Erschienen in:
15.11.2023 | Original Article
Proton Pump Inhibitors Increases Longitudinal Risk of Mortality, Decompensation, and Infection in Cirrhosis: A Meta-Analysis
verfasst von:
Zhen Yu Wong, Jia Hong Koh, Mark Muthiah, Benjamin Koh, Elden Yen Hng Ong, Christen En Ya Ong, Kai Qi Ou, Wen Hui Lim, Darren Jun Hao Tan, Douglas Chee, Kewin Tien Ho Siah, Yujun Wong, Apichat Kaewdech, Karn Wijarnpreecha, Anand V. Kulkarni, Benjamin Nah, Daniel Q. Huang, Mazen Noureddin, Cheng Han Ng, Margaret Teng
Erschienen in:
Digestive Diseases and Sciences
|
Ausgabe 1/2024
Einloggen, um Zugang zu erhalten
Abstract
Background/Aims
Proton pump inhibitors (PPIs) are frequently prescribed to cirrhotic patients, but there is limited longitudinal evidence regarding their effects. This study aimed to assess the impact of PPIs on adverse events in cirrhotic patients.
Methods
A comprehensive search was conducted using the Medline and Embase databases to identify relevant articles. Pooled hazard ratios (HRs) using DerSimonian and Laird random-effects model were calculated to evaluate the risk of adverse events such as long-term mortality, hepatic decompensation, hepatic encephalopathy (HE), spontaneous bacterial peritonitis (SBP), and overall infection in cirrhotic patients with PPI use.
Results
The analysis included 28 studies with 260,854 cirrhotic patients. The prevalence of PPI use among cirrhotic patients was 55.93%. The use of PPIs was not significantly associated with short-term mortality in cirrhotic patients. However, long-term mortality (HR 1.321, 95% CI 1.103–1.581, P = 0.002), decompensation (HR 1.646, 95% CI 1.477–1.835, P < 0.001), HE (HR 1.968, 95% CI 1.372–2.822, P < 0.001), SBP (HR 1.751, 95% CI 1.649–1.859, P < 0.001), and infection (HR 1.370, 95% CI 1.148–1.634, P < 0.001) were significantly associated with PPI use. Sensitivity analysis with prospective studies yielded similar results.
Conclusion
PPIs should be reserved for appropriate indications at lowest effective dose for cirrhotic patients due to the potential harm.