Black box warning contraindicated comedications: concordance among three major drug interaction screening programs

Abstract

Background: Black box warnings represent the strongest safety warning that the Food and Drug Administration can issue for a marketed prescription drug. Some black box warnings recommend against coadministration of specific medications due to an increased risk for serious, perhaps life-threatening, effects.

Objective: To determine the level of agreement in presence, clinical severity scores level of documentation ratings, and alert content among 3 leading drug interaction screening programs with regard to contraindicated comedications that are mentioned in black box warnings.

Methods: We reviewed the prescribing information for currently marketed prescription drugs with a black box warning that mentioned a contraindicated drug combination. We selected the drug interaction databases Facts & Comparisons 4.0, MICROMEDEX DRUG-REAX, and Lexi-Comp Lexi-Interact to evaluate the interactions. Discrepancies in the inclusion of interactions and level of agreement in clinical severity scores and level of documentation ratings for each interaction were assessed, using descriptive statistics, Spearman's correlation coefficient, Kendall-Stuart tau-c, and Cronbach's alpha.

Results: We identified 11 drugs with black box warnings that contained information on 59 unique contraindicated drug combinations, only 68% of which were covered by any source. Lexi-Comp detected the most interactions (n = 29) and DRUG-REAX the least (n = 18). Only 3 drug combinations were detected and rated as contraindicated or potentially life-threatening in all 3 databases. The severity scores and level of documentation ratings varied widely.

Conclusions: There are discrepancies among major drug interaction screening programs in the inclusion, severity, and level of documentation of contraindicated drug combinations mentioned in black box warnings. Further studies could explore the implications of these inconsistencies, particularly with regard to the integration of black box warning information in clinical practice. Clinicians should consult multiple drug resources to maximize the potential for detecting a potentially severe drug interaction.