Purpose: Few studies have reported an association between proton pump inhibitor (PPI) use and tuberculosis. Tuberculosis incidence is relatively high in Asian people, and an increase in PPI prescriptions has been reported in South Korea. Thus, we investigated the association between PPI use and tuberculosis development.
Methods: We conducted a case-control study on 25 672 newly diagnosed tuberculosis patients using the National Health Insurance Service-National Sample Cohort (NHIS-NSC) database (2002-2013). We selected a control group without tuberculosis using 1:1 exact matching based on age, sex, index year, insurance type, and income level. We investigated PPI exposure 2 years prior to the index date and classified the subjects into nonuser, continuous user, recent user, and former user groups. Odds ratios (ORs) and 95% confidence intervals (CIs) for tuberculosis development were calculated using conditional logistic regression.
Results: A total of 51 344 cases and controls were analyzed. Recent PPI use (adjusted odds ratio [aOR], 1.28; 95% CI, 1.18-1.39) and continuous PPI use (aOR, 1.13; 95% CI 1.10-1.28) were significantly associated with tuberculosis development, compared with nonuse of PPIs. An increased tuberculosis incidence was not observed in the former use group compared with the nonuse group (aOR 1.05, 95% CI 0.95-1.17).
Conclusions: In this case-control study, we found that recent PPI use and continuous PPI use were associated with increased tuberculosis development. Although further investigation is needed, the tuberculosis risk accompanying PPI treatment should be considered.
Keywords: association; case-control study; pharmacoepidemiology; proton pump inhibitor; tuberculosis.
© 2019 John Wiley & Sons, Ltd.