Prevalence of Clinically Important Drug-Drug Interactions in the Ambulatory Setting and the Impact of a Systems Approach to Improving Patient Safety: Evidence from a Large Nationwide Pharmaceutical Claims Database

Presenter: David S. Hutchins, MBA, MHSA

Institution: AdvancePCS, 9501 East Shea Boulevard, Scottsdale AZ, 85260-6719

Co-Authors: Heather Haupert, MS (1); Daniel C. Malone, RPh, PhD (2); Philip Hansten, PharmD (3); Babette Duncan, Pharm.D (1); Robin C. Van Bergen, BS (1); Steven L. Solomon, MD (4); Richard B. Lipton, MD (5)

Introduction: Technological advances simplify identifying and preventing potentially harmful, even lethal, drug-drug interaction (DDI) combinations. In an earlier study, we systematically reviewed numerous potentially harmful DDIs, likely to occur in an outpatient setting, selecting 25 more clinically important DDIs for further analysis. Our purpose in this study was to quantify the frequency with which these interactions occurred, warnings were generated, and claims aborted in an ambulatory setting.

Methods: A retrospective analysis was performed on prescription claims submitted between April 2000 and June 2002 for about 45 million individuals and maintained in a longitudinal data base. A DDI case was counted when the object and precipitant medications were dispensed within 30-days of each other to the same individual during a 25-month interval. Prevalence of each DDI divided the number of cases by the total membership. Case-exposure rates divided the number of DDI cases by the number of individuals exposed to one drug in a DDI pair.

Results: Prevalence was highest for the nonsteroidal anti-inflammatory drugs (NSAID)/warfarin DDI (prevalence:279/100,000 members), with a case-exposure rate of 243/1000 for warfarin recipients. Prevalence was lowest for the cyclosporine/rifamycins (0.10/100,000) while the case-exposure rate was lowest for azoles/pimozide (0.028/1000 for azole recipients). Prevalence and case-exposure rates varied by sex and generally increased with age. Among the drugs receiving the highest proportion of warnings per paid claim were cyclosporine (61%), warfarin/anticoagulants (41%), and methotrexate (40%). Propoxyphene had the lowest percentage of warnings (0.02%), although 13 other drugs had warnings for less than 1% of paid claims. The percent of aborted claims ranged from 20% to 46%.

Conclusion: Current DUR efforts prevent some potential DDIs, but serious DDI combinations including warfarin/NSAID, warfarin/thyroid hormones, azoles/benzodiazepines, and quinolones/theophylline still occur frequently enough, especially in the elderly, to warrant additional efforts to further reduce their potential threat to public safety.