A detection algorithm for drug‐induced liver injury in medical information databases using the Japanese diagnostic scale and its comparison with the Council for International Organizations of Medical Sciences/the Roussel Uclaf Causality Assessment Method scale. (5th March 2014)

• Record Type: Journal Article
• Authors: Hanatani, Tadaaki; Sai, Kimie; Tohkin, Masahiro; Segawa, Katsunori; Kimura, Michio; Hori, Katsuhito; Kawakami, Junichi; Saito, Yoshiro
• Is Part Of: Pharmacoepidemiology and drug safety. Volume 23:Number 9(2014:Sep.)
• Page Start: 984
• Page End: 988
• Publication Date: 2014-03-05
• Subjects: Pharmacoepidemiology -- Periodicals; Chemotherapy -- Periodicals; Epidemiology -- Periodicals; 615.705
• DOI: 10.1002/pds.3603
• Languages: English
• ISSNs: 1053-8569

Abstract

<abstract abstract-type="main"> <title>ABSTRACT</title> <sec id="pds3603-sec-0001" sec-type="section"> <title>Purpose</title> <p>Drug‐induced liver injury (DILI) is one of the primary targets for pharmacovigilance using medical information databases (MIDs). Because of diagnostic complexity, a standardized method for identifying DILI using MIDs has not yet been established. We applied the Digestive Disease Week Japan 2004 (DDW‐J) scale, a Japanese clinical diagnostic criteria for DILI, to a DILI detection algorithm, and compared it with the Council for International Organizations of Medical Sciences/the Roussel Uclaf Causality Assessment Method (CIOMS/RUCAM) scale to confirm its consistency. Characteristics of DILI cases identified by the DDW‐J algorithm were examined in two Japanese MIDs.</p> </sec> <sec id="pds3603-sec-0002" sec-type="section"> <title>Methods</title> <p>Using an MID from the Hamamatsu University Hospital, we constructed a DILI detection algorithm on the basis of the DDW‐J scale. We then compared the findings between the DDW‐J and CIOMS/RUCAM scales. We examined the characteristics of DILI after antibiotic treatment in the Hamamatsu population and a second population that included data from 124 hospitals, which was derived from an MID from the Medical Data Vision Co., Ltd. We performed a multivariate logistic regression analysis to assess the possible DILI risk factors.</p> </sec> <sec id="pds3603-sec-0003" sec-type="section"> <title>Results</title> <p>The concordance rate was 79.4% between DILI patients identified by the DDW‐J and CIOMS/RUCAM; the Spearman rank correlation coefficient was 0.952 (<italic>P</italic> < 0.0001). Men showed a significantly higher risk for DILI after antibiotic treatments in both MID populations.</p> </sec> <sec id="pds3603-sec-0004" sec-type="section"> <title>Conclusions</title> <p>The DDW‐J and CIOMS/RUCAM algorithms were equivalent for identifying the DILI cases, confirming the utility of our DILI detection method using MIDs. This study provides evidence supporting the use of MID analyses to improve pharmacovigilance. Copyright © 2014 John Wiley & Sons, Ltd.</p> </sec> </abstract>