Prescriber perceptions of a near real-time fax alert program for potential drug-drug interactions

Health systems have developed interventions to reduce harm associated with drug-drug interactions. Pharmacy benefit managers are in an important position to identify the coprescribing of medications known to interact, since they process data on a large portion of prescription claims in the United States. Electronic health records and electronic prescribing also include alerts through their systems' clinical decision support. However, limited data are available that assess prescribers' perceptions of processes that screen for potential drug-drug interactions (PDDIs).

To determine prescribers' perceptions of near real-time fax alerts for PDDIs.

This was a 6-month prospective study where a pharmacy benefit manager distributed evidence-based summaries of 18 different PDDIs that included references and suggested management strategies. Fax alerts were individualized letters sent to the prescriber of the second drug of a PDDI pair for an individual patient. A 16-item questionnaire to assess prescribers' perceptions of the intervention accompanied each individualized PDDI evidence-based summary.

A total of 8,075 fax alerts were distributed with 977 returning questionnaires, yielding a 12.1% response rate. There were 848 (86.8%) responses completed by physicians, and 71 (7.3%) completed by nurse practitioners. The most common PDDI fax alerts sent were for warfarin-statin (3,511, 43.5%) and warfarin-thyroid (2,111, 26.1%) interactions. 42.6% of respondents agreed or strongly agreed that fax alerts were a good way to communicate with them. However, 37.5% of respondents either agreed or strongly agreed that the fax alert was a "waste of my time." In contrast, respondents thought notification of carbamazepine-macrolide (mean 1.5 ± 0.71), ciprofloxacin-tizanidine (mean 2.3 ± 1.0), and statin-macrolide (mean 2.3 ± 1.1) was not a waste of time. Also, 59.1% of respondents either disagreed or strongly disagreed that they would prefer to receive a telephone call when interactions like this occur. Half (50.5%) of the respondents indicated their computer systems provided drug interaction alerts. Prescribers who had previously received alerts and specialists were less likely to respond to the questionnaire (OR = 0.685, P ≤ 0.0001 and OR = 0.851, P = 0.0205, respectively).

The positive response to fax alerts by physicians varies by the component drugs of the PDDI alerts.