, cephalexin, or clindamycin for cellulitis. Am J Med. 2010;123(10):942-950.

Patient-Specific Screening with Health Information Technology Prevents Falls

Clinical question: Does a fall-prevention toolkit using health information technology decrease patient falls in acute-care hospitals?

Background: Inpatient falls and fall-related injuries result in substantial morbidity and additional healthcare costs. While specific fall-prevention strategies were a longstanding target for intervention, little evidence exists to link them with decreased fall rates.

Study design: Cluster-randomized study.

Setting: Four urban hospitals in Massachusetts.

Synopsis: Comparing patient fall rates in four acute-care hospitals between units providing usual care (5,104 patients) and units using a health information technology (HIT)-linked fall prevention toolkit (5,160 patients), this study demonstrated significant fall reduction in older inpatients. The intervention integrated existing workflow and validated fall risk assessment (Morse Falls Scale) into an HIT software application that tailored fall-prevention interventions to patients’ specific fall risk determinants. The toolkit produced bed posters, patient education handouts, and plans of care communicating patient-specific alerts to key stakeholders.

The primary outcome was patient falls per 1,000 patient-days during the six-month intervention period. The number of patients with falls was significantly different (P=0.02) between control (n=87) and intervention (n=67) units. The toolkit prevented one fall per 862 patient-days.

This nonblinded study was limited by the fact that it was conducted in a single health system. The toolkit was not effective in patients less than 65 years of age. Additionally, the sample size did not have sufficient power to detect effectiveness in preventing repeat falls or falls with injury.

Bottom line: Patient-specific fall prevention strategy coupled with HIT reduces falls in older inpatients.

Citation: Dykes PC, Carroll DL, Hurley A, et al. Fall prevention in acute care hospitals: a randomized trial. JAMA. 2010;304(17):1912-1918.

Lack of Timely Outpatient Follow-Up Results in Higher Readmission Rates

Clinical question: Does timely primary-care-physician (PCP) follow-up improve outcomes and prevent hospital readmissions?

Background: Active PCP involvement is key to interventions aimed at reducing readmissions and ensuring effective ongoing patient care. Some studies suggest increased overall resource utilization when PCP follow-up occurs after hospitalization. Resource utilization and clinical outcomes after hospitalization related to timely PCP follow-up have not been adequately studied.

Study design: Prospective cohort.

Setting: An urban, academic, 425-bed tertiary-care center in Colorado.

Synopsis: From a convenience sample of 121 patients admitted to general medicine services during winter months, 65 patients completed the study. Demographics, diagnosis, payor source, and PCP information were collected upon enrollment. Post-discharge phone calls and patient surveys were used to determine follow-up and readmission status. Timely PCP follow-up was defined as a visit with a PCP or specialist related to the discharge diagnosis within four weeks of hospital discharge.

Thirty-day readmission rates and hospital length of stay were compared for those with timely PCP follow-up and those without. Less than half of general-medicine inpatients received timely PCP follow-up post-discharge. Lack of timely PCP follow-up was associated with younger age, a 10-fold increase in 30-day readmission for the same condition, and a trend toward longer length of stay. However, hospital readmission for any condition did not differ with lack of timely PCP follow-up.

This small, single-center study with convenience sample enrollment might not represent all medical inpatients or diagnoses. Determination of same-condition readmission was potentially subjective.

Bottom line: Patients who lack timely post-discharge follow-up have higher readmission rates for the same medical condition.

Citation: Misky GJ, Wald HL, Coleman EA. Post-hospitalization transitions: examining the effects of timing of primary care provider follow-up. J Hosp Med. 2010;5(7):392-397.

Compared with Uninsured and Medicaid Patients, Privately Insured Patients Admitted for Acute MI, Stroke, and Pneumonia Have Better Mortality Rates

Clinical question: Do outcomes for insured and underinsured patients vary for three of the most common medical conditions for which patients are hospitalized: acute myocardial infarction (AMI), stroke, and pneumonia?

Background: The ideal healthcare system would provide quality care to all individuals regardless of insurance status. Nevertheless, disparities in outcomes for the insured and underinsured or uninsured are well-documented in the outpatient setting but not as well in the inpatient setting. More needs to be done to address these potential disparities.

Study design: Retrospective database analysis.

Setting: Database including 20% of all U.S. community hospitals, including public hospitals, academic medical centers, and specialty hospitals.

Synopsis: This study utilized a database of 8 million discharges from more than 1,000 hospitals and isolated patients 18-64 years old (154,381 patients). Privately insured, uninsured, and Medicaid patients’ data were reviewed for in-hospital mortality, length of stay (LOS), and cost per hospitalization. The analysis took into account disease severity, comorbidities, and the proportion of underinsured patients receiving care in each hospital when insurance-related disparities were examined.

Compared with the privately insured, in-hospital mortality and LOS for AMI and stroke were significantly higher for uninsured and Medicaid patients. Among pneumonia patients, Medicaid patients had significantly higher in-hospital mortality and LOS than the other two groups. Cost per hospitalization was highest for all three conditions in the Medicaid group; the uninsured group had the lowest costs for all three conditions.

Unfortunately, the three conditions analyzed only comprise 8% of annual hospital discharges, so the findings cannot be generalized. Also, deaths occurring soon after hospital discharge were not included, and uninsured and Medicaid patients are likely to have more severe diseases, which, rather than insurance status, could account for the mortality differences.

Bottom line: In-hospital mortality and resource use for three common medical conditions vary significantly between privately insured and uninsured or Medicaid patients, highlighting the need to take measures to close this gap.

Citation: Hasan O, Orav EJ, LeRoi LS. Insurance status and hospital care for myocardial infarction, stroke, and pneumonia. J Hosp Med. 2010;5(8);452-459.

Warfarin Monotherapy Best in Prevention of Thromboembolic Events for Atrial Fibrillation Patients

Clinical question: Is there a benefit to adding an antiplatelet agent to warfarin for the prevention of thromboembolic stroke in atrial fibrillation?

Background: Many physicians prescribe various combinations of aspirin, clopidogrel, and warfarin, as these treatments are endorsed in guidelines and expert statements. The use of these medications, however, has not been studied in a setting large enough to understand the safety of these therapies.

Study design: Retrospective cohort.

Setting: All Danish hospitals.

Synopsis: All hospitalized patients in Denmark from 1997 to 2006 who were identified with new onset atrial fibrillation (n=118,606) were monitored for outcomes and the use of aspirin, clopidogrel, and warfarin. These patients were followed for a mean of 3.3 years with the primary endpoint being admission to a hospital for a diagnosis of bleeding and a secondary endpoint of stroke.

Bleeding occurred in 13,573 patients (11.4%). The incidence of bleeding was highest in the first 180 days and then leveled off. Hazard ratios were computed with warfarin monotherapy as a reference. Only the hazard ratio for aspirin monotherapy (0.93) was lower (confidence interval [CI] 0.88-0.98). The highest risk of bleeding was with the triple therapy warfarin-aspirin-clopidogrel, which had a hazard ratio of 3.70 (CI 2.89-4.76).

For strokes, the hazard ratio was slightly better for warfarin-clopidogrel (0.70), although the CI was wide at 0.35-1.4 compared to warfarin monotherapy as a reference. Hazard ratios for monotherapy with clopidogrel or aspirin, dual therapy, and triple therapy all were worse, ranging from 1.27 to 1.86.

Bottom line: Warfarin as a monotherapy might have a bleeding risk comparable to that of aspirin or clopidogrel alone, and prevents more strokes than various combinations of these medications.

Citation: Hansen ML, Sørensen R, Clausen MT, et al. Risk of bleeding with single, dual, or triple therapy with warfarin, aspirin, and clopidogrel in patients with atrial fibrillation. Arch Intern Med. 2010;170 (16):1433-1441.

Cognitive and Physical Function Declines in Elderly Severe Sepsis Survivors

Clinical question: Is there a change in cognitive and physical functioning after severe sepsis?

Background: Disability is associated with increased mortality, decreased quality of life, and increased burdens by families and healthcare costs. After severe sepsis, the lasting effects of debility have not been investigated in any large studies.

Study design: Prospective cohort.

Setting: Hospitalized Medicare patients participating in the Health and Retirement Study.

Synopsis: Patients (n=1194) were followed for a minimum of one year between 1998 and 2006. The outcomes were measured by multiple personal interviews before and after a severe sepsis episode. Cognitive impairment was measured using three validated questionnaires dependent upon age or if a proxy was the respondent. For functional limitations, a questionnaire concerning instrumental and basic activities of daily living was used.

Cognitive impairment for those with moderate to severe impairment increased to 16.7% from 6.1% after a sepsis episode with an odds ratio of 3.34 (95% confidence interval 1.53-7.25). There was no significant increase in nonsevere sepsis hospitalized comparison patients (n=5574). All survivors with severe sepsis had a functional decline of 1.5 activities. The comparison group had about a 0.4 activity decline. All of these deficits endured throughout the study.

The authors provide comments that there should be a system-based approach in preventing severe sepsis, its burdens, and its costs. Suggestions include preventing delirium, initiating better standards of care, and involving therapists earlier to prevent immobility.

Bottom line: Severe sepsis is independently associated with enduring cognitive and physical functional declines, which strain families and our healthcare system.

Citation: Iwashyna TJ, Ely EW, Smith DM, Langa KM. Long-term cognitive impairment and functional disability among survivors of severe sepsis. JAMA. 2010;304(16):1787-1794.

Furosemide on Day of Elective Noncardiac Surgery Does Not Increase Risk of Intraoperative Hypotension

Clinical question: For patients chronically treated with loop diuretics, does withholding furosemide on the day of elective noncardiac surgery prevent intraoperative hypotension?

Background: Recent studies have questioned the safety of blood-pressure-lowering medications administered on the day of surgery. Beta-blockers have been associated with an increase in strokes and death perioperatively, and angiotensin receptor blockers (ARB) are frequently withheld on the day of surgery to avoid intraoperative hypotension. The effect of loop diuretics is uncertain.

Study design: Double-blind, randomized, placebo-controlled study.

Setting: Three North American university centers.

Synopsis: One hundred ninety-three patients were instructed to take furosemide or placebo on the day they underwent noncardiac surgery. The primary outcome measure was perioperative hypotension defined as a SBP <90 mmHg for more than five minutes, a 35% drop in the mean arterial blood pressure, or the need for a vasopressor agent. The number of cardiovascular complications (acute heart failure, acute coronary syndrome, arrhythmia, acute cerebrovascular event) and deaths also were analyzed.

Concerns have been raised that loop diuretics might predispose patients to a higher risk of intraoperative hypotension during noncardiac surgery. This trial showed no significant difference in the rates of intraoperative hypotension in patients who were administered furosemide versus those who were not. Although cardiovascular complications occurred more frequently in the furosemide group, the difference was not statistically significant.

Important limitations of the study were recognized. A larger population of patients could have revealed a statistically significant difference in cardiovascular outcomes in the furosemide group. Also, an anesthetic protocol was not utilized, which raises questions about the interaction of furosemide and effect on blood pressure with certain anesthetics.

Bottom line: Administering furosemide prior to surgery in chronic users does not appreciably increase the rate of intraoperative hypotension or cardiovascular events.

Citation: Khan NA, Campbell NR, Frost SD, et al. Risk of intraoperative hypotension with loop diuretics: a randomized controlled trial. Am J Med. 2010;123(11):1059e1-1059e8.

Pediatric HM Literature

Computerized Physician Order Entry Decreases Hospitalwide Mortality

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 effect of computerized physician order entry (CPOE) on hospitalwide mortality?

Background: CPOE has been touted as a key method for improving the quality and safety of patient care. To date, however, studies have not demonstrated definitive benefits, with one study demonstrating an increase in mortality rates in a critical-care unit after CPOE implementation.

Study design: Retrospective cohort study.

Setting: Quaternary-care academic children’s hospital.

Synopsis: More than 97,000 nonobstetric patients admitted to Lucile Packard Children’s Hospital (LCPH) at Stanford University in Palo Alto, Calif., from January 2001 to April 2009 were included in the analysis. The patients were divided into pre- and postintervention groups. The intervention was defined as a “big bang” activation of CPOE on Nov. 4, 2007, affecting 90% of the inpatient beds at LCPH.

After the intervention, mean monthly unadjusted mortality rates decreased to 0.716 deaths from 1.008 per 100 discharges. After adjustment for severity of illness and a rapid-response team (which had previously been shown to decrease mortality rates), mortality rates postintervention continued to remain significantly lower.

This study revealed an impressive reduction in perhaps the ultimate outcome measure: hospitalwide mortality. The authors went to extensive lengths to adjust for potential confounding factors to include delayed implementation of CPOE in the ICUs. As with many complex process improvement efforts, the devil might be in the details of local context and culture. Thus, the software might have been less important than the manner in which the institution adapted to its functionality. Those seeking to replicate these results will need to be mindful of the requisite organizational changes in communication, order standardization, and data retrieval.

Bottom line: CPOE implementation has the potential to significantly reduce hospitalwide mortality rates.

Citation: Longhurst CA, Parast L, Sandborg CI, et al. Decrease in hospital-wide mortality rate after implementation of a commercially sold computerized physician order entry system. Pediatrics. 2010;126(1):e1-e8.

In This Edition

Literature at a Glance

A guide to this month’s studies

• Risk of adverse events with opioid use
• Drug of choice for outpatient treatment of cellulitis
• Preventing hospital falls
• Post-hospital outcomes based on status of PCP follow-up
• LOS, mortality, and readmission based on insurance
• Antiplatelets added to warfarin for atrial fibrillation.
• Cognitive effects of severe sepsis
• Effect of preoperative furosemide use

ED Visits Are Higher among Recipients of Chronic Opioid Therapy

Clinical question: Is there an association between the use of prescription opioids and adverse outcomes?

Background: Chronic opioid therapy is a common strategy for managing chronic, noncancer pain. There has been an increase in overdose deaths and ED visits (EDV) involving the use of prescription opioids.

Study design: Retrospective study from claims records.

Setting: Population in the Health Core Integrated Research Database, containing large, commercial insurance plans in 14 states, and Arkansas Medicaid.

Synopsis: Patients 18 and older without cancer diagnoses who used prescription opioids for at least 90 continuous days within a six-month period from 2000 to 2005 were examined for risk factors for EDVs and alcohol- or drug-related encounters (ADEs) in the 12 months following 90 days or more of prescribed opioids.

Patients with diagnoses of headache, back pain, and pre-existing substance-use disorders had significantly higher EDVs and ADEs. Opioid dose at morphine-equivalent doses over 120 mg per day doubled the risk of ADEs. The use of short-acting Schedule II opioids was associated with EDVs (relative risk, 1.09-1.74). The use of long-acting Schedule II opioids was strongly associated with ADEs (relative risk, 1.64-4.00).

Bottom line: In adults with noncancer pain prescribed opioids for 90 days or more, short-acting Schedule II opioid use was associated with an increased number of EDVs, and long-acting opioid use was associated with an increased number of ADEs. Minimizing Schedule II opioid prescription in these higher-risk patients might be prudent to increase patient safety.

Citation: Braden JB, Russo J, Fan MI, et al. Emergency department visits among recipients of chronic opioid therapy. Arch Intern Med. 2010; 170(16):1425-1432.

Empiric Outpatient Therapy with Trimethoprim-Sulfamethoxazole or Clindamycin Is Preferred for Cellulitis

Clinical question: What is the best empiric outpatient oral antibiotic treatment for cellulitis in areas with a high prevalence of community-associated MRSA infections?

Background: The increasing rates of community-associated MRSA skin and soft-tissue infections have raised concerns that such beta-lactams as cephalexin and other semisynthetic penicillins are not appropriate for empiric outpatient therapy for cellulitis.

Study design: Three-year, retrospective cohort study.

Setting: A teaching clinic of a tertiary-care medical center in Hawaii.

Synopsis: More than 540 patients with cellulitis were identified from January 2005 to December 2007. Of these, 139 patients were excluded for reasons such as hospitalization, surgical intervention, etc. In the final cohort of 405 patients, the three most commonly prescribed oral antibiotics were cephalexin (44%), trimethoprim-sulfamethoxazole (38%), and clindamycin (10%). Other antibiotics accounted for the remaining 8%.

MRSA was recovered in 62% of positive culture specimens. The success rate of trimethoprim-sulfamethoxazole was 91% vs. 74% in the cephalexin group (P<0.001). Clindamycin success rates were higher than those of cephalexin in patients who had subsequently confirmed MRSA infections (P=0.01) and moderately severe cellulitis (P=0.03) and were obese (P=0.04).

Bottom line: Antibiotics with activity against community-acquired MRSA (e.g. trimethroprim-sulfamethoxazole and clinidamycin) are the preferred empiric outpatient therapy for cellulitis in areas with a high prevalence of community-acquired MRSA.

Citation: Khawcharoenporn T, Tice A. Empiric outpatient therapy with trimethoprim-sulfamethoxazole, cephalexin, or clindamycin for cellulitis. Am J Med. 2010;123(10):942-950.

Patient-Specific Screening with Health Information Technology Prevents Falls

Clinical question: Does a fall-prevention toolkit using health information technology decrease patient falls in acute-care hospitals?

Background: Inpatient falls and fall-related injuries result in substantial morbidity and additional healthcare costs. While specific fall-prevention strategies were a longstanding target for intervention, little evidence exists to link them with decreased fall rates.

Study design: Cluster-randomized study.

Setting: Four urban hospitals in Massachusetts.

Synopsis: Comparing patient fall rates in four acute-care hospitals between units providing usual care (5,104 patients) and units using a health information technology (HIT)-linked fall prevention toolkit (5,160 patients), this study demonstrated significant fall reduction in older inpatients. The intervention integrated existing workflow and validated fall risk assessment (Morse Falls Scale) into an HIT software application that tailored fall-prevention interventions to patients’ specific fall risk determinants. The toolkit produced bed posters, patient education handouts, and plans of care communicating patient-specific alerts to key stakeholders.

The primary outcome was patient falls per 1,000 patient-days during the six-month intervention period. The number of patients with falls was significantly different (P=0.02) between control (n=87) and intervention (n=67) units. The toolkit prevented one fall per 862 patient-days.

This nonblinded study was limited by the fact that it was conducted in a single health system. The toolkit was not effective in patients less than 65 years of age. Additionally, the sample size did not have sufficient power to detect effectiveness in preventing repeat falls or falls with injury.

Bottom line: Patient-specific fall prevention strategy coupled with HIT reduces falls in older inpatients.

Citation: Dykes PC, Carroll DL, Hurley A, et al. Fall prevention in acute care hospitals: a randomized trial. JAMA. 2010;304(17):1912-1918.

Lack of Timely Outpatient Follow-Up Results in Higher Readmission Rates

Clinical question: Does timely primary-care-physician (PCP) follow-up improve outcomes and prevent hospital readmissions?

Background: Active PCP involvement is key to interventions aimed at reducing readmissions and ensuring effective ongoing patient care. Some studies suggest increased overall resource utilization when PCP follow-up occurs after hospitalization. Resource utilization and clinical outcomes after hospitalization related to timely PCP follow-up have not been adequately studied.

Study design: Prospective cohort.

Setting: An urban, academic, 425-bed tertiary-care center in Colorado.

Synopsis: From a convenience sample of 121 patients admitted to general medicine services during winter months, 65 patients completed the study. Demographics, diagnosis, payor source, and PCP information were collected upon enrollment. Post-discharge phone calls and patient surveys were used to determine follow-up and readmission status. Timely PCP follow-up was defined as a visit with a PCP or specialist related to the discharge diagnosis within four weeks of hospital discharge.

Thirty-day readmission rates and hospital length of stay were compared for those with timely PCP follow-up and those without. Less than half of general-medicine inpatients received timely PCP follow-up post-discharge. Lack of timely PCP follow-up was associated with younger age, a 10-fold increase in 30-day readmission for the same condition, and a trend toward longer length of stay. However, hospital readmission for any condition did not differ with lack of timely PCP follow-up.

This small, single-center study with convenience sample enrollment might not represent all medical inpatients or diagnoses. Determination of same-condition readmission was potentially subjective.

Bottom line: Patients who lack timely post-discharge follow-up have higher readmission rates for the same medical condition.

Citation: Misky GJ, Wald HL, Coleman EA. Post-hospitalization transitions: examining the effects of timing of primary care provider follow-up. J Hosp Med. 2010;5(7):392-397.

Compared with Uninsured and Medicaid Patients, Privately Insured Patients Admitted for Acute MI, Stroke, and Pneumonia Have Better Mortality Rates

Clinical question: Do outcomes for insured and underinsured patients vary for three of the most common medical conditions for which patients are hospitalized: acute myocardial infarction (AMI), stroke, and pneumonia?

Background: The ideal healthcare system would provide quality care to all individuals regardless of insurance status. Nevertheless, disparities in outcomes for the insured and underinsured or uninsured are well-documented in the outpatient setting but not as well in the inpatient setting. More needs to be done to address these potential disparities.

Study design: Retrospective database analysis.

Setting: Database including 20% of all U.S. community hospitals, including public hospitals, academic medical centers, and specialty hospitals.

Synopsis: This study utilized a database of 8 million discharges from more than 1,000 hospitals and isolated patients 18-64 years old (154,381 patients). Privately insured, uninsured, and Medicaid patients’ data were reviewed for in-hospital mortality, length of stay (LOS), and cost per hospitalization. The analysis took into account disease severity, comorbidities, and the proportion of underinsured patients receiving care in each hospital when insurance-related disparities were examined.

Compared with the privately insured, in-hospital mortality and LOS for AMI and stroke were significantly higher for uninsured and Medicaid patients. Among pneumonia patients, Medicaid patients had significantly higher in-hospital mortality and LOS than the other two groups. Cost per hospitalization was highest for all three conditions in the Medicaid group; the uninsured group had the lowest costs for all three conditions.

Unfortunately, the three conditions analyzed only comprise 8% of annual hospital discharges, so the findings cannot be generalized. Also, deaths occurring soon after hospital discharge were not included, and uninsured and Medicaid patients are likely to have more severe diseases, which, rather than insurance status, could account for the mortality differences.

Bottom line: In-hospital mortality and resource use for three common medical conditions vary significantly between privately insured and uninsured or Medicaid patients, highlighting the need to take measures to close this gap.

Citation: Hasan O, Orav EJ, LeRoi LS. Insurance status and hospital care for myocardial infarction, stroke, and pneumonia. J Hosp Med. 2010;5(8);452-459.

Warfarin Monotherapy Best in Prevention of Thromboembolic Events for Atrial Fibrillation Patients

Clinical question: Is there a benefit to adding an antiplatelet agent to warfarin for the prevention of thromboembolic stroke in atrial fibrillation?

Background: Many physicians prescribe various combinations of aspirin, clopidogrel, and warfarin, as these treatments are endorsed in guidelines and expert statements. The use of these medications, however, has not been studied in a setting large enough to understand the safety of these therapies.

Study design: Retrospective cohort.

Setting: All Danish hospitals.

Synopsis: All hospitalized patients in Denmark from 1997 to 2006 who were identified with new onset atrial fibrillation (n=118,606) were monitored for outcomes and the use of aspirin, clopidogrel, and warfarin. These patients were followed for a mean of 3.3 years with the primary endpoint being admission to a hospital for a diagnosis of bleeding and a secondary endpoint of stroke.

Bleeding occurred in 13,573 patients (11.4%). The incidence of bleeding was highest in the first 180 days and then leveled off. Hazard ratios were computed with warfarin monotherapy as a reference. Only the hazard ratio for aspirin monotherapy (0.93) was lower (confidence interval [CI] 0.88-0.98). The highest risk of bleeding was with the triple therapy warfarin-aspirin-clopidogrel, which had a hazard ratio of 3.70 (CI 2.89-4.76).

For strokes, the hazard ratio was slightly better for warfarin-clopidogrel (0.70), although the CI was wide at 0.35-1.4 compared to warfarin monotherapy as a reference. Hazard ratios for monotherapy with clopidogrel or aspirin, dual therapy, and triple therapy all were worse, ranging from 1.27 to 1.86.

Bottom line: Warfarin as a monotherapy might have a bleeding risk comparable to that of aspirin or clopidogrel alone, and prevents more strokes than various combinations of these medications.

Citation: Hansen ML, Sørensen R, Clausen MT, et al. Risk of bleeding with single, dual, or triple therapy with warfarin, aspirin, and clopidogrel in patients with atrial fibrillation. Arch Intern Med. 2010;170 (16):1433-1441.

Cognitive and Physical Function Declines in Elderly Severe Sepsis Survivors

Clinical question: Is there a change in cognitive and physical functioning after severe sepsis?

Background: Disability is associated with increased mortality, decreased quality of life, and increased burdens by families and healthcare costs. After severe sepsis, the lasting effects of debility have not been investigated in any large studies.

Study design: Prospective cohort.

Setting: Hospitalized Medicare patients participating in the Health and Retirement Study.

Synopsis: Patients (n=1194) were followed for a minimum of one year between 1998 and 2006. The outcomes were measured by multiple personal interviews before and after a severe sepsis episode. Cognitive impairment was measured using three validated questionnaires dependent upon age or if a proxy was the respondent. For functional limitations, a questionnaire concerning instrumental and basic activities of daily living was used.

Cognitive impairment for those with moderate to severe impairment increased to 16.7% from 6.1% after a sepsis episode with an odds ratio of 3.34 (95% confidence interval 1.53-7.25). There was no significant increase in nonsevere sepsis hospitalized comparison patients (n=5574). All survivors with severe sepsis had a functional decline of 1.5 activities. The comparison group had about a 0.4 activity decline. All of these deficits endured throughout the study.

The authors provide comments that there should be a system-based approach in preventing severe sepsis, its burdens, and its costs. Suggestions include preventing delirium, initiating better standards of care, and involving therapists earlier to prevent immobility.

Bottom line: Severe sepsis is independently associated with enduring cognitive and physical functional declines, which strain families and our healthcare system.

Citation: Iwashyna TJ, Ely EW, Smith DM, Langa KM. Long-term cognitive impairment and functional disability among survivors of severe sepsis. JAMA. 2010;304(16):1787-1794.

Furosemide on Day of Elective Noncardiac Surgery Does Not Increase Risk of Intraoperative Hypotension

Clinical question: For patients chronically treated with loop diuretics, does withholding furosemide on the day of elective noncardiac surgery prevent intraoperative hypotension?

Background: Recent studies have questioned the safety of blood-pressure-lowering medications administered on the day of surgery. Beta-blockers have been associated with an increase in strokes and death perioperatively, and angiotensin receptor blockers (ARB) are frequently withheld on the day of surgery to avoid intraoperative hypotension. The effect of loop diuretics is uncertain.

Study design: Double-blind, randomized, placebo-controlled study.

Setting: Three North American university centers.

Synopsis: One hundred ninety-three patients were instructed to take furosemide or placebo on the day they underwent noncardiac surgery. The primary outcome measure was perioperative hypotension defined as a SBP <90 mmHg for more than five minutes, a 35% drop in the mean arterial blood pressure, or the need for a vasopressor agent. The number of cardiovascular complications (acute heart failure, acute coronary syndrome, arrhythmia, acute cerebrovascular event) and deaths also were analyzed.

Concerns have been raised that loop diuretics might predispose patients to a higher risk of intraoperative hypotension during noncardiac surgery. This trial showed no significant difference in the rates of intraoperative hypotension in patients who were administered furosemide versus those who were not. Although cardiovascular complications occurred more frequently in the furosemide group, the difference was not statistically significant.

Important limitations of the study were recognized. A larger population of patients could have revealed a statistically significant difference in cardiovascular outcomes in the furosemide group. Also, an anesthetic protocol was not utilized, which raises questions about the interaction of furosemide and effect on blood pressure with certain anesthetics.

Bottom line: Administering furosemide prior to surgery in chronic users does not appreciably increase the rate of intraoperative hypotension or cardiovascular events.

Citation: Khan NA, Campbell NR, Frost SD, et al. Risk of intraoperative hypotension with loop diuretics: a randomized controlled trial. Am J Med. 2010;123(11):1059e1-1059e8.

Pediatric HM Literature

Computerized Physician Order Entry Decreases Hospitalwide Mortality

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 effect of computerized physician order entry (CPOE) on hospitalwide mortality?

Background: CPOE has been touted as a key method for improving the quality and safety of patient care. To date, however, studies have not demonstrated definitive benefits, with one study demonstrating an increase in mortality rates in a critical-care unit after CPOE implementation.

Study design: Retrospective cohort study.

Setting: Quaternary-care academic children’s hospital.

Synopsis: More than 97,000 nonobstetric patients admitted to Lucile Packard Children’s Hospital (LCPH) at Stanford University in Palo Alto, Calif., from January 2001 to April 2009 were included in the analysis. The patients were divided into pre- and postintervention groups. The intervention was defined as a “big bang” activation of CPOE on Nov. 4, 2007, affecting 90% of the inpatient beds at LCPH.

After the intervention, mean monthly unadjusted mortality rates decreased to 0.716 deaths from 1.008 per 100 discharges. After adjustment for severity of illness and a rapid-response team (which had previously been shown to decrease mortality rates), mortality rates postintervention continued to remain significantly lower.

This study revealed an impressive reduction in perhaps the ultimate outcome measure: hospitalwide mortality. The authors went to extensive lengths to adjust for potential confounding factors to include delayed implementation of CPOE in the ICUs. As with many complex process improvement efforts, the devil might be in the details of local context and culture. Thus, the software might have been less important than the manner in which the institution adapted to its functionality. Those seeking to replicate these results will need to be mindful of the requisite organizational changes in communication, order standardization, and data retrieval.

Bottom line: CPOE implementation has the potential to significantly reduce hospitalwide mortality rates.

Citation: Longhurst CA, Parast L, Sandborg CI, et al. Decrease in hospital-wide mortality rate after implementation of a commercially sold computerized physician order entry system. Pediatrics. 2010;126(1):e1-e8.

In This Edition

Literature at a Glance

A guide to this month’s studies

• Risk of adverse events with opioid use
• Drug of choice for outpatient treatment of cellulitis
• Preventing hospital falls
• Post-hospital outcomes based on status of PCP follow-up
• LOS, mortality, and readmission based on insurance
• Antiplatelets added to warfarin for atrial fibrillation.
• Cognitive effects of severe sepsis
• Effect of preoperative furosemide use

ED Visits Are Higher among Recipients of Chronic Opioid Therapy

Clinical question: Is there an association between the use of prescription opioids and adverse outcomes?

Background: Chronic opioid therapy is a common strategy for managing chronic, noncancer pain. There has been an increase in overdose deaths and ED visits (EDV) involving the use of prescription opioids.

Study design: Retrospective study from claims records.

Setting: Population in the Health Core Integrated Research Database, containing large, commercial insurance plans in 14 states, and Arkansas Medicaid.

Synopsis: Patients 18 and older without cancer diagnoses who used prescription opioids for at least 90 continuous days within a six-month period from 2000 to 2005 were examined for risk factors for EDVs and alcohol- or drug-related encounters (ADEs) in the 12 months following 90 days or more of prescribed opioids.

Patients with diagnoses of headache, back pain, and pre-existing substance-use disorders had significantly higher EDVs and ADEs. Opioid dose at morphine-equivalent doses over 120 mg per day doubled the risk of ADEs. The use of short-acting Schedule II opioids was associated with EDVs (relative risk, 1.09-1.74). The use of long-acting Schedule II opioids was strongly associated with ADEs (relative risk, 1.64-4.00).

Bottom line: In adults with noncancer pain prescribed opioids for 90 days or more, short-acting Schedule II opioid use was associated with an increased number of EDVs, and long-acting opioid use was associated with an increased number of ADEs. Minimizing Schedule II opioid prescription in these higher-risk patients might be prudent to increase patient safety.

Citation: Braden JB, Russo J, Fan MI, et al. Emergency department visits among recipients of chronic opioid therapy. Arch Intern Med. 2010; 170(16):1425-1432.

Empiric Outpatient Therapy with Trimethoprim-Sulfamethoxazole or Clindamycin Is Preferred for Cellulitis

Clinical question: What is the best empiric outpatient oral antibiotic treatment for cellulitis in areas with a high prevalence of community-associated MRSA infections?

Background: The increasing rates of community-associated MRSA skin and soft-tissue infections have raised concerns that such beta-lactams as cephalexin and other semisynthetic penicillins are not appropriate for empiric outpatient therapy for cellulitis.

Study design: Three-year, retrospective cohort study.

Setting: A teaching clinic of a tertiary-care medical center in Hawaii.

Synopsis: More than 540 patients with cellulitis were identified from January 2005 to December 2007. Of these, 139 patients were excluded for reasons such as hospitalization, surgical intervention, etc. In the final cohort of 405 patients, the three most commonly prescribed oral antibiotics were cephalexin (44%), trimethoprim-sulfamethoxazole (38%), and clindamycin (10%). Other antibiotics accounted for the remaining 8%.

MRSA was recovered in 62% of positive culture specimens. The success rate of trimethoprim-sulfamethoxazole was 91% vs. 74% in the cephalexin group (P<0.001). Clindamycin success rates were higher than those of cephalexin in patients who had subsequently confirmed MRSA infections (P=0.01) and moderately severe cellulitis (P=0.03) and were obese (P=0.04).

Bottom line: Antibiotics with activity against community-acquired MRSA (e.g. trimethroprim-sulfamethoxazole and clinidamycin) are the preferred empiric outpatient therapy for cellulitis in areas with a high prevalence of community-acquired MRSA.

Citation: Khawcharoenporn T, Tice A. Empiric outpatient therapy with trimethoprim-sulfamethoxazole, cephalexin, or clindamycin for cellulitis. Am J Med. 2010;123(10):942-950.

Patient-Specific Screening with Health Information Technology Prevents Falls

Clinical question: Does a fall-prevention toolkit using health information technology decrease patient falls in acute-care hospitals?

Background: Inpatient falls and fall-related injuries result in substantial morbidity and additional healthcare costs. While specific fall-prevention strategies were a longstanding target for intervention, little evidence exists to link them with decreased fall rates.

Study design: Cluster-randomized study.

Setting: Four urban hospitals in Massachusetts.

Synopsis: Comparing patient fall rates in four acute-care hospitals between units providing usual care (5,104 patients) and units using a health information technology (HIT)-linked fall prevention toolkit (5,160 patients), this study demonstrated significant fall reduction in older inpatients. The intervention integrated existing workflow and validated fall risk assessment (Morse Falls Scale) into an HIT software application that tailored fall-prevention interventions to patients’ specific fall risk determinants. The toolkit produced bed posters, patient education handouts, and plans of care communicating patient-specific alerts to key stakeholders.

The primary outcome was patient falls per 1,000 patient-days during the six-month intervention period. The number of patients with falls was significantly different (P=0.02) between control (n=87) and intervention (n=67) units. The toolkit prevented one fall per 862 patient-days.

This nonblinded study was limited by the fact that it was conducted in a single health system. The toolkit was not effective in patients less than 65 years of age. Additionally, the sample size did not have sufficient power to detect effectiveness in preventing repeat falls or falls with injury.

Bottom line: Patient-specific fall prevention strategy coupled with HIT reduces falls in older inpatients.

Citation: Dykes PC, Carroll DL, Hurley A, et al. Fall prevention in acute care hospitals: a randomized trial. JAMA. 2010;304(17):1912-1918.

Lack of Timely Outpatient Follow-Up Results in Higher Readmission Rates

Clinical question: Does timely primary-care-physician (PCP) follow-up improve outcomes and prevent hospital readmissions?

Background: Active PCP involvement is key to interventions aimed at reducing readmissions and ensuring effective ongoing patient care. Some studies suggest increased overall resource utilization when PCP follow-up occurs after hospitalization. Resource utilization and clinical outcomes after hospitalization related to timely PCP follow-up have not been adequately studied.

Study design: Prospective cohort.

Setting: An urban, academic, 425-bed tertiary-care center in Colorado.

Synopsis: From a convenience sample of 121 patients admitted to general medicine services during winter months, 65 patients completed the study. Demographics, diagnosis, payor source, and PCP information were collected upon enrollment. Post-discharge phone calls and patient surveys were used to determine follow-up and readmission status. Timely PCP follow-up was defined as a visit with a PCP or specialist related to the discharge diagnosis within four weeks of hospital discharge.

Thirty-day readmission rates and hospital length of stay were compared for those with timely PCP follow-up and those without. Less than half of general-medicine inpatients received timely PCP follow-up post-discharge. Lack of timely PCP follow-up was associated with younger age, a 10-fold increase in 30-day readmission for the same condition, and a trend toward longer length of stay. However, hospital readmission for any condition did not differ with lack of timely PCP follow-up.

This small, single-center study with convenience sample enrollment might not represent all medical inpatients or diagnoses. Determination of same-condition readmission was potentially subjective.

Bottom line: Patients who lack timely post-discharge follow-up have higher readmission rates for the same medical condition.

Citation: Misky GJ, Wald HL, Coleman EA. Post-hospitalization transitions: examining the effects of timing of primary care provider follow-up. J Hosp Med. 2010;5(7):392-397.

Compared with Uninsured and Medicaid Patients, Privately Insured Patients Admitted for Acute MI, Stroke, and Pneumonia Have Better Mortality Rates

Clinical question: Do outcomes for insured and underinsured patients vary for three of the most common medical conditions for which patients are hospitalized: acute myocardial infarction (AMI), stroke, and pneumonia?

Background: The ideal healthcare system would provide quality care to all individuals regardless of insurance status. Nevertheless, disparities in outcomes for the insured and underinsured or uninsured are well-documented in the outpatient setting but not as well in the inpatient setting. More needs to be done to address these potential disparities.

Study design: Retrospective database analysis.

Setting: Database including 20% of all U.S. community hospitals, including public hospitals, academic medical centers, and specialty hospitals.

Synopsis: This study utilized a database of 8 million discharges from more than 1,000 hospitals and isolated patients 18-64 years old (154,381 patients). Privately insured, uninsured, and Medicaid patients’ data were reviewed for in-hospital mortality, length of stay (LOS), and cost per hospitalization. The analysis took into account disease severity, comorbidities, and the proportion of underinsured patients receiving care in each hospital when insurance-related disparities were examined.

Compared with the privately insured, in-hospital mortality and LOS for AMI and stroke were significantly higher for uninsured and Medicaid patients. Among pneumonia patients, Medicaid patients had significantly higher in-hospital mortality and LOS than the other two groups. Cost per hospitalization was highest for all three conditions in the Medicaid group; the uninsured group had the lowest costs for all three conditions.

Unfortunately, the three conditions analyzed only comprise 8% of annual hospital discharges, so the findings cannot be generalized. Also, deaths occurring soon after hospital discharge were not included, and uninsured and Medicaid patients are likely to have more severe diseases, which, rather than insurance status, could account for the mortality differences.

Bottom line: In-hospital mortality and resource use for three common medical conditions vary significantly between privately insured and uninsured or Medicaid patients, highlighting the need to take measures to close this gap.

Citation: Hasan O, Orav EJ, LeRoi LS. Insurance status and hospital care for myocardial infarction, stroke, and pneumonia. J Hosp Med. 2010;5(8);452-459.

Warfarin Monotherapy Best in Prevention of Thromboembolic Events for Atrial Fibrillation Patients

Clinical question: Is there a benefit to adding an antiplatelet agent to warfarin for the prevention of thromboembolic stroke in atrial fibrillation?

Background: Many physicians prescribe various combinations of aspirin, clopidogrel, and warfarin, as these treatments are endorsed in guidelines and expert statements. The use of these medications, however, has not been studied in a setting large enough to understand the safety of these therapies.

Study design: Retrospective cohort.

Setting: All Danish hospitals.

Synopsis: All hospitalized patients in Denmark from 1997 to 2006 who were identified with new onset atrial fibrillation (n=118,606) were monitored for outcomes and the use of aspirin, clopidogrel, and warfarin. These patients were followed for a mean of 3.3 years with the primary endpoint being admission to a hospital for a diagnosis of bleeding and a secondary endpoint of stroke.

Bleeding occurred in 13,573 patients (11.4%). The incidence of bleeding was highest in the first 180 days and then leveled off. Hazard ratios were computed with warfarin monotherapy as a reference. Only the hazard ratio for aspirin monotherapy (0.93) was lower (confidence interval [CI] 0.88-0.98). The highest risk of bleeding was with the triple therapy warfarin-aspirin-clopidogrel, which had a hazard ratio of 3.70 (CI 2.89-4.76).

For strokes, the hazard ratio was slightly better for warfarin-clopidogrel (0.70), although the CI was wide at 0.35-1.4 compared to warfarin monotherapy as a reference. Hazard ratios for monotherapy with clopidogrel or aspirin, dual therapy, and triple therapy all were worse, ranging from 1.27 to 1.86.

Bottom line: Warfarin as a monotherapy might have a bleeding risk comparable to that of aspirin or clopidogrel alone, and prevents more strokes than various combinations of these medications.

Citation: Hansen ML, Sørensen R, Clausen MT, et al. Risk of bleeding with single, dual, or triple therapy with warfarin, aspirin, and clopidogrel in patients with atrial fibrillation. Arch Intern Med. 2010;170 (16):1433-1441.

Cognitive and Physical Function Declines in Elderly Severe Sepsis Survivors

Clinical question: Is there a change in cognitive and physical functioning after severe sepsis?

Background: Disability is associated with increased mortality, decreased quality of life, and increased burdens by families and healthcare costs. After severe sepsis, the lasting effects of debility have not been investigated in any large studies.

Study design: Prospective cohort.

Setting: Hospitalized Medicare patients participating in the Health and Retirement Study.

Synopsis: Patients (n=1194) were followed for a minimum of one year between 1998 and 2006. The outcomes were measured by multiple personal interviews before and after a severe sepsis episode. Cognitive impairment was measured using three validated questionnaires dependent upon age or if a proxy was the respondent. For functional limitations, a questionnaire concerning instrumental and basic activities of daily living was used.

Cognitive impairment for those with moderate to severe impairment increased to 16.7% from 6.1% after a sepsis episode with an odds ratio of 3.34 (95% confidence interval 1.53-7.25). There was no significant increase in nonsevere sepsis hospitalized comparison patients (n=5574). All survivors with severe sepsis had a functional decline of 1.5 activities. The comparison group had about a 0.4 activity decline. All of these deficits endured throughout the study.

The authors provide comments that there should be a system-based approach in preventing severe sepsis, its burdens, and its costs. Suggestions include preventing delirium, initiating better standards of care, and involving therapists earlier to prevent immobility.

Bottom line: Severe sepsis is independently associated with enduring cognitive and physical functional declines, which strain families and our healthcare system.

Citation: Iwashyna TJ, Ely EW, Smith DM, Langa KM. Long-term cognitive impairment and functional disability among survivors of severe sepsis. JAMA. 2010;304(16):1787-1794.

Furosemide on Day of Elective Noncardiac Surgery Does Not Increase Risk of Intraoperative Hypotension

Clinical question: For patients chronically treated with loop diuretics, does withholding furosemide on the day of elective noncardiac surgery prevent intraoperative hypotension?

Background: Recent studies have questioned the safety of blood-pressure-lowering medications administered on the day of surgery. Beta-blockers have been associated with an increase in strokes and death perioperatively, and angiotensin receptor blockers (ARB) are frequently withheld on the day of surgery to avoid intraoperative hypotension. The effect of loop diuretics is uncertain.

Study design: Double-blind, randomized, placebo-controlled study.

Setting: Three North American university centers.

Synopsis: One hundred ninety-three patients were instructed to take furosemide or placebo on the day they underwent noncardiac surgery. The primary outcome measure was perioperative hypotension defined as a SBP <90 mmHg for more than five minutes, a 35% drop in the mean arterial blood pressure, or the need for a vasopressor agent. The number of cardiovascular complications (acute heart failure, acute coronary syndrome, arrhythmia, acute cerebrovascular event) and deaths also were analyzed.

Concerns have been raised that loop diuretics might predispose patients to a higher risk of intraoperative hypotension during noncardiac surgery. This trial showed no significant difference in the rates of intraoperative hypotension in patients who were administered furosemide versus those who were not. Although cardiovascular complications occurred more frequently in the furosemide group, the difference was not statistically significant.

Important limitations of the study were recognized. A larger population of patients could have revealed a statistically significant difference in cardiovascular outcomes in the furosemide group. Also, an anesthetic protocol was not utilized, which raises questions about the interaction of furosemide and effect on blood pressure with certain anesthetics.

Bottom line: Administering furosemide prior to surgery in chronic users does not appreciably increase the rate of intraoperative hypotension or cardiovascular events.

Citation: Khan NA, Campbell NR, Frost SD, et al. Risk of intraoperative hypotension with loop diuretics: a randomized controlled trial. Am J Med. 2010;123(11):1059e1-1059e8.

Pediatric HM Literature

Computerized Physician Order Entry Decreases Hospitalwide Mortality

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 effect of computerized physician order entry (CPOE) on hospitalwide mortality?

Background: CPOE has been touted as a key method for improving the quality and safety of patient care. To date, however, studies have not demonstrated definitive benefits, with one study demonstrating an increase in mortality rates in a critical-care unit after CPOE implementation.

Study design: Retrospective cohort study.

Setting: Quaternary-care academic children’s hospital.

Synopsis: More than 97,000 nonobstetric patients admitted to Lucile Packard Children’s Hospital (LCPH) at Stanford University in Palo Alto, Calif., from January 2001 to April 2009 were included in the analysis. The patients were divided into pre- and postintervention groups. The intervention was defined as a “big bang” activation of CPOE on Nov. 4, 2007, affecting 90% of the inpatient beds at LCPH.

After the intervention, mean monthly unadjusted mortality rates decreased to 0.716 deaths from 1.008 per 100 discharges. After adjustment for severity of illness and a rapid-response team (which had previously been shown to decrease mortality rates), mortality rates postintervention continued to remain significantly lower.

This study revealed an impressive reduction in perhaps the ultimate outcome measure: hospitalwide mortality. The authors went to extensive lengths to adjust for potential confounding factors to include delayed implementation of CPOE in the ICUs. As with many complex process improvement efforts, the devil might be in the details of local context and culture. Thus, the software might have been less important than the manner in which the institution adapted to its functionality. Those seeking to replicate these results will need to be mindful of the requisite organizational changes in communication, order standardization, and data retrieval.

Bottom line: CPOE implementation has the potential to significantly reduce hospitalwide mortality rates.

Citation: Longhurst CA, Parast L, Sandborg CI, et al. Decrease in hospital-wide mortality rate after implementation of a commercially sold computerized physician order entry system. Pediatrics. 2010;126(1):e1-e8.