CRD summary

The authors concluded that acid suppressing drugs, particularly with concomitant antibiotics, were associated with increased risk of Clostridium difficile infection. The authors acknowledged heterogeneity and other limitations of the evidence, and that further research was likely to have an important impact on the findings. Therefore, increased risk may be overstated.

Authors' objectives

To assess the risk of Clostridium difficile infection in individuals using acid suppressing drugs and antibiotics (proton pump inhibitors).

Searching

PubMed and EMBASE were searched from inception to December 2011, without language restrictions. Search terms were available online. Reference lists of included studies and relevant reviews were manually screened for additional studies.

Study selection

Eligible for inclusion were controlled observational studies that assessed the risk of C. difficile infection in individuals using acid suppressing drugs (proton pump inhibitors) and concomitant antibiotics versus individuals not using these medications. Eligible studies had to report incidence and recurrence of C. difficile infection as odds ratios/risk ratios, or provide sufficient data to allow these statistics to be calculated.

Where reported, included studies were conducted between 1994 and 2009. Most studies were conducted in the USA, five were conducted in the UK, and others were undertaken in Canada, other parts of Europe or Asia. Most studies included patients with diarrhoea in single hospitals. Where reported, the mean age of participants ranged from 8.5 to 82.3 years. Most studies reported incidence of C. difficile infection, but three studies reported recurrence.

Participant selection criteria and methods to ascertain use of acid suppressing drugs varied across studies. Details on acid suppressing drugs and regimens and the diagnosis of C. difficile infection were lacking in the original articles.

Two reviewers independently screened studies for inclusion.

Assessment of study quality

Study risk of bias was assessed in accordance with the Cochrane Adverse Effects Methods Group, taking into consideration participant selection, ascertainment of proton pump inhibitor exposure, definition of C. difficile infection and statistical adjustment for confounders.

The authors did not explicitly state how many reviewers performed the assessment for study risk of bias.

Data extraction

Odds ratios and risk ratios were considered similar as incidence levels were low. Adjusted or unadjusted odds ratios/risk ratios and 95% confidence intervals were extracted, with preference for adjusted data. Where this was not reported, raw data were used to calculate unadjusted odds ratios.

Primary authors were contacted for clarification, where necessary.

Two reviewers independently extracted or calculated outcome data. Discrepancies were resolved through rechecking data and referral to a third reviewer.

Methods of synthesis

A random-effects model (using inverse variance method) was used to pool odds ratios and 95% confidence intervals. Statistical heterogeneity was assessed using I²; 30% to 60% represented moderate level of heterogeneity. Where possible, subgroup analyses were performed according to study design, outcome data (adjusted versus unadjusted), healthcare setting and method of diagnosing C. difficile infection. The number of patients requiring treatment for one additional person to have an adverse outcome (number-needed-to-harm) was also calculated.

Adjusted indirect comparisons were performed using Bucher's method. Comparisons in outcomes were made between participants using concomitant antibiotics with acid suppressing drugs, or participants using histamine receptor-2 antagonists (H2RAs) alone, versus participants using acid suppressing drugs alone. The proportion of risk attributable to interaction between acid suppressing inhibitors and antibiotics was also calculated.

Publication bias was assessed using funnel plots if sufficient studies were available and there was no evidence of significant statistical heterogeneity.

Results of the review

Forty-two observational studies (30 case-control and 12 cohorts) including 313,000 participants, were included in the review. Twenty-eight studies performed some form of adjustment for potential confounding variables. Thirty-eight studies reported positive C. difficile infection toxin tests. All studies were considered to be at some risk of bias.

Acid suppressing drugs: The risk of C. difficile incidence was significantly increased in patients using acid suppressing drugs compared to participants not using these drugs (OR 1.74, 95% CI 1.47 to 2.05; 39 studies), but with substantial statistical heterogeneity (I²=85%). The results were consistent in studies that provided adjusted data, but the difference was no longer significant in studies that provided unadjusted data. The numbers-needed-to-harm (reported by one study) were fully reported in the review.

The risk of C. difficile recurrence was significantly increased in participants using acid suppressing drugs (OR 2.51, 95% CI 1.16 to 5.44; three studies), again with substantial statistical heterogeneity (I²=78%). Findings were no longer significant when the two adjusted studies were pooled.

Subgroup analyses did not identify the source for statistical heterogeneity. Publication bias was not assessed due to substantial statistical heterogeneity.

Adjusted indirect comparisons: Concomitant acid suppressing drugs and antibiotics increased the risk of C. difficile infection when compared to acid suppressing drugs alone (OR 1.96, 95% CI 1.03 to 3.70; six studies). Use of histamine receptor-2 antagonists significantly reduced the risk of C. difficile infection compared to acid suppressing drugs (OR 0.71, 95% CI 0.53 to 0.97; 15 studies). Other findings were reported in the review.

Authors' conclusions

Despite statistical and clinical heterogeneity, the evidence indicated that acid suppressing drugs were associated with increased risk and recurrence of difficile infection, and this risk increased with concomitant antibiotics. Histamine receptor-2 antagonists were associated with a lower risk of C. difficile infection compared to acid suppressing drugs.

CRD commentary

The review question and supporting inclusion criteria were broad but clearly stated. The literature search was limited to two databases and it was unclear whether unpublished data were sought. There were no language restrictions which reduced potential for language bias. Study selection and data extraction were performed in duplicate, but it was unclear whether this was true for quality assessment, which means that reviewer error and bias cannot be ruled out. All studies were observational and the criteria used to assess study risk of bias indicated that all studies were at some risk of bias.

A large body of evidence was included in the review, but few study and participant details were provided in the original articles. There was substantial unexplained heterogeneity between studies, and confidence intervals were sometimes wide, which reduced the robustness of the findings. Findings from the indirect comparison analyses should be interpreted with caution as these methods have their own limitations and are based on studies with substantial heterogeneity.

The authors acknowledged that because of heterogeneity, imprecision, potential role of confounders from observational designs, and lack of data on different doses or duration of treatment, further research is very likely to have an important impact on the confidence in the estimate of effect and is likely to change the estimate. The review was published in 2012 and the literature search was up to 2011, which means that further relevant research may have been carried out since.

The evidence appeared to suggest some increased risk of C. difficile infection with acid suppressing drugs, but limitations of the evidence suggest the findings may be overstated. The indirect comparisons for concomitant antibiotics and histamine receptor-2 antagonists had additional limitations, which suggest the findings may not be reliable.

Implications of the review for practice and research

Practice: The authors stated that it may be prudent to withhold acid suppressing drugs in patients receiving antibiotics, unless there were clear gastrointestinal indications for acid-suppression therapy. Discontinuation of acid suppressing drugs should be strongly considered in patients diagnosed with C. difficile infection.

Research: The authors stated that further randomised controlled trials and prospective cohort studies were needed to investigate the effect of acid suppressing drugs on risk of C. difficile infection.

Funding

None.

Bibliographic details

Kwok CS, Arthur AK, Anibueze CI, Singh S, Cavallazzi R, Loke YK. Risk of Clostridium difficile infection with acid suppressing drugs and antibiotics: meta-analysis. American Journal of Gastroenterology 2012; 107(7): 1011-1019. [PubMed: 22525304]

Original Paper URL

http://www.nature.com/ajg/journal/v107/n7/abs/ajg2012108a.html

Indexing Status

Subject indexing assigned by NLM

MeSH

Anti-Bacterial Agents /adverse effects; Clostridium difficile; Enterocolitis, Pseudomembranous /chemically induced; Histamine H2 Antagonists /adverse effects; Humans; Incidence; Proton Pump Inhibitors /adverse effects; Recurrence; Risk Factors

AccessionNumber

12012036872

Database entry date

12/06/2013

Record Status

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.