Comparison of alphabetic and phonetic retrieval of online drug information

Alphabetic, phonetic, and combined alphabetic and phonetic methods of retrieving online drug information were compared. Twenty-four volunteers participated in the study representing four user groups: physicians, nurses, pharmacists, and nonhealth-care hospital staff. Each subject performed 150 searches, 50 by each retrieval method. Using the alphabetic method, drug information was retrievable only if the drug name was spelled correctly. Using the phonetic method, searches were conducted based on the phonetic spelling of requests (e.g., "symetadine" for cimetidine). The combined method used a phonetic search only after an initial alphabetic search was unsuccessful. The elapsed time between the first entry and an indication that the information had been found or could not be found was determined, and the number of drug names not found and the number of excess tries were counted. There were no significant differences in elapsed time among the three methods. Pharmacists had the shortest mean elapsed time and physicians the longest. The average number of excess tries using the phonetic system was a third of the number required using the alphabetic method. The number of drugs not found showed only slight differences among the three methods. The subjects found the desired information on the first try 67% of the time with the alphabetic method, 66% with the combined method, and 90% with the phonetic method. The phonetic method had an average of 75 matches versus 20 for the alphabetic and combined methods. These results support use of a combined alphabetic and phonetic system for retrieving drug information.