Proton Pump Inhibitors Increase Significantly the Risk of Clostridium difficile Infection in a Low-Endemicity, Non-Outbreak Hospital Setting
Dalton BR, Lye-Maccannell T, Henderson EA, Maccannell DR, Louie TJ
Aliment Pharmacol Ther. 2009;29:626-634
Clostridium difficile is a highly prevalent cause of infectious diarrhea in the community and increasingly in hospitalized patients. The rate of C difficile infection has risen dramatically, particularly in hospitalized patients in whom the disease incidence has been climbing at a rate of approximately 25% per year since 2000. Current estimates are that C difficile infection complicates 0.5% to 1.0% of all hospital admissions. Because many hospitalized patients receive acid-suppressive therapy with proton-pump inhibitors (PPIs), several recent reports have suggested a possible causal relationship between PPI therapy and C difficile infection. The premise is that reduction of gastric acid secretion may allow an infectious agent, such as C difficile, to be ingested and perhaps survive a "first wall of host defense" -- the acidic gastric pool. Reduced gastric acid might facilitate survival of these potentially noxious bacteria in the upper gastrointestinal tract, leading to disease-associated sequelae.
This retrospective cohort study was conducted at 2 hospitals in Canada. Data from patients who had been hospitalized for ≥ 7 days and received antibiotics were included in the analysis. Demographics, exposure to PPIs, and other variables were assessed by univariate and multivariate analyses. Of 14,719 patients, 149 (1%) had first-episode C difficile infection. Exposure to PPIs increased the risk for C difficile infection in these patients (odds ratio [OR] 1.96, 95% confidence interval [CI] 1.42-2.72). Other risks were the use of anticonvulsant therapy, antiplatelet agents, osteoporosis medication, angiotensin II receptor antagonists, asthma medications, systemic corticosteroids, antidiabetic agents, or antidepressant medications. Additionally, age, antidepressants, length of admission, and admission service all significantly increased the OR for incremental risk. The highest identified risk was actually with use of antidepressants (OR 2.99, 95% CI 2.16-4.15). By logistic regression, the number of days of PPI use was a significant predictor of C difficile infection (OR 1.01/day, 95% CI 1.00-1.02). Although a statistically significant increase in C difficile infection was observed in the patients who received antibiotics and PPIs, the absolute risk was modest, and the study authors suggest that the benefits of PPI therapy outweigh the risk of developing C difficile infection.
Clearly, there has been a lot of "noise" in regard to the indications for PPIs and the potential adverse consequences associated with PPI therapy. Reports have suggested a relationship not only between PPI use and C difficile infection, but also between PPI use and other notable sequelae, such as hip fracture and vitamin B12 depletion. There is also increasing interest in the possibility that PPI therapy may interfere with the antiplatelet activity of clopidogrel (Plavix®).
The putative mechanism of increased C difficile infection with PPI use is the agent's acid-suppressive effects. There is a paradox here, however, because C difficile spores are resistant to gastric acid. Hence, the mechanism of the reported association between PPI therapy and C difficile infection is unclear. If acid suppression is the true cause of C difficile infection in patients receiving PPIs, then the highest risk for infection should be seen with achlorhydia (no acid production), a state associated with pernicious anemia. Because pernicious anemia is a code-based diagnosis, retrospective studies could determine whether an association existed between this diagnosis and C difficile infection. Such an association would provide an excellent if not pivotal cornerstone to the acid-suppression hypothesis. If an association between the achlorhydric state and C difficile infection is found, the relationship between lesser degrees of acid suppression and C difficile infection could be further evaluated.
The data in regard to PPI use suggest that a pH target of > 4 is achieved for 11-14 hours with a standard once-daily dosing. By no means do PPIs raise patients' gastric pH to levels as high as 7 (coinciding with achlorhydria). A nested case-control study within any of the retrospective studies, to include patients with pernicious anemia, could provide the evidence needed to determine whether acid suppression is associated with C difficile infection. However, to date, none of the retrospective studies looking at outcomes associated with PPI use employed this scientific design, considerably weakening the reported associations.
The study authors are to be commended for acknowledging the relatively modest potential risk for C difficile infection associated with PPI use. The unimpressively low relative risk is notable and should be kept in perspective with the balanced risks and benefits of PPI therapy. Clearly, if patients do not need PPIs, these agents should not be prescribed. However, data still suggest that these agents have a significant benefit for management of acid secretion-related diseases and are the mainstay of prophylactic strategies to reduce bleeding of peptic or stress ulcers associated with nonsteroidal anti-inflammatory use.
Finally, a reality check in regard to allegations of adverse consequences of PPI use is in order. The message for healthcare providers is that balancing the risks and benefits is a key concept of any disease management strategy. The epidemiologic studies of PPI use that document potential adverse consequences, such as C difficile infection, have been both supported and refuted by other evidence.[2,3] The present study suggests that if there is an increased risk, it is very modest. However, the greater message is that current data support the concept that the benefits of PPI therapy -- when appropriately justified for use -- outweigh the risk of developing C difficile infection.
Medscape Gastroenterology © 2009 Medscape, LLC
Cite this: David A. Johnson. Proton-Pump Inhibitors and Clostridium difficile Infection Risk: New Insights Toward an Understanding of Potential Risks - Medscape - May 19, 2009.