Dr Bajaj and colleagues' study of the rates of certain serious infections among US veterans with decompensated cirrhosis who were, or were not, subsequently started on proton pump inhibitor (PPI) therapy adds to the ongoing controversy about their potential risks.
In recent years, multiple studies have focused on putative risks of PPI therapy without considering its potential benefits; this bandwagon rolls merrily along. The patients whom Bajaj et al. studied were obviously very sick and more than half were started on some form of acid suppression – usually a PPI. As numerous other studies have demonstrated, PPI therapy is often a surrogate for poor health.[2, 3]
As the indications for PPI therapy are not addressed by Bajaj et al., it is impossible to determine how many patients received these agents appropriately. Therefore, we cannot quantify the benefit side in any risk–benefit analysis.
Bajaj et al. sought to determine the risks of PPI therapy on all serious infections and on those that they postulated were secondary to gastric acid suppression. However, one could question the validity of placing pneumonia in the latter category. Pneumonia is multifactorial and US veterans may have many of its predisposing factors that are aetiologically more important than (the oft-blamed) gastric acid suppression.
This study was necessarily retrospective in design. Although the summary hazard ratios were statistically significant, they were small. Therefore, as with many of these types of study, we should interpret the findings cautiously. When used appropriately, PPI therapy offers more benefit than risk.
To help ensure continued appropriate use, I agree that clinicians should re-evaluate the reason(s) for prescribing a PPI in all patients. However, there is currently no persuasive reason to throw out this baby with its bathwater and embrace the widespread use of a potentially less effective drug class (i.e. H2-receptor antagonists) in this particularly vulnerable group of patients.