Incidence of drug-induced liver injury in medical inpatients.

Objectives: Drug-induced liver injury (DILI) is a common concern. However, data on DILI epidemiology in inpatients are sparse.

Methods: To investigate the incidence of DILI, we screened all patients in the pharmacoepidemiological inpatient database according to the CIOMS (Council for International Organisation of Medical Science) criteria, which consist of the evaluation of some clinical chemistry laboratory liver parameters (CIOMS laboratory criteria) and the exclusion of any disease-related causes for the liver injury. Thus, only cases with probable or certain causality according to the World Health Organization criteria were included.

Results: Among a total of 6383 patients, liver parameters were determined in 4610, and 489 among them fulfilled the CIOMS laboratory criteria. However, 401 patients had to be excluded because of disease-related liver injury and, thus, the study cohort consisted of 4209 patients at risk for DILI. Among a total of 88 DILI cases, 31 had no documented normal baseline liver parameters and, thus, represented prevalent cases. The remaining 57 represented incident DILI cases. Thus, the incidence of DILI was 1.4% (95% CI 1.0, 1.7). The drug classes most frequently causing DILI were heparins, antibacterials, tuberculostatics and antineoplastic agents. Among those, antineoplastic agents and tuberculostatics showed the highest incidence. Liver injury was not mentioned among the diagnoses or in the physician's discharge letter in about 52-68% of all cases.

Conclusion: Approximately 1 in 100 patients develops DILI during hospitalisation in a department of medicine. Incidences of DILI were highest for antineoplastic agents and tuberculostatics. DILI is frequently missed and, therefore, DILI detection by diagnoses will result in misleadingly low incidence rates.