A prospective nationwide study of drug-induced liver injury in Korea.

OBJECTIVES: To address a growing concern about drug-induced liver injury (DILI), a nationwide study was performed to investigate the significance of DILI in Korea. METHODS: From May 2005 to May 2007, cases of DILI (alanine transferase > 3 × upper normal limit or total bilirubin > 2 × upper normal limit) from 17 referral university hospitals were prospectively enrolled. Adjudication by the seven review boards was considered for the confirmation of causality and the Roussel Uclaf Causality Assessment Method (RUCAM) scale was used. RESULTS: A total of 371 cases were diagnosed with DILI. The extrapolated incidence of hospitalization at university hospital in Korea was 12/100,000 persons/year. The causes included "herbal medications" (102, 27.5%), "prescription or non-prescription medications" (101, 27.3%), "health foods or dietary supplements" (51, 13.7%), "medicinal herbs or plants" (35, 9.4%), "folk remedies" (32, 8.6%), "combined" (30, 8.2%), "herbal preparations" (12, 3.2%), and others (8, 2.2%). Nine cases were linked to acetaminophen. The frequencies of hepatocellular, mixed, and cholestatic types were 76.3, 14.8, and 8.9%, respectively. A total of 234 cases met the criteria for Hy's law. Five patients died or underwent transplantation. Twenty-five cases (21 herbs and 4 medications) did not meet the time-to-onset criteria of the RUCAM. CONCLUSIONS: DILI appears to be a highly relevant health problem in Korea. "Herbal medications" are the principal cause of DILI. A more objective and reproducible causality assessment tool is strongly desired as the RUCAM scale frequently undercounts the cases caused by herbs owing to a lack of previous information and incompatible time criteria.