Patient Profiling and Provider Feedback to Reduce Adverse Drug Events
Peter A. Glassman MBBS MSc
VA Greater Los Angeles Health Care System
West Los Angeles, CA
Funding Period: January 2000 - December 2003

BACKGROUND/RATIONALE:
Adverse drug events (ADE) present a unique focus for error reduction. Computerized provider order entry, with embedded clinical decision support, has great promise in reducing medication errors but preventable adverse drug events may still occur despite such systems.

OBJECTIVE(S):
The purpose of the study was to evaluate whether adding medication profiling (by using a retrospective drug utilization review program) to computerized provider order entry with embedded order checks (drug alerts) reduces the incidence of adverse drug events.

METHODS:
Medication profiles mainly focused on possible drug-drug and drug-disease interactions, with some drug duplications. To do the medication profiles we licensed a proprietary computerized retrospective drug utilization review system. We randomly assigned over 900 patients to Usual Care or Provider Feedback. For patients in the latter group, selected providers were contacted by letter with pertinent information; electronic mail was used for follow-up contact. Clinical and other relevant data was retrospectively abstracted from the medical records for up to one year from the last medication profile for all patients. This was done by a pharmacist reviewer, using a study-derived instrument, and blinded to patient assignment. ADE incidence is the primary outcome of interest, with other outcomes such as ADE severity and preventability also assessed. We also developed and implemented provider surveys in pre- and post-profiling periods.

FINDINGS/RESULTS:
Currently, we are in the process of analyzing and submitting substantive findings from the profiling aspect of the study. In the surveys we noted, among other issues, that recognition of common selected drug interactions was not always comprehensive and that knowledge of those selected interacting drug pairs did not substantially change over time with exposure to order checks within provider order entry. Providers appeared to value order checks (alerts) but signal to noise ratio (e.g., non-relevant warnings) was considered a substantive barrier to utilization of order checks (alerts) in both time periods.

IMPACT:
We are assessing whether combining computerized provider order entry with medication profiling (by retrospective drug utilization review) impacts medication safety in terms of reducing incidence, preventability and severity of adverse drug events.

PUBLICATIONS:

Journal Articles

1. Glassman PA, Belperio P, Simon B, Lanto A, Lee M. Exposure to automated drug alerts over time: effects on clinicians' knowledge and perceptions. Medical Care. 2006; 44(3): 250-6.
2. Glassman PA, Simon B, Belperio P, Lanto A. Improving recognition of drug interactions: benefits and barriers to using automated drug alerts. Medical Care. 2002; 40(12): 1161-71.

DRA: Health Services and Systems
DRE: Prevention, Quality of Care
Keywords: Adverse events, Education (provider), Safety
MeSH Terms: Ambulatory Care, Databases, Health Care Costs, Behavior, Behavior and Behavior Mechanisms, Controlled Clinical Trials, Decision Support Techniques, Delivery of Health Care, Integrated, Drug Therapy, Safety