Proton pump inhibitors use and clinical outcomes in critically ill patients with pulmonary fibrosis: A retrospective cohort study with propensity score matching

Objective Proton pump inhibitors (PPIs) are prescribed in the intensive care unit (ICU), but their association with outcomes in patients with pulmonary fibrosis remains unclear. Methods We conducted a retrospective cohort study using the Medical Information Mart for Intensive Care IV (MIMIC-IV) database (v3.1). Propensity score matching (PSM) and multivariate regression models adjusted for baseline differences. Landmark, subgroup, dose-response, and exploratory drug-specific analyses were performed. Results Among 551 patients, 286 (51.9%) received PPIs. After PSM, PPI use was associated with higher 30-day mortality (HR 1.59, 95% CI 1.08–2.34; p = 0.019), 90-day mortality (HR 1.62, 95% CI 1.15–2.30; p = 0.006), ICU mortality (OR 3.22, 95% CI 1.57–6.59; p = 0.001), and longer ICU stay (β 1.90, 95% CI 0.91–2.88; p < 0.001). Associations weakened in the 48-hour landmark analysis, with only 90-day mortality remaining significant (adjusted HR 1.40, 95% CI 1.02–1.90; p = 0.035). Exploratory drug-specific estimates varied across agents but were imprecise. In exploratory cumulative-dose analyses after PSM and adjustment, each 10-mg increase in omeprazole-equivalent exposure was associated with higher ICU mortality (OR 1.06, 95% CI 1.01–1.10; p = 0.012) and longer ICU stay (β 0.34 days, 95% CI 0.28–0.40; p < 0.001), but not 30-day or 90-day mortality. Initial timing was not associated with mortality after PSM and adjustment but was associated with longer ICU stays. Subgroup associations were consistent, with an interaction between PPI and H2RA use for ICU length of stay. Conclusions PPI exposure was associated with higher short-term mortality and longer ICU stay, although the associations weakened in the 48-hour landmark analysis, with only 90-day mortality remaining significant. Higher cumulative PPI doses were associated with higher ICU mortality and longer ICU stay. These findings are hypothesis-generating and require prospective confirmation before informing changes in clinical practice.