Relative prevalence of 10 types of pharmacodynamic interactions in psychiatric treatment

To assess the relative prevalence and factors affecting the prescription of medication combinations with a theoretical efficacy limiting pharmacodynamic interaction, defined as two medications with opposing indications and side effects or antagonistic action at the primary receptor of mechanism of action.

One hundred sixteen combinations were identified for 10 types of pharmacodynamic interactions. PubMed was searched for each combination to assess the quality of evidence either supporting clinical use or verifying reduced efficacy. Micromedex was searched to determine the presence of warnings to prescribers of reduced efficacy. The prevalence in clinical practice was determined by computer review of the Genoa Healthcare database for all prescribers at 10 participating community mental health centers. The expected prevalence was calculated as the product of the probability of each medication prescribed alone and was compared with the actual prevalence of the combination using the test of proportions.

The frequency of prescription of eight combinations met the Bonferroni corrected level of significance of p < 0.001. Four were combinations of amphetamine and D2 antagonists and each were prescribed less often than chance, p = 0.0001 consistent with epidemiological studies and multiple animal studies verifying an efficacy limiting interaction. Despite epidemiological studies indicating increased risk of accidents, alprazolam and amphetamine were prescribed more often than chance, p = 0.0001. Micromedex generated warnings for efficacy limiting interactions for five other combinations, but with no subsequent change in prescription frequency.

Neither presence of medical evidence nor warnings from Micromedex consistently affect the prescription of combinations with pharmacodynamic efficacy limiting interactions.