Impact of a pharmacotherapy alerting system on medication errors.

PURPOSE

An evaluation of a rules-based pharmacotherapy alerting system configured to identify improperly verified new medication orders in an inpatient setting is described.

METHODS

A retrospective pre-post cohort study was conducted to assess order-verification alerts and pharmacy interventions at a 900-bed hospital before and after implementation of a commercial pharmacotherapy alerting system. In the preintervention phase of the study, the pharmacotherapy alerting system was used on a limited basis, with clinical pharmacists responding to all alerts and the resulting data used to refine the trigger rules; for the intervention phase, the pharmacotherapy alerting system was programmed to alert only on order-verification errors involving four medications (darbepoetin, filgrastim, fondaparinux, and warfarin). In the event of alerts, a pharmacy response team provided nearly real-time feedback to the order-verification pharmacist, mainly via e-mail or paging.

RESULTS

From the preintervention period to the intervention period, there was a 36% decrease in the frequency of order-verification alerts (p = 0.035), and the average number of alerts per day declined from 1.0 to 0.6, suggesting that the pharmacotherapy alerting system and associated oversight mechanisms were effective in enabling pharmacy staff to prevent future errors at the order-verification step before such errors could result in patient harm. The review team spent an average of 10.2 minutes carrying out interventions in response to alerts during the intervention phase.

CONCLUSION

Incorporation of a real-time pharmacotherapy alerting system with an oversight response process reduced the number of pharmacotherapy alerts and facilitated interception and prevention of adverse drug events.