Evaluation of electronic databases used to identify solid oral dosage forms.

The ability of electronic drug identification databases to identify solid oral dosage forms by their imprint codes was studied. The following seven commercially available electronic drug identification databases were selected to identify 500 solid oral dosage forms by their imprint codes: Clinical Pharmacology (Gold Standard Media, Tampa, FL), eFacts (Facts and Comparison, St. Louis, MO), Ident-A-Drug (Therapeutic Research, Stockton, CA), Identidex (Micromedex, Greenwood Village, CO), Clinical Reference Library (Lexi-Comp, Hudson, OH), Physicians' Desk Reference (PDR) Electronic Library (Medical Economics, Montvale, NJ), and RxList (RxList LLC, San Francisco, CA), Chi-square test was used to compare the percentages of medications identified by each of the seven electronic references. The ability of the databases to identify medication by specific characteristics, such as brand name versus generic, prescription versus nonprescription, commercially available for more than one year versus less than one year, colored versus white drug products, and controlled versus noncontrolled substances was evaluated. A logistic regression model was used to determine the probability of a drug product being identified by one of the electronic references based on these characteristics. All seven electronic databases combined identified 95.6% of the unknown medications by imprint code, color, shape, and scoring. Ident-A-Drug and Identidex identified the most drugs. The PDR Electronic Library and Facts and Comparisons Identified the least number of drugs. Solid oral dosage forms more likely to be identified were those that were on the market for more than a year, brand-name products, and prescription medications. Generic products on the market for less than a year and nonprescription products were particularly difficult to identify. A combination of electronic drug identification databases provides the best method of drug identification in an institutional setting.