To verify safety profiles of generic statins in a real-world setting, the risk of liver dysfunction as a common adverse event was compared between generic and brand drugs. A new user cohort design was employed in which patients prescribed one of six statins (atorvastatin, simvastatin, pitavastatin, pravastatin, fluvastatin, or rosuvastatin) in Japan between January 1, 2014 and March 31, 2022 were identified in the MID-NET® database. Adjusted hazard ratios (aHRs) for generic drugs compared with their corresponding brand drug for the occurrence of first liver dysfunction were estimated using high-dimensional propensity score-weighted Cox models. Among the six statins, no increased trends in aHRs were observed in the primary analysis, except for atorvastatin. The primary analysis showed an aHR of 2.08 (95% confidence interval [CI]: 1.20-3.63) for atorvastatin. In an additional analysis with shorter follow-up periods, aHRs for atorvastatin gradually approached 1.00 (1.74 [95% CI: 0.94-3.22] within 360 days, 1.65 [95% CI: 0.84-3.25] within 180 days, 1.49 [95% CI: 0.73-3.01] within 90 days, and 1.33 [95% CI: 0.60-2.96] within 30 days). Results suggest that risks of liver dysfunction by generic statins are similar to those for brand drugs, facilitating our understanding about the safety of generic drugs. The aHR for atorvastatin was inconsistent between the primary and additional analyses, which suggests that the observed increased risk of generic atorvastatin may be affected by other factors and does not necessarily indicate a different safety profile between generic and brand drugs.
Keywords: Database Study; Generic drugs; High dimensional propensity score; Liver dysfunction; Pharmacoepidemiology; Statin.
© 2025. The Author(s).