Comparison of prospective pharmacy monitoring vs. medical record discharge coding for ADR surveillance


Eric Boyce: 0000-0002-5447-9016

Document Type


Conference Title/Conference Publication

American Society for Clinical Pharmacology & Therapeutics


San Francisco, CA

Conference Dates

March 21-23, 1990

Date of Presentation



Hospitals monitor for ADRs by retrospective reviews (RR) of computerized discharge diagnosis, looking for ADR codes. This study evaluates this approach comparing it to a prospective (PR) ADR program. The PR program monitored ADRs via a clinical pharmacist on 5 medicine teams for 2 months. Suspected ADRs were investigated for severity & causality. PR ADRs were compared to ADRs from RR performed by Medical Records. 39 ADRs were identified by these approaches. Of these, 2 were detected by both approaches, 4 by RR, and 33 by PR. The sens., spec., PPV,NPV, of PR were 89,100,100, & 99%, respectively. The RR results were 15,99,29, & 97%. RR ADRs resulted in hospital admission rather than devloping in hospital. Serious and unexpected ADRs by PR & RR were 89,98,46, & 99%, and 17,98,14, & 99%, respectively. In conclusion, a retrospective discharge-based diagnosis system for collecting hospital ADRs is uacceptable as a sole surveillance system due to low sensitivity.

This document is currently not available here.