In February 2012, the FDA issued a safety alert regarding an association between proton-pump inhibitors (PPIs) and Clostridium difficile infection. In new meta-analyses, two groups of researchers used slightly different criteria to select studies in which this association could be evaluated; all included studies (23 and 42, respectively) were observational (cohort or case-control). Each meta-analysis involved roughly 300,000 patients.
In both meta-analyses, risk for C. difficile infection was significantly higher in PPI users than in nonusers (risk ratio, about 1.7). Although results across individual studies were heterogeneous, nearly all trended toward higher risk. Most of the included studies were adjusted for confounding variables, including antibiotic use. Concomitant use of both PPIs and antibiotics — examined in one meta-analysis — was associated with greater risk for C. difficile infection than was use of PPIs alone or antibiotics alone. Risk for C. difficile infection was higher with histamine (H)2-receptor antagonists than with no acid-suppressive therapy, but lower with H2-receptor antagonists than with PPIs.
Comment: The opportunity for residual confounding in these studies is substantial, because sicker patients are more likely both to receive PPIs and to be vulnerable to C. difficile infection. Still, these worrisome findings should remind clinicians to initiate PPIs only for valid indications and to stop PPIs in patients who take them for unclear reasons.