Consultant evaluation of a hospital medication system: analysis of the existing system.

A consultant team's evaluation of a system for distributing and controlling medications in a large teaching hospital is described. Through interviews with key personnel from administration, pharmacy, nursing, and the medical staff, an interdisciplinary research group identified problems in the reliability and response times of the hospital's existing medication system. After assessing staff expectations regarding acceptable standards for medication errors and response times and their attitudes toward proposed changes in the medication system, medication-error rates were determined using a pharmacist-observer method. Observations during 34 five-hour periods on four nursing units were conducted over a 17-day period. Medication-error rates were calculated as the frequency of medication errors during the observation period divided by the total opportunities for error (OE), which were defined as doses ordered plus unauthorized doses given. Response times for processing "now," "stat," and routine orders were also determined using work-sampling methods. The total medication-error rate for the nursing units studied was 9% excluding wrong-time errors; more than a third of doses were given more than 30 minutes before or after their scheduled administration times. Response times for "now" and "stat" orders averaged about 23 minutes, in conformance with the desired standard of 30 minutes. However, processing of routine orders required an average of two hours and seven minutes, much of which was attributed to delays in the messenger service. The basic design of the existing unit dose medication system contributed to problems in the reliability and efficiency of the system.