Secondary Logo

Journal Logo

Institutional members access full text with Ovid®

A controlled trial of smart infusion pumps to improve medication safety in critically ill patients*

Rothschild, Jeffrey M. MD, MPH; Keohane, Carol A. BSN, RN; Cook, E Francis ScD; Orav, E John PhD; Burdick, Elisabeth MS; Thompson, Sarah RN; Hayes, Judy RN; Bates, David W. MD, MSc

doi: 10.1097/01.CCM.0000155912.73313.CD
Clinical Investigations

Objective: Intravenous medications are vital during inpatient management. Errors associated with the administration of medications through intravenous infusion pumps to critically ill patients can result in adverse drug events. We sought to assess the impact of smart pumps with integrated decision support software on the incidence and nature of medication errors and adverse drug events.

Design: We performed a prospective, randomized time-series trial and compared the serious medication error rate between intervention (decision support on) and control (decision support off) periods. Serious medication errors included both near-misses and preventable adverse drug events. Pump software produced log reports to help identify potential events. Events were presented to physicians for rating of event type, preventability, and severity.

Setting: Cardiac surgical intensive care and step-down units between February and December 2002.

Patients: Pump data were available for 744 cardiac surgery admissions.

Interventions: Decision support during medication administration provided feedback including alerts, reminders, and unit-specific drug rate limits.

Measurements and Main Results: We found a total of 180 serious medication errors, including 14 and 11 preventable adverse drug events and 73 and 82 nonintercepted potential adverse drug events in the control and intervention periods, respectively. The serious medication error rates in the control and intervention periods were 2.03 and 2.41 per 100 patient-pump-days, respectively (p = .124). We also found numerous opportunities for safety improvement. Violations of infusion practice during the intervention periods included 571 (25%) bypasses of the drug library. Medications were also frequently administered without documentation of physician orders in both periods (n = 823; 7.7%).

Conclusion: Intravenous medication errors and adverse drug events were frequent and could be detected using smart pumps. We found no measurable impact on the serious medication error rate, likely in part due to poor compliance. Although smart pumps have great promise, technological and nursing behavioral factors must be addressed if these pumps are to achieve their potential for improving medication safety.

From the Division of General Medicine and Department of Medicine (JMR, CAK, EFC, EJO, EB, DWB), Brigham and Women's Hospital, Harvard Medical School (JMR, DWB), Harvard School of Public Health (EFC, EJO), Department of Nursing, Brigham and Women's Hospital (ST, JH), Boston, MA.

Supported, in part, by AHRQ grant PO1 HHS11534–01 and ALARIS Medical Systems.

Presented, in part, at the Annual Symposium of the American Medical Informatics Association, Washington DC, November 2003.

Address requests for reprints to: Jeffrey M. Rothschild, MD, MPH, Division of General Medicine, Brigham and Women's Hospital, 1620 Tremont Street, Boston, Massachusetts 02120-1613. E-mail:

© 2005 by the Society of Critical Care Medicine and Lippincott Williams & Wilkins