Background Little agreement exits between different drug interaction databases.

Purpose To compare the frequency and severity of potential drug-drug interactions (DDIs) occurring in a haematological unit and detected by two drug interactions databases.

Materials and methods A prospective, observational and descriptive study was carried out from November 2012 to February 2013. Twice a week, every patient’s treatment sheet was collected and screened through two drug interactions databases: Thomson Micromedex and Drug Interaction Facts. All potential DDIs identified were recorded and graded by their level of severity.

Results Among 317 analysed treatment sheets, a total of 2373 potential DDIs were detected by the two databases. According to Micromedex, 1348 potential DDIs were found, counting 176 different pairs of drugs; of these DDIs, 64 were classified as contraindicated, 538 as major, 718 as moderate and 28 as minor. Regarding Drug Interaction Facts, 1025 potential DDIs were found, counting 124 different pairs of drugs; of these DDIs, 203 were classified as major, 537 as moderate and 285 as minor. There was a pool of 225 different pairs of drugs detected by both databases, irrespective of how many times these interactions appeared. Upon assessing the total number of pairs of drugs identified by the two databases, Micromedex identified 78.2% (176/225) and Drug Interaction Facts, 55.1% (124/225) of the potential interactions. Upon evaluation of the congruence of severity ratings between both databases, there was an agreement in 16.4% of the 225 pairs of drugs identified (37/225).

Conclusions The lack of agreement between different databases shows how complicated it is to detect potentially significant drug interactions in clinical practice.

No conflict of interest.