Determining the accuracy of open-access databases for identifying commonly prescribed oral medications

To compare the accuracy of open-access medication identification databases in identifying commonly prescribed oral tablets and capsules.

We compiled a dataset from the most frequently prescribed medications in 2012 and 2013 by total prescription volume. Imprints for the medications were obtained from the Food and Drug Administration label information accessed via the National Library of Medicine's (NLM) DailyMed. The imprints were used to determine if the correct medication could be identified in the open-access medication identification databases Drugs.com, Healthline, NLM's Pillbox, RxList, and WebMD by means of the imprints alone. A general web search using the imprints as key words via Google search engine was also conducted to mimic the lay public's approach to identifying an unknown medication.

Database accuracy of search results ranged from 26.5% to 89.27% with the NLM Pillbox having the highest accuracy when using tablet/capsule imprint information alone. The secondary web search with the use of the Google search engine yielded 75.7% accuracy in using imprints as search terms to identify websites leading to the appropriate drugs. Drugs.com was listed as the first site for the majority of Google searches and was 86.44% accurate.

Open-access databases are available to help identify oral tablets and capsules and they have a relatively high success rate of identifying commonly prescribed medications. However no database is 100% accurate, which means that the risk involved with patients identifying their own medications is substantial. Patients should be highly encouraged by all health care practitioners to consult their pharmacist or primary health care provider to prevent misidentification and subsequent misadministration.