Year in Review

EBM Focus - Volume 8, Issue 52

Read the Complete EBM Focus and Earn CME

Correction to DynaMed EBM Focus 52: Year in Review 2013 December 27, 2013

The DynaMed EBM Focus 52: Year in Review 2013 included a conclusion of “Low Sodium Diets Increase Mortality and Readmission Rate in Patients with Systolic Heart Failure Taking Daily Diuretics” based on a systematic review reported in DynaMed EBM Focus in January 2013. The systematic review was retracted in March 2013, and the retraction notice replaced the content in DynaMed in March 2013.

It remains uncertain whether sodium restriction is helpful or harmful in most patients with heart failure, but sodium restriction appears harmful in patients treated with very high doses of furosemide. The original trials are summarized in the Heart failure topic. The diuretic doses used (furosemide dosing ranging from 250 mg/day to 1,000 mg/day) were so much higher than routine clinical practice that we downgraded the conclusion to “level 3 [lacking direct] evidence.”

Brian S. Alper, MD, MSPH, FAAFP
Founder of DynaMed
Vice President of EBM Research and Development, Quality & Standards

Year in Review

In 2013, the DynaMed Editors considered 26,463 articles and 11,499 articles were included in DynaMed. Each week, from an average of 222 articles added to DynaMed, editors selected articles that were considered “most likely to change clinical practice.”

In this “Year in Review” issue, the DynaMed Editors share a recap of the 5 most important articles from 2013. We welcome your comments on how your clinical practice was affected by the information presented. Comments can be sent to

We would like to thank our readers for their support and input throughout the year. Your feedback helps improve the DynaMed EBM Focus for the global DynaMed Community.

With our very best wishes for a joyous and peaceful New Year,
The DynaMed Editors.

Donor Feces Infusion with Short Antibiotic Course May Cure Recurrent C. difficile Diarrhea Reference: N Engl J Med 2013 Jan 31;368(5):407 (level 2 [mid-level] evidence)

About a quarter of patients treated for Clostridium difficile infections develop recurrent symptoms after a standard course of antibiotics. Recurrent C. difficile infections can be particularly resistant to further antibiotic treatment, and with each recurrence the risk of another recurrence increases. Probiotics have been tried for treating recurrent C. difficile, but evidence for efficacy is limited and inconsistent (Am J Gastroenterol 2006 Apr;101(4):812, Clin Infect Dis 2000 Oct;31(4):1012). However “stool transplants” of donor feces, an alternative approach to restoring the diversity of intestinal flora, have been reported to result in high rates of resolution in case series (Arch Intern Med 2012 Jan 23;172(2):191, Clin Infect Dis 2003 Mar 1;36(5):580). Evidence from a randomized trial for this treatment had been lacking until now.
In a recent small trial, 43 patients (mean age 68 years) with recurrent C. difficile diarrhea after at least 1 course of antibiotics were randomized to 1 of 3 treatments. The first group had a 4-5 day course of vancomycin followed by bowel lavage and then donor feces infusion through nasoduodenal tube. Feces were diluted in 500 ml of 0.9% saline, stirred, and the supernatant was then administered within 6 hours following collection from donor. All donors were < 60 years old and screened for risk factors associated with potentially transmissible diseases. An additional infusion with feces from a different donor was given if infection recurred. The second group received a 14-day course of vancomycin plus lavage, and the third received vancomycin for 14 days only. The primary outcome was cure of C. difficile infection within 10 weeks, defined as absence of diarrhea (or persistent diarrhea explained by other cause) plus 3 consecutive stool tests negative for C. difficile. The cure rates were 94% with feces infusion, 23% with vancomycin plus lavage, and 31% with vancomycin only (p < 0.001 for feces infusion vs. each group, NNT 2). Symptoms were resolved in all but 3 patients (81%) in the infusion group after the first treatment. Two additional patients were cured after a second infusion. Diarrhea was common on the day of infusion in that group. The planned enrollment was 120 patients, but the trial was terminated early after an interim analysis showed a significant difference in efficacy (N Engl J Med 2013 Jan 16 early online). For more information, see the Clostridium difficile infection topic in DynaMed.

Low Sodium Diets Increase Mortality and Readmission Rate in Patients with Systolic Heart Failure Taking Daily Diuretics Reference: Am J Med Sci 2011 Jul;342(1):27 (level 1 [likely reliable] evidence)

Guidelines from the Heart Failure Society of America (HFSA) recommend daily sodium intake of 2-3 g in most patients with heart failure, and suggest limiting sodium intake to < 2 g daily in cases of moderate-severe heart failure (J Card Fail 2010 Jun;16(6):e1). These recommendations have been based largely on observational data. A Cochrane review in 2011 identified a single randomized trial evaluating low sodium diets in patients with heart failure taking high-dose furosemide, and found an association between low sodium and increased mortality (Cochrane Database Syst Rev 2011 Jul 6;(7):CD009217). A new systematic review compared low sodium vs. normal sodium diets with data from 6 randomized trials including 2,747 patients with heart failure with a broader mix of diuretic treatment dosages. In all trials, patients were randomized to a low-sodium diet of 1.8 grams daily vs. a normal sodium diet of 2.8 grams daily. Most trials had a fluid restriction for all patients of 1 liter daily. Furosemide doses (IV or oral) ranged from a low of 50-125 mg twice daily in 1 trial with 1,771 patients to a high of 500-1000 mg twice daily in 2 trials. Compared to normal sodium diets, low sodium diets significantly increased all-cause mortality (risk ratio [RR] 1.95, 95% CI 1.66-2.29), heart failure-related mortality (RR 2.23, 95% CI 1.77-2.81) and heart failure readmissions (RR 2.1, 95% CI 1.67-2.64). The numbers needed to harm (NNH) were 6-12 for all-cause mortality and 2-7 for readmission. Low sodium diets were also associated with increased risk of sudden death (RR 1.72, 95% CI 1.21-2.44) (Heart 2013 Jan 24 early online). In the largest trial, with the most clinically applicable furosemide dosing, the low sodium diet was associated with increased mortality (23.8% vs. 12.9%, p < 0.0001, NNH 9) and increased heart failure readmissions (34.2% vs. 18.5%, p < 0.0001, NNH 6) in follow-up of 57 months. In this trial, the normal sodium group received hypertonic saline solution 150 mL with each furosemide infusion. For more information, see the Heart failure topic in DynaMed.

Antibiotics May Not Improve Outcomes in Uncomplicated Left-Sided Diverticulitis Reference: Cochrane Database Syst Rev 2012 Nov 14;(11):CD009092 (level 2 [mid-level] evidence)

Practice parameters from the American Society of Colon and Rectal Surgeons (ASCRS) recommend oral or IV broad-spectrum antibiotics for nonoperative management of acute left-sided diverticulitis (Dis Colon Rectum. 2006 Jul;49(7):939-44), and antibiotic treatment has become the standard of care for uncomplicated disease. However, little research has been performed to assess the efficacy of this approach. A recent Cochrane review identified only 1 unblinded trial that compared antibiotics to no antibiotics in patients with uncomplicated left-sided diverticulitis. In this trial, 623 patients (mean age 57 years) with acute uncomplicated left-sided diverticulitis verified by computed tomography were randomized to broad-spectrum antibiotics for ≥ 7 days vs. no antibiotics and followed for 1 year (Br J Surg 2012 Apr;99(4):532). Initial IV treatment was a combination of a second- or third-generation cephalosporin and metronidazole, or carbapenem antibiotics, or piperacillin-tazobactam. IV treatment was followed by oral antibiotics (ciprofloxacin or cefadroxil with metronidazole). All patients received IV fluids. There were no significant differences between the groups in any clinical outcomes. Recurrent diverticulitis with hospital readmission occurred in 15.8% with antibiotics vs. 16.2% without antibiotics. Complications (including sigmoid perforations or abscesses) occurred in 1% with antibiotics vs. 1.9% without, and sigmoid resections were performed in 1.6% vs. 2.3%. The median hospital stay was 3 days in each group. The 2 other trials included in the Cochrane review addressed the comparative efficacy of different antibiotic regimens, and found no significant differences in outcomes comparing short duration of IV treatment vs. > 7 days of IV treatment (1 trial with 44 patients) or comparing cefoxitin vs. gentamicin plus clindamycin (1 trial with 51 patients). For more information, see the Diverticulitis topic in DynaMed.

Apixaban Is as Effective for Treatment of Acute Venous Thromboembolism as Conventional Therapy and Reduces Bleeding Risk Reference: N Engl J Med 2013 Jul 1 early online (level 1 [likely reliable] evidence)

The standard treatment for venous thromboembolism (VTE) includes short-term treatment with a parenteral heparin-type anticoagulant (e.g. enoxaparin) and long-term treatment with warfarin, which requires frequent monitoring and dose adjustment. Newer anticoagulants, like fixed-dose oral factor Xa inhibitors, could simplify VTE management by reducing the need for injections and monitoring. The AMPLIFY trial compared the efficacy of the factor Xa inhibitor apixaban to conventional therapy for treatment of acute VTE in 5,395 adult patients. Patients (mean age 57 years) with acute symptomatic proximal deep vein thrombosis or pulmonary embolism were randomized to 1 of 2 interventions and followed for 7 months. The conventional therapy group received enoxaparin 1 mg/kg subcutaneously every 12 hours for ≥ 5 days plus warfarin with target international normalized ratio (INR) 2-3 for 6 months. The apixaban group received apixaban 10 mg orally twice daily for the first 7 days, then 5 mg twice daily for 6 months. Blinding was maintained through placebo subcutaneous injections and placebo warfarin with sham INR monitoring. The primary outcome was a composite of recurrent symptomatic VTE and VTE-related death. Primary outcome events occurred in 2.3% with apixaban vs. 2.7% with conventional therapy group (not significant). The apixaban group had lower rates of major bleeding (0.6% vs. 1.8%, p < 0.001, NNT 84) and clinically-relevant nonmajor bleeding (3.8% vs. 8%, p < 0.05, NNT 24). There were no significant differences in all-cause mortality (1.5% vs. 1.9%) or in serious adverse events. These data in patients with acute VTE add to recently published evidence regarding the efficacy and safety of long-term apixaban use for prevention of recurrent VTE (DynaMed EBM Focus Volume 8, Issue 11). Apixaban is currently FDA-approved only for use in patients with nonvalvular atrial fibrillation. For more information, see the Anticoagulant therapy for venous thromboembolism topic in DynaMed.

New ACC/AHA Guidelines for Statins/Calculator Overestimates Risk/DynaMed Provides Patient-Specific Benefit Estimation Reference: Circulation 2013 Nov 12 early online PDF, Lancet 2013 Nov 19 early online PDF (level 2 [mid-level] evidence)

Last week the American College of Cardiology/American Heart Association (ACC/AHA) published guidelines on cardiovascular risk assessment, and lifestyle changes and cholesterol-lowering treatment to reduce cardiovascular risk (J Am Coll Cardiol 2013 Nov 12 early online PDF, Circulation 2013 Nov 12 early online PDF). Many DynaMed topics have been updated to incorporate these new guidelines and replace the 2002 National Cholesterol Education Program (NCEP) guidelines. For the Statins for prevention of cardiovascular disease topic 3 major issues are of exceptional interest.

1. The ACC/AHA guidelines do not identify target cholesterol levels as the goal when treating dyslipidemia. This is now consistent with guidelines from other countries (such as 2008 NICE guidelines in the United Kingdom) which recognize benefit in treatment from overall risk reduction but not specific benefit from treating to target levels. The ACC/AHA guidelines (unlike the NICE guidelines) recommend monitoring cholesterol levels on treatment, especially for the purpose of monitoring and encouraging treatment adherence.

2. The threshold for when statin therapy is recommended are much lower in the ACC/AHA guidelines than in the NICE guidelines. The ACC/AHA guidelines recommend treatment with a 10-year risk for cardiovascular disease events > 7.5% and suggest treatment may be reasonable at a risk of 5%-7.5%, while NICE recommends statins therapy for adults with a 10-year risk > 20%. Both groups recommend statin therapy for most patients with diabetes or established cardiovascular disease.

3. Use of Pooled Cohort Equations is proposed for estimating the 10-year risk of cardiovascular disease events. The Pooled Cohort Equations appear to overestimate cardiovascular risk. This was found in comparison to actual event rates in the 2 cohorts used for independent external validation of these equations, and was also reported in 3 large primary prevention cohorts (Lancet 2013 Nov 19 early online PDF).

The DynaMed topic on Cardiovascular risk prediction includes multiple approaches for risk estimation, including links for their use and evidence supporting them. To put this into perspective for patients: To understand the benefits for an individual patient it is useful to estimate the absolute benefit of treatment for that patient. This can be understood as the number of patients who would need to be treated for 5 years to prevent one adverse outcome (NNT). We made estimates of the NNT for selected major adverse outcomes at different levels of predicted 10-year risk, using estimates for risk reductions derived from systematic reviews: NNT for Statins for 5 Years: 10-year risk of CVD events 5-year NNT for CVD events 5-year NNT for myocardial infarction 5-year NNT for stroke 5-year NNT for mortality 5% 160 278 910 * 7.5% 108 186 606 * 10% 80 140 456 * 15% 54 94 304 334 20% 40 70 228 250 Abbreviations: CVD, cardiovascular disease; NNT, number needed to treat (to prevent 1 outcome) * no apparent mortality reduction in lowest-risk patients (BMJ 2013 Oct 22;347:f6123) Courtesty: DynaMed The guidelines consider risks for adverse effects to be minimal. Randomized controlled data find low rates of serious adverse events, such as rate of myopathy 1 per 10,000 person-years, and a modest increase in diabetes (NNH 255 over 4 years [Lancet 2010 Feb 27;375(9716):735]) Observational studies have reported up to 17.4% rates of statin-related adverse events (Ann Intern Med 2013 Apr 2;158(7):526) For more information, see the Statins for prevention of cardiovascular disease topic in DynaMed.

Other EBSCO Sites +