Dimitri M. Drekonja, MD, MS; Mary Butler, PhD, MBA; Roderick MacDonald, MS; Donna Bliss, PhD, RN; Gregory A. Filice, MD; Thomas S. Rector, PhD; Timothy J. Wilt, MD, MPH
Note: A clinician's guide and consumer guide, produced by AHRQ, are available online (Supplements).
Disclaimer: The authors are responsible for the content of the report. Statements should not be construed as endorsements by AHRQ or the U.S. Department of Health and Human Services.
Acknowledgment: The authors thank Indulus Rutks and James Tacklind for their technical expertise in the preparation of this manuscript.
Grant Support: By contract HHSA 290 2007 10064 I from AHRQ, U.S. Department of Health and Human Services, and a Veterans Affairs Career Development Award through the Minneapolis Veterans Affairs Center for Epidemiological and Clinical Research (Dr. Drekonja) and Veterans Affairs Health Services Research and Development Service grant HFP-98-001 (Dr. Rector).
Potential Conflicts of Interest: Disclosures can be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M11-1356.
Requests for Single Reprints: Dimitri M. Drekonja, MD, MS, Infectious Disease Section (111F), Minneapolis Veterans Affairs Health Care System, One Veterans Drive, Minneapolis, MN 55417; e-mail, email@example.com.
Current Author Addresses: Drs. Drekonja and Filice: Infectious Disease Section (111F), Minneapolis Veterans Affairs Health Care System, One Veterans Drive, Minneapolis, MN 55417.
Dr. Butler: Division of Health Policy and Management, University of Minnesota, D330-4 Mayo Building, Minneapolis, MN 55455.
Mr. MacDonald and Dr. Wilt: Center for Chronic Disease Outcomes Research (111-0), Minneapolis Veterans Affairs Health Care System, One Veterans Drive, Minneapolis, MN 55417.
Dr. Bliss: University of Minnesota School of Nursing, 6-189A WDH, 308 Harvard Street SE, Minneapolis, MN 55455.
Dr. Rector: Center for Chronic Disease Outcomes Research, Minneapolis Veterans Affairs Health Care System, 152/2E, One Veterans Drive, Minneapolis, MN 55417.
Author Contributions: Conception and design: D.M. Drekonja, M. Butler, D. Bliss, G.A. Filice, T.J. Wilt.
Analysis and interpretation of the data: D.M. Drekonja, M. Butler, R. MacDonald, D. Bliss, G.A. Filice, T.S. Rector, T.J. Wilt.
Drafting of the article: D.M. Drekonja, M. Butler, R. MacDonald, D. Bliss, G.A. Filice.
Critical revision of the article for important intellectual content: D.M. Drekonja, M. Butler, D. Bliss, G.A. Filice, T.S. Rector, T.J. Wilt.
Final approval of the article: D.M. Drekonja, M. Butler, D. Bliss, T.S. Rector, T.J. Wilt.
Provision of study materials or patients: R. MacDonald.
Statistical expertise: R. MacDonald, T.S. Rector, T.J. Wilt.
Obtaining of funding: T.J. Wilt.
Collection and assembly of data: D.M. Drekonja, M. Butler, R. MacDonald, D. Bliss.
Administrative, technical, or logistic support: T.J. Wilt.
Drekonja DM, Butler M, MacDonald R, Bliss D, Filice GA, Rector TS, et al. Comparative Effectiveness of Clostridium difficile Treatments: A Systematic Review. Ann Intern Med. 2011;155:839-847. doi: 10.7326/0003-4819-155-12-201112200-00007
Download citation file:
Published: Ann Intern Med. 2011;155(12):839-847.
Clostridium difficile infection is increasing in incidence and severity. The optimal treatment is unknown.
To determine whether, among adults with C. difficile infection, treatment with certain antibiotics compared with others results in differences in initial cure, recurrence, and harms.
MEDLINE, AMED, ClinicalTrials.gov, and Cochrane databases (search dates: inception through August 2011, limited to English-language reports); bibliography review.
Randomized, controlled trials of adults with C. difficile infection, independent of outcomes, who were treated with medications available in the United States. Observational studies reporting strain were included.
Study design, inclusion and exclusion criteria, quality and strength of evidence as assessed by 2 reviewers, study definitions, and duration of treatment and follow-up. Outcomes included initial cure, recurrence, and treatment harms.
11 trials that included 1463 participants were identified. Three trials compared metronidazole with vancomycin; 8 compared metronidazole or vancomycin with another agent, combined agents, or placebo. Strain was analyzed in 1 trial and 2 cohort studies. No study comparing 2 antimicrobial agents demonstrated a statistically significant difference for initial cure; all comparisons were of low to moderate strength of evidence. Moderate-strength evidence from 1 study demonstrated that recurrence was decreased with fidaxomicin versus vancomycin (15% vs. 25%; difference, −10 percentage points [95% CI, −17 to −3 percentage points]; P = 0.005). Subgroup analysis of a single study comparing metronidazole with vancomycin for patients who have severe C. difficile infection showed no difference by intention-to-treat analysis; this was rated as insufficient-strength evidence. Harms, when reported, did not differ between treatments in any study.
Definitions of diarrhea, C. difficile infection, initial cure, and relapse varied. Some studies reported insufficient detail to allow assessment of all randomly assigned participants or of harms.
No antimicrobial agent is clearly superior for the initial cure of C. difficile infection. Recurrence is less frequent with fidaxomicin than with vancomycin.
U.S. Department of Health and Human Services.
Several antibiotics are available for treatment of Clostridium difficile infections.
This comparative effectiveness review of 11 randomized trials involving adults with C. difficile found that initial cure rates did not statistically significantly differ for fidaxomicin, vancomycin, and metronidazole. Recurrent disease was common with all agents but was less frequent with fidaxomicin (15%) than with vancomycin (25%).
There were few studies, and most had small sample sizes. Studies used heterogeneous definitions for cure. Potential harms were inconsistently reported.
Any of several agents could be used for initial treatment of C. difficile infection. Fidaxomicin may be a superior agent for recurrent infection.
Clostridium difficile infection is a common infection that usually causes diarrhea but occasionally more severe disease, including pseudomembranous colitis and toxic megacolon (1). Antibiotic use precedes most episodes of this infection, and nosocomial acquisition is common (1). Attributable mortality increased from 5.7 to 23.7 deaths per 1 million persons from 1999 to 2004 in the United States (2). Therapy for C. difficile infection has been studied for more than 3 decades, with numerous randomized trials evaluating multiple agents and dosing strategies (3-17). Table 1 presents current treatment options available in the United States. Authors of a 1995 position paper found that comparative clinical trials demonstrated equivalent rates of clinical cure for metronidazole and vancomycin (6-7); however, they stated that metronidazole may be preferred because of concerns regarding the emergence of vancomycin resistance, particularly vancomycin-resistant enterococci (18-19), and costs (20). A 2007 Cochrane review concluded that no antibiotic was clearly superior for the treatment of C. difficile infection (21).
Coinciding with increasing incidence and severity of C. difficile infection (22), further treatment studies have been published (3, 17. Largely on the basis of these new data, recent reviews and guidelines endorse vancomycin as the preferred treatment for this severe infection (23-24), and vancomycin use for it has increased (25). More recently, data from a large trial comparing vancomycin with fidaxomicin has led to U.S. Food and Drug Administration approval of fidaxomicin for treatment of C. difficile infection (26). Because of these changes, an updated assessment of the evidence evaluating antimicrobial treatment of C. difficile infection is necessary. Accordingly, as part of an Agency for Healthcare Research and Quality (AHRQ)–funded review (27), we conducted a systematic review of the literature regarding C. difficile infection therapy, including updating the AHRQ review to include the current literature. Specifically, we assessed the comparative effectiveness and harms of different antibiotic treatments in prospective randomized trials, including whether effectiveness varied according to disease severity or strain.
This topic was nominated by the Medicaid Medical Directors Learning Network and selected by AHRQ for systematic review. The Minnesota Evidence-based Practice Center followed a standardized protocol. Details of the protocol can be found elsewhere (27).
We searched MEDLINE, AMED (Allied and Complementary Medicine), Cochrane databases, and ClinicalTrials.gov for studies of C. difficile infection published from database inception through August 2011. We used keywords to identify condition (difficile) and interventions and used a filter to identify study design. Appendix Table 1 shows our search strategy. In addition, we reviewed bibliographies, received suggestions regarding possibly relevant literature from topic experts, and searched the U.S. Food and Drug Administration Web site for relevant study data.
Appendix Table 1.
Two extractors independently screened all abstracts, then screened all potentially relevant, full-text articles for final inclusion. We included randomized, controlled trials published in English of adult patients with C. difficile infection treated with medications available to clinicians in the United States. We anticipated that there would be few controlled trials for the subquestions of whether treatment efficacy differed when stratified by disease severity or strain; accordingly, for these questions we included studies of less robust design, including retrospective cohorts. We excluded studies presented only at professional meetings.
Two trained extractors abstracted study and patient characteristics and clinically relevant outcomes, including initial cure, recurrence, duration of diarrhea, and overall mortality. We used the definitions of initial cure and recurrence reported in the individual studies and recorded between-study differences. We also abstracted data on clearance of organism or toxin. When possible, we used a strict intention-to-treat analysis, defined as including all randomly assigned patients who met inclusion criteria.
We evaluated the overall strength of evidence using methods developed by AHRQ and the Effective Health Care Program (28). Evidence quality is rated as high (further research is very unlikely to change the confidence in the estimate of effect, meaning that the evidence reflects the true effect), moderate, low, or insufficient (evidence is unavailable or does not permit a conclusion). Domains evaluated to rate the strength of evidence included 1) risk for bias, 2) consistency, 3) directness, and 4) precision (28). Study quality was assessed for allocation concealment (whether treatment assignment remains unknown until randomization), blinding, and type of analysis (29).
The limited number of studies and study heterogeneity precluded a pooled analysis. Accordingly, we present qualitative summaries and data syntheses of individual studies.
Funding was from AHRQ, U.S. Department of Health and Human Services. A representative from AHRQ provided technical assistance and provided comments on draft versions of the full evidence report. The AHRQ staff did not directly participate in the literature search; determination of study eligibility criteria; data analysis or interpretation; or preparation, review, or approval of the manuscript for publication.
The search identified 1078 possible relevant articles (Appendix Figure). After initial abstract review, we excluded 1065 articles. The remaining 13 articles met eligibility criteria. Two were observational studies comparing the efficacy of different antimicrobial agents based on strain of C. difficile(30-31), whereas 11 were randomized trials evaluating different antimicrobial agents (or different doses of a single drug) for the treatment of C. difficile infection (3-8, 11, 13, 15, 17, 26. The 11 randomized trials enrolled 1463 patients and reported efficacy analysis on 1239 patients. A total of 1324 patients met inclusion criteria and had data available for an intention-to-treat analysis (Appendix Table 2). Vancomycin, the most frequently assessed antimicrobial agent, was included in 9 of 11 studies (3, 4, 6, 7, 11, 13, 15, 17, 26. Metronidazole was the next most frequently studied antimicrobial agent (5 studies) (3, 6-8, 10, followed by bacitracin (4, 15 and nitazoxanide (8, 17 (2 studies each), then fidaxomicin (26) and rifampin (with metronidazole) (10) (1 study each).
Appendix Table 2.
The 11 trials varied considerably for several characteristics, including location; methods; allocation concealment; demographic characteristics; duration of treatment and follow-up; and definitions of C. difficile infection, initial cure, and recurrence (Appendix Tables 2 and 3). Six studies were conducted in the United States (3, 7, 8, 13, 15, 17, whereas a single study each was conducted in Australia (4), Austria (6), Canada (10), the United Kingdom (11), and both the United States and Canada (26). Most studies were direct comparisons between 2 drugs (8 studies) (3, 4, 6-8, 15, 17, 26; the remaining 3 compared different doses of the same drug (13), the same drug with another antimicrobial agent added (10), or a drug with placebo (11). Table 2 documents study quality elements (allocation concealment, blinding, and presence of an intention-to-treat analysis). Although an “intention-to-treat analysis” was presented in 5 studies (4, 10, 11, 17, 26, patients were removed in 4 studies after being randomly assigned because exclusion criteria were later discovered (7, 15, 17, 26. Patients were predominantly inpatients. One study included a single infant (13); otherwise, all patients were adults. In 2 studies, all or most patients reportedly were on a surgical service (7, 11, whereas the type of service was not specified in all other studies. The definition of C. difficile infection invariably included a test for C. difficile toxin or organisms combined with diarrhea; however, definitions of diarrhea varied, and some studies required an additional manifestation of C. difficile infection, including fever, abdominal pain, or leukocytosis.
Appendix Table 3.
All randomized trials reported the outcomes of initial cure and recurrence, although definitions varied (Appendix Table 2). For instance, Teasley and colleagues (7) considered a patient initially cured if “diarrhoea resolved within 6 treatment days,” Lagrotteria and coworkers (10) defined initial cure as the patient becoming “asymptomatic during the treatment course,” whereas Zar and associates (3) defined initial cure as “resolution of diarrhea and a negative result of a C. difficile toxin A assay at days 6 and 10 of treatment.” This last definition is notable for including a test for C. difficile toxin in the determination of initial cure; most studies used clinical assessment only, although 3 did incorporate C-reactive protein or peripheral leukocyte measurements (6, 8, 17. Definitions of recurrence were similar in that most required both reappearance of symptoms and a confirmatory microbiological test result; however, the duration of follow-up varied from 21 to 60 days. Authors were inconsistent in their reporting of duration of diarrhea and clearance of organism or toxin. Mortality was also inconsistently reported; 8 studies reported overall mortality stratified by treatment received, 2 reported overall mortality without specifying treatment received, and 1 reported incomplete mortality data.
The most frequently compared antimicrobial agents were vancomycin and metronidazole, which were compared in 3 studies (3, 6-7). Study size ranged from 62 to 172 patients, totaling 335, all inpatients. Dosage of study medications varied (Appendix Table 2), but duration was uniformly 10 days. Each study reported initial cure, recurrence, and mortality; however, mortality was not reported by treatment received in 1 study (6). Reports of duration of diarrhea and clearance of organism or toxin varied.
The percentage of patients initially cured with vancomycin ranged from 84% to 94% among individual studies (Figure 1); the percentage of patients initially cured with metronidazole ranged from 73% to 94%. No study demonstrated a statistically significant difference between treatments. A single study reported a prespecified subgroup analysis on patients with severe disease, with severity based on a clinical score and classification occurring before randomization (3). By strict intention-to-treat analysis of all 82 randomly assigned patients with severe disease, the percentage with initial cure was not significantly different for vancomycin compared with metronidazole (79% vs. 66%; P = 0.22), corresponding to a risk ratio for initial cure with vancomycin of 1.20 (95% CI, 0.92 to 1.57).
The percentage of participants meeting the various definitions of recurrent disease (after meeting criteria for initial cure) ranged from 7% to 17% with vancomycin (Figure 2). For metronidazole, the range was 5% to 21%. No difference was statistically significant.
Death was rare; 5 patients died in each group among the 335 patients enrolled in studies comparing vancomycin with metronidazole. Death was not reported by treatment group in 1 study, which included 2 other antimicrobial agents (fusidic acid and teicoplanin) that are not discussed here because they are unavailable in the United States (6).
Among studies that reported these outcomes, there were no statistically significant differences between vancomycin and metronidazole. Mean duration of diarrhea was 2.4 to 3.2 days, and toxin or organism cleared in 60% to 74% of patients.
Eight studies, ranging in size from 21 to 629 patients, compared metronidazole or vancomycin with a different antimicrobial agent, placebo, or the same drug with another antimicrobial added (Appendix Table 2). Two studies compared vancomycin with bacitracin (4, 15, and 1 each compared metronidazole with nitazoxanide (8), vancomycin with nitazoxanide (17), high-dose with low-dose vancomycin (13), metronidazole with metronidazole plus rifampin (10), vancomycin with fidaxomicin (26), and vancomycin with placebo (11). Dosages, therapy duration, definition of outcomes, and reporting of outcomes varied among studies. Patients were inpatients except for 3 studies with 5%, 14%, and 41% outpatients, respectively.
Figure 3 presents results of the 6 studies comparing vancomycin with other agents. Except for the study comparing vancomycin with fidaxomicin, all analyzed fewer than 60 patients. The single statistically significant difference in percentage of patients initially cured was in a study of 21 patients comparing vancomycin with placebo (11). In contrast, in the 4 comparisons of vancomycin with an active alternative regimen, the frequency of initial cure did not significantly differ between groups. Likewise, in the 2 studies comparing metronidazole with other agents, the percentage of patients initially cured did not differ significantly by treatment group (data not shown).
The percentage of patients experiencing recurrence ranged from 9% (17) to 37% (4) in individual studies. Only the comparison between fidaxomicin and vancomycin showed a statistically significant difference (15% vs. 25%; difference, −10 percentage points [CI, −17 to −3 percentage points]; P = 0.005); in all other trials, the percentage of patients with recurrence did not significantly differ. Between-trial comparisons for the percentage of patients with recurrence are of uncertain relevance because of the variable definitions of recurrence and duration of follow-up.
In 1 study (10), 32% of patients (6 of 19) receiving metronidazole and rifampin died compared with 5% of those (1 of 20) receiving metronidazole alone (P = 0.044). The authors attributed this high mortality rate to elderly patients with multiple severe comorbid conditions. In the 6 remaining studies, the mortality rate ranged from 0% to 9% and did not differ between groups.
Although authors inconsistently reported duration of diarrhea, there were no statistically significant between-treatment differences in duration among studies comparing metronidazole or vancomycin with other agents (4, 6, 7, 10, 13. In 2 studies comparing vancomycin with bacitracin (4, 15, stool specimens obtained 7 to 10 days after initiation of treatment were less likely to contain toxin (risk ratio, 1.68 [CI, 1.14 to 2.47]) or viable C. difficile organisms (risk ratio, 1.67 [CI, 1.12 to 2.49]) among patients receiving vancomycin than among those receiving bacitracin. Vancomycin was also superior to placebo for both clearance of toxin (risk ratio, 2.75 [CI, 1.17 to 6.44]) and organism (risk ratio, 8.25 [CI, 1.29 to 52.77]).
Authors inconsistently reported harms attributed to study medications, with some authors reporting only adverse events leading to cessation of therapy (3, 7, others reporting all adverse events (6, 10, 17, 26, and others including a statement to the effect that all medications were well-tolerated (11, 13, 15. One study reported unspecified serious adverse events in each treatment group, but the investigators stated that none was judged to be due to study medications (8). One study did not address harms in any way (4). The study reporting the highest mortality rates also reported the highest rates of adverse effects, with 15 of 39 patients (38%) experiencing rash, nausea, or vomiting (10). There were no statistically significant differences between treatments in any study reporting adverse events by therapy received. Four studies assessed resistance to the studied antimicrobial agents (4, 8, 13, 26; no resistant C. difficile isolates were detected. No studies assessed the development of resistance in non–C. difficile organisms (such as vancomycin-resistant enterococci).
A single trial included data on efficacy stratified by strain, which was available for 415 of 602 patients (26). Initial cure and recurrence were assessed by strain, categorized as North American pulsed-field gel electrophoresis type 1 (NAP1) vs. non-NAP1. For initial cure, no significant difference was observed between patients treated with fidaxomicin and those receiving vancomycin, regardless of strain. However, among patients with non-NAP1 strains, those treated with fidaxomicin experienced recurrence less frequently than those treated with vancomycin (10% vs. 28%; P < 0.001), whereas recurrence was similarly frequent regardless of treatment among patients with the NAP1 strain.
Because trial data addressing the effect of strain on treatment were limited, we also looked for less robust studies stratifying efficacy of treatment by strain type. We identified 2 studies that attempted to address this question. One was a prospective observational study of 52 patients receiving metronidazole or vancomycin for C. difficile infection, in which the authors categorized patients as having the “epidemic” strain (specified as NAP1) or a “nonepidemic” strain (31). In this study, in which 60% of patients had the epidemic strain, a significantly greater proportion of vancomycin recipients had resolution of diarrhea at day 5 compared with patients receiving metronidazole. However, between days 5 and 10, almost all patients who continued taking either agent had symptom resolution, and response rates did not differ between epidemic and nonepidemic strains. A larger retrospective study looked at the incidence of severe or complicated C. difficile infection during a period that included the widespread emergence of the NAP1 strain, stratified by initial treatment with metronidazole or vancomycin. From 1991 to 2002 (defined as the preepidemic strain period), vancomycin use was associated with a lower probability of developing severe or complicated C. difficile infection compared with metronidazole use, but this difference disappeared in 2003 to 2006 (defined as the epidemic strain period) (30).
We rated the strength of the evidence as moderate for comparisons of metronidazole with vancomycin for the initial cure of C. difficile infection, of fidaxomicin with vancomycin for both initial cure and recurrence, and of vancomycin compared with placebo for initial cure. We rated the strength of the evidence for the remaining comparisons as low or insufficient (Table 3); no comparison received a high rating in strength of evidence.
In this systematic review of antimicrobial treatment for adults with C. difficile infection, we found moderate-strength evidence that fidaxomicin and vancomycin did not differ for the outcome of initial cure but that recurrence was less frequent with fidaxomicin. We also found moderate-strength evidence that outcomes did not differ among patients treated with metronidazole or vancomycin, the two most commonly used agents. Almost all other comparisons resulted in imprecise estimates of comparative effectiveness and frequently were from a single study of relatively small size. We found insufficient evidence that vancomycin is superior to metronidazole for the outcome of initial cure in patients with severe disease, with a per-protocol subgroup analysis of a single study demonstrating a significant difference. Of the other active agents compared with metronidazole or vancomycin, we found low-strength evidence that there was no difference in the outcomes of initial cure or recurrence. Definitions and timing of outcome assessment varied between studies.
The available evidence suggests that for patients with nonsevere disease, fidaxomicin, metronidazole, and vancomycin all provide effective treatment. Recurrence is common with all agents but develops less frequently with fidaxomicin when compared with vancomycin. Whether this decrease in recurrence will lead to widespread use of fidaxomicin is unknown. Multiple factors will probably influence use, including the willingness of providers to prescribe a new medication, safety data as use becomes more widespread, local patterns of incidence and recurrence of C. difficile infection, and costs (including both drug costs and those associated with recurrent C. difficile infection).
Currently, metronidazole and vancomycin are the most commonly used drugs to treat C. difficile infection, and despite no evidence for superior efficacy in mild to moderate disease, vancomycin use is increasing (25). This increase has multiple potential causes, including the finding that both metronidazole and vancomycin have been linked to the emergence of vancomycin-resistant enterococci (32), the availability of intravenous vancomycin compounded as an oral suspension that provides a less costly alternative to vancomycin capsules (33), and an increased incidence of severe disease (22, 34. Clinicians may also be extrapolating the reported superiority of vancomycin in severe disease (3) to patients with mild to moderate disease, or may have clinical experience that influences their decision making.
The only comparative evidence for vancomycin versus metronidazole in severe disease comes from a per-protocol subgroup analysis of 69 patients in a single trial (3). Whether vancomycin was superior to metronidazole depends on the method of data analysis; a strict intention-to-treat analysis yielded no significant difference in the percentage of patients initially cured, and the authors did not report a statistical test of interaction for this subgroup. However, both per-protocol and modified intention-to-treat analyses (presented in a subsequent response (35) to several letters to the editor (36-38)) identified a statistically significant difference favoring vancomycin. Finally, this study was the only included trial incorporating toxin-detection tests into the definition of initial cure, although subsequent letters addressed this by reclassifying 2 participants who met the definition for initial cure but were toxin-positive (35-36). However, because severe C. difficile infection is associated with considerable morbidity, mortality, and costs (22, 23, 39, more liberal use of vancomycin treatment is reasonable, especially because of decreased drug costs (Table 1) and lessened concerns regarding vancomycin resistance (32).
Two other studies assessed disease severity by using different definitions, with neither showing any statistically significant difference in outcomes. Because of the low level of evidence regarding the most effective treatment of severe C. difficile infection, we recommend that future trials address the comparative effectiveness of different treatment options in patients with severe C. difficile infection, preferably with a validated and easily applied definition of severity. One such definition has been presented in conference proceedings (40), but further validation of this or other classification systems is needed.
Information on the effect of C. difficile strain on treatment response comes from a single trial and 2 observational studies (26, 30-31). In the trial, only patients with non-NAP1 strains had a decrease in recurrent disease with fidaxomicin (compared with vancomycin), whereas the observational studies provide indirect and partially conflicting data on the comparative efficacy of metronidazole and vancomycin. Although future studies in this area would be of interest, the clinical relevance of such data are currently limited because strain typing of stool isolates is rare in practice.
Few studies compared other agents with metronidazole or vancomycin. In addition to the finding that fidaxomicin use resulted in less recurrence than vancomycin, the only significant difference that emerged was the superiority of vancomycin to bacitracin for clearance of organism and toxin. The clinical relevance of this finding is uncertain because bacitracin is rarely used to treat C. difficile infection, clearance of organism or toxin in an otherwise-well patient is of unknown benefit, and assessment of multiple outcomes may lead to identification of statistically significant findings by chance.
Most studies did not show a mortality difference between study drugs, although few deaths occurred. An exception was the study that added rifampin to metronidazole, in which the mortality rate was greater among rifampin recipients. The authors clarified that 1 patient died before receiving any drug and that none of the deaths was attributed to rifampin (10). However, the higher mortality rate is consistent with another report describing increased mortality associated with use of adjunctive rifampin (41).
Limitations of the available evidence include substantial variability among studies, including the definitions used for C. difficile infection, initial cure, and recurrence, and the durations of treatment and follow-up. In aggregate, these differences make between-study comparisons difficult, and we caution against drawing definitive conclusions. Particularly problematic are the differences in initial cure definitions, because some studies used clinical assessment alone, whereas others included laboratory variables. Similarly, differences in follow-up duration make comparisons of recurrence difficult because patients followed for 60 days after treatment (15) have a longer period at risk for recurrence than do those followed for 21 days (3, 7, 8, 17. A standard approach to defining disease, outcomes, and duration of follow-up would enhance future trials of treatment of C. difficile infection by making between-trial comparisons more valid. In addition, few trials reported outcomes on all randomly assigned patients, although most provided sufficient detail as to which group the excluded patients were assigned. Finally, harms were variably reported, although it is encouraging that the more recent studies include systematic collection and enhanced reporting of harms.
In summary, we found that the strength of evidence regarding the comparative effectiveness and harms of treatment for C. difficile infection is generally low to moderate. Only a few clinical trials are available to help establish the most effective treatment for this common and serious illness, and follow-up studies to reevaluate findings from initial studies are even scarcer. We found moderate-strength evidence that fidaxomicin, metronidazole, and vancomycin are effective for initial cure and that no agent is clearly superior for this outcome. Recurrent C. difficile infection is frequent with all agents, but we found moderate-strength evidence that fidaxomicin, recently approved for this infection in the United States, leads to fewer recurrences than vancomycin for non-NAP1 strains. Although alternative agents are available, limited data available suggest that none differs significantly from metronidazole or vancomycin for any assessed outcomes. For patients with severe disease, there is insufficient evidence that vancomycin is superior to metronidazole (3).
The In the Clinic® slide sets are owned and copyrighted by the American College of Physicians (ACP). All text, graphics, trademarks, and other intellectual property incorporated into the slide sets remain the sole and exclusive property of the ACP. The slide sets may be used only by the person who downloads or purchases them and only for the purpose of presenting them during not-for-profit educational activities. Users may incorporate the entire slide set or selected individual slides into their own teaching presentations but may not alter the content of the slides in any way or remove the ACP copyright notice. Users may make print copies for use as hand-outs for the audience the user is personally addressing but may not otherwise reproduce or distribute the slides by any means or media, including but not limited to sending them as e-mail attachments, posting them on Internet or Intranet sites, publishing them in meeting proceedings, or making them available for sale or distribution in any unauthorized form, without the express written permission of the ACP. Unauthorized use of the In the Clinic slide sets will constitute copyright infringement.
Martin H., Floch, Clinical Professor of Medicine
Yale University School of Medicine
December 30, 2011
Probiotics and fecal microbial transplants are treatments for Clostridium difficile infection
To the Editor:
Probiotics and fecal microbial transplants are treatments for Clostridium difficile infection.
In the December 20, 2011 issue of the Annals, the article entitled "Comparative Effectiveness of Clostridium difficile Treatments. A Systemic Review" as published by Drekonja and colleagues reports on a systemic review of antibiotic therapy but makes no mention of other treatments for recurrent and severe Clostridia difficile diarrhea infection. It is surprising that their discussion does not mention these treatments. Although there is controversy, the literature and studies in many institutions revealed that Saccharomyces boulardii and Lactobacillus GG have been reported repeatedly to be helpful in recurrent Clostridia difficile infection (1-3). Furthermore, the use of fecal microbial transplant (FMT) is now demonstrating surprising cures in patients with severe and resistant C. difficile infection. The review article published in the Annals leaves the reader with the impression that only antibiotics are treatments for this severe infection. Certainly, these other modalities of treatment, probiotics and FMT, are important and at times have been shown to be life-saving (4,5). They certainly should be mentioned at least in the discussion, and I would hope that the authors will correct the erroneous impression from this systemic review of antibiotics that only antibiotics are helpful in the treatment of Clostridia difficile.
1. Floch MH, Madsen KK, Jenkins DJA, et al. Recommendations for probiotic use. J Clin Gastroenterol 2006;40:275-278
2. Floch MH, Walker WA, Guandalini S, et al. Recommendations for probiotic use-2008. J Clin Gastroenterol 2008;42:S104-S108
3. Floch MH, Walker WA, Madsen K, et al. Recommendations for probiotic use -2011 Update. J Clin Gastroenterol 2011;45:S168-S171
4. Brandt L, Reddy SS. Fecal microbiota transplantation for recurrent Clostridium difficile infection. J Clin Gastroenterol 2001:45:S159-S167
5. Persky SE, Brandt LJ. Treatment of recurrent Clostridium difficile- associated diarrhea by administration of donated stool directly through a colonscope. Am J Gastroenterol 2000;95:3283-3295
Gastroenterology/Hepatology, Infectious Disease, Healthcare Delivery and Policy, High Value Care, Diarrhea.
Results provided by:
Copyright © 2016 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use
This PDF is available to Subscribers Only