Medical News Stories
Histamine-2 Receptor Antagonists vs. Proton Pump Inhibitors on Gastrointestinal Tract Hemorrhage and Infectious Complications in the Intensive Care Unit
Study Question: Is the occurrence of gastrointestinal hemorrhage (GIH), pneumonia (PNA) and Clostridium difficile infection (CDI) different between proton pump inhibitor (PPI) and histamine-2 receptor antagonist (H2RA) therapies?
Study Description: This was a retrospective, multi-center, pharmacoepidemiologic, cohort study evaluating ICD-9 codes from the Premier Perspective database (Premier Inc.). Patients were included if requiring invasive mechanical ventilation (MV) for ≥ 24 hours and PPI or H2RA for ≥ 48 hours while on MV. Primary outcomes included secondary diagnosis (occurring during admission) of GIH, PNA, and CDI. Propensity-match multivariate analyses were performed to account for confounders.
Results: A total of 35,312 patients were included (38.1% in the H2RA group vs. 69.1% in the PPI group). Famotidine and pantoprazole were the most common H2RA and PPI, with 47% of PPI patients receiving > 40 mg/day. After matching (n = 17,598), the PPI group had greater antibiotic (77.3% vs. 81.2%, p < 0.001) and platelet inhibitor (30.8% vs. 33.4%, p < 0.001) use, while the H2RA group had greater anticoagulant use (36.5% vs. 31.1%, p < 0.001). The PPI group also had statistically longer therapy and MV duration. GIH (2.4% vs. 4.7%, p < 0.001), PNA (30.7% vs. 34%, p < 0.001), CDI (2.6% vs. 3.4%, p = 0.002), and ICU mortality (12.3% vs. 15.3%, p < 0.001) occurred more frequently in the PPI group. Multivariate analyses revealed PPI use had a higher odds ratio for GIH (1.95, 95% CI 1.44-2.65), PNA (1.23, 95% CI 10.7-1.43), and CDI (1.31, 95% CI 1.04-1.64).
Conclusion: PPIs are associated with greater risks of GIH, PNA, and CDI than H2RA in a heterogeneous ICU population receiving MV.
Perspective: Previous studies evaluating PPI versus H2RA have demonstrated increased risk of pneumonia and CDI with PPI. This is the first study to evaluate infectious outcomes in conjunction with GI hemorrhage prevention. This study is limited by indication bias, which the matching attempts to overcome. Prospective studies powered around clinically significant GIH and infectious complications are warranted, particularly for patients not primarily receiving famotidine or pantoprazole. Studies evaluating mechanisms behind these results are also desirable.