Physician versus computer knowledge of potential drug interactions in the emergency department

Proliferation of Food and Drug Administration-approved drugs makes it impossible for emergency medicine (EM) faculty to stay current on potential interactions between drugs, and with diseases, laboratory tests, and ethanol. A computer database may augment physician knowledge.

To compare the performance of EM faculty and an "expert" emergency physician (EP) with that of a criterion standard computer database in identifying potential drug interactions, and to report the incidence of drug-ethanol and drug-laboratory test interactions.

This was a retrospective review of 276 emergency department charts for drug, ethanol, lab, and medical history. Evaluation by both EM faculty and an "expert" EP of patient history was done to identify potential interactions, and comparison with the Micromedex Drug-Reax database for potential interactions (graded minor, moderate, or major) was made. Clinical significance of potential interactions was judged by a second EM faculty member.

Seventeen percent of the patients had potential drug-drug interactions, and 25% of these were judged to be clinically significant. Up to 52% of the patients had potential drug-ethanol interactions, while 38% of the patients could have potential drug-lab interactions. Sensitivity, specificity, and positive and negative predictive values of the EM faculty for potential drug-drug interactions compared with the computer were poor, at 14%, 58%, 6%, and 23%, respectively. The corresponding values for the "expert" EP were 25%, 86%, 26%, and 85%. The "expert" EP was statistically better than the EM faculty, but still less sensitive and predictive than the computer.

A computer can aid the physician in avoiding potential drug interactions. Prospective validation of these findings should be done.