Proton Pump Inhibitors Use Is Not Associated with the Risk of Covid-19-Related Mortality or Hospitalization: Croatian Nationwide Cohort Study

ABSTRACT

Introduction: Proton pump inhibitors (PPIs) play an indispensable role in the treatment of acid-secretion disorders and are one of the most widely used drugs. This study aimed to investigate the association between proton pump inhibitors (PPI) use and COVID-19-related mortality and hospitalizations. Aims & Methods: This population-based matched cohort study included all individuals diagnosed with the first episode of COVID-19 up to August 15, 2021, in Croatia. We classified patients based on exposure to PPIs and burden of PPI-requiring conditions as 1. Non-users (patients without issued PPI prescriptions and treatmentrequiring conditions), 2. Possible users (patients without issued PPI prescriptions but with recorded treatment-requiring conditions), and;3. Users (patients with issued PPI prescriptions). Users were further divided into three groups based on the intensity of PPI prescriptions to investigate the dose effect of PPIs. In addition to the comparison of users to non-users, we compared 1. Users to possible users to isolate the effect of PPIs and 2. Possible users to non-users to isolate the effect of comorbidities treated with PPIs on COVID-19-related mortality and hospitalization. Log-binomial regression with robust sandwich variance estimation was used to calculate relative risk and 95% confidence intervals after exact matching in respect to a range of pre-COVID-19 characteristics (in primary analysis: age (binned to 5 years), sex, vaccination status, time period in the course of the pandemic, Charlson comorbidity index, presence of ACE inhibitor therapy and comorbidities atrial fibrillation, autoimmune diseases, cancer, chronic heart failure, chronic obstructive lung disease, ischemic or cerebrovascular diseases, chronic renal disease and immunocompromised state;in sensitivity analysis with an alternative set of covariates). Result(s) Among 433609 COVID-19 patients, 332389 were identified as nonusers, 18170 as possible users, and 55098 as users of PPIs. Users to non-users, users to possible users, and possible users to non-users were matched 48453 to 325005, 41195 to 17334, and 17466 to 316168 subjects per group, respectively. A small difference in COVID-19 related mortality and hospitalizations was observed after matching users to non-users [RRmortality = 1.23 (95%CI 1.16 - 1.30) and RRhospitalization = 1.46 (95%CI 1.38 - 1.54)] and possible users to non-users [RRmortality = 1.24 (95%CI 1.13 to 1.37) and RRhospitalization = 1.26 (95%CI 1.16 - 1.37)]. However, there was no relevant difference between users and possible users in COVID-19-related mortality [RR= 0.93 (95%CI 0.85 - 1.02)] or hospitalizations [RR = 1.04 (0.97 - 1.13)]. Dose effect was not observed in any comparison involving users. Sensitivity analysis yielded comparable results. Conclusion(s) The comparison of possible to non-users, and users to possible users indicates that the risk observed in the comparison of users and non-users of PPI is likely attributable to the burden of comorbidities treated with PPIs and not the effect of the PPIs.