|Statins for the prevention and treatment of infections: a systematic review and meta-analysis
|Tleyjeh I M, Kashour T, Hakim F A, Zimmerman V A, Erwin P J, Sutton A J, Ibrahim T
This review concluded that statin use may be associated with benefits in treating and preventing infections, but given the heterogeneity between studies and evidence of publication bias, randomised controlled trials are needed to confirm this benefit of statin use. These conclusions are likely to be reliable.
To evaluate the effects of statins in the prevention and treatment of infections.
MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials, Scopus and Web of Science were searched, without language restriction, for dates ranging from 1990 to December 2007. Search terms were reported. References of all included studies were also checked.
Randomised controlled trials (RCTs) or cohort studies that included at least 50 patients, and examined the association between statin use and the incidence or outcome (including complications, severity and mortality) of infections, were eligible for inclusion. Cohort studies were required to report an adjusted effect estimate for at least one potential confounding variable.
All included studies were prospective or retrospective cohort studies. Inclusion criteria for patients varied considerably. Treatment studies assessed patients with bacteraemia, pneumonia, sepsis and bacterial infection. Prevention studies assessed patients with vascular diseases, chronic kidney diseases, diabetes, intensive care unit-acquired infections, and those from general practice. Varying definitions of statin use were employed; statin use ascertainment methods included medical records, filing of prescriptions, interviews, and hospital records and drug charts. Included studies considered potential confounders such as demographic factors, co-morbidities, other prognostic factors, severity of illness (for treatment studies) and health care use (for prevention studies). Statins used were atorvastatin, cerivastatin, fluvastatin, lovastatin, pravastatin and simvastatin.
Two reviewers independently assessed the studies for inclusion in the review.
Assessment of study quality
Two reviewers independently assessed the studies for validity using the Newcastle-Ottawa Quality Assessment Scale for cohort studies.
Data were extracted on outcomes for prevention or treatment of infections, including adjusted effect estimates with 95% confidence intervals (CI) and the variables used to generate them. Two reviewers independently extracted the data using a predefined form. Disagreements were discussed with two other reviewers and agreement was reached through consensus. Authors were contacted for missing data.
Methods of synthesis
Prevention and treatment studies were combined in separate meta-analyses using the DerSimonian and Laird random-effects model to calculate pooled adjusted effect estimates. Heterogeneity for each analysis and overall was assessed using the I2 statistic. The following predefined sources of heterogeneity in treatment studies were investigated using univariate meta-regressions: use of odds ratio as effect measure; population-based study; use of propensity score analysis; use of mortality as an outcome measure; and underlying disease was community-acquired pneumonia versus bacteraemia versus sepsis. Identical analyses were conducted for prevention studies, with the exception of underlying disease where the comparison was between vascular diseases and others. Publication bias was assessed by visual inspection of funnel plots and Egger's test.
Results of the review
Sixteen cohort studies were included in the review, of which nine were treatment studies and seven were prevention studies. Quality varied in both treatment and prevention cohorts. Of the nine treatment studies, four were population-based or multi-centre, two used blinded outcome assessment and eight adjusted for multiple confounders. Of the seven prevention studies, four were population-based or multi-centre, one used blinded outcome assessment and six adjusted for multiple confounders.
The pooled adjusted effect estimate for treatment studies suggested an improvement in outcomes in patients taking statins (0.55, 95% CI: 0.36 to 0.83). However, there was highly significant statistical heterogeneity (I2=76.5%). The pooled adjusted effect estimate for prevention studies also suggested an improvement in outcomes in patients taking statins (0.57, 95% CI: 0.43 to 0.75), and also showed highly significant statistical heterogeneity (I2=82%). In neither case did meta-regression identify factors that explained the observed heterogeneity. There was some evidence of publication bias in funnel plots for both treatment and prevention studies but this only showed statistical significance on the Egger test for the prevention studies.
Statin use may be associated with benefits in treating and preventing infections. Given the heterogeneity between studies and evidence of publication bias there is a need for RCTs to confirm a benefit of statin use for these outcomes.
The review question and the inclusion criteria were clear. The authors searched several relevant databases, including one which listed conference abstracts. This, together with the lack of language restrictions, reduced the chances of relevant studies being omitted or biases being introduced. However, when publication bias was assessed, it was found to be present, and the authors took clear account of this in their conclusions. Rigorous methodology was used at all stages of the review process. An appropriate scale was used to assess study validity. There was considerable clinical heterogeneity between the studies in both meta-analyses, and this was reflected in significant statistical heterogeneity. Reasonable attempts at exploration of this heterogeneity failed to identify factors which were likely causes. Despite this potential concern, the authors' cautious conclusions reflected both the outcomes and the nature of the evidence and are likely to be reliable.
Implications of the review for practice and research
Practice: The authors stated that several issues require clarification before statins are used for the purpose of prevention or treatment of infections.
Research: The authors stated that there is a need for well-conducted placebo-controlled RCTs of statins in the context of infection prevention and treatment. Several such trials are underway for treatment, but these do not assess mortality as the primary outcome. The safety of statins in patients with sepsis also requires assessment, while the lack of an intravenous formulation or a clear pharmokinetic profile of statins would also need to be addressed before widespread use would be possible.
Tleyjeh I M, Kashour T, Hakim F A, Zimmerman V A, Erwin P J, Sutton A J, Ibrahim T. Statins for the prevention and treatment of infections: a systematic review and meta-analysis. Archives of Internal Medicine 2009; 169(18): 1658-1667
Other publications of related interest
Kopterides P, Falagas ME. Statins for sepsis: a critical and updated review. Clin Microbiol Infect 2009;15:325-34
Correction of error in Figure Legends: Tleyjeh IM, Kashour T, Hakim FA, Zimmerman VA, Erwin PJ, Sutton AJ, Ibrahim T. Statins for the prevention and treatment of infections: a systematic review and meta-analysis. Arch Intern Med 2010; 170(1): 42.
Subject indexing assigned by NLM
Bacteremia /prevention & Cross Infection /prevention & Diabetes Complications /prevention & Humans; Hydroxymethylglutaryl-CoA Reductase Inhibitors /therapeutic use; Infection Control; Pneumonia /prevention & Randomized Controlled Trials as Topic; Renal Insufficiency, Chronic /complications; Sepsis /prevention & Vascular Diseases /complications; control; control; control; control; control
Date bibliographic record published
Date abstract record published
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.