Identification and Prevalence of Serious Drug-Drug Interactions in the Ambulatory Care Setting (Part 1)

Presenter: Jacob Abarca, PharmD, MS

Institution: University of Arizona College of Pharmacy

Co-Authors: Daniel C. Malone, Ph.D.; Edward P. Armstrong, Pharm.D.; Amy J. Grizzle, Pharm.D.; Richard B. Lipton, M.D.; Philip D. Hansten, Pharm.D.; Babette Edgar, Pharm.D., M.B.A.

Introduction: The first objective of this study was to evaluate the agreement on the combined listing and rating of ‘major’ drug-drug interaction (DDIs) among commonly used drug interaction compendia. The second objective was to develop a list of the clinically important DDIs that always merit intervention and are likely to be detected by ambulatory pharmacy computer systems.

Methods: First, a systematic review and identification of major drug interactions in tertiary drug interaction references, including Evaluation of Drug Interactions, Drug Interaction Facts, Drug Interactions: Analysis and Management, and MicroMedex (Drug-REAX) system, was conducted. To accommodate the inconsistent rating systems used in DDI compendia, individual criteria were developed to select 'major' interactions from each compendia. A list of all selected 'major' interactions was developed. Medications involved in each DDI were aggregated by medication class when appropriate. For example, DDIs listed at the medication level (e.g., sertraline -- MAO inhibitor, paroxetine -- MAO inhibitor, fluoxetine -- MAO inhibitor) were aggregated to created a more inclusive DDI listing (e.g., SSRI -- MAO inhibitor) Agreement on the combined listing and rating of DDIs across the four compendia was evaluated by calculating the intraclass correlation coefficient (ICC). Second, an expert panel, comprised of 3 pharmacists and 2 physicians, systematically evaluated the evidence supporting each ‘major’ DDI listed in at least three of the four compendia. Each expert panel member was asked to assess each DDI using a 16-item instrument that assessed the severity of the interaction, the evidence supporting the interaction, the probability of an interaction when the medications are co-administered, and the probability of the medications being co-administered in routine clinical practice. Each item used a 10 point scale. During this process, expert panel members could also submit for consideration DDIs that were not listed in three of the four compendia as 'major' interactions, but were considered clinically important. After all the interactions were rated, a modified Delphi-process was used to identify a subset of clinically important DDIs.

Results: The ICC for the combined listing and rating of 'major' DDIs among the four compendia was -0.092. A total of 56 DDIs were rated by the expert panel. Twenty-eight were rated with a clinical significance > 8.0. The range for clinical significance ranged from 3.2-9.6, with a mean (SD) of 7.5±1.5 across the DDIs examined. The mean score for the quality of data suggesting the interaction exists ranged from 1.0-9.6, with a mean of 5.8±2.5. In terms of evidence substantiating the interaction, the mean rating was 6.3±2.2, with range from 1.4-9.2. After completion of the modified Delphi process, a total of 25 interactions were considered to be clinically important. They include:

Discussion: There is little agreement among DDI compendia concerning the severity rating for DDIs. Using an expert panel and a standard evaluation tool, a total of 25 interactions were rated as clinical important and should be included in computerized systems to prevent inadvertent co-administration.