Observational study of accuracy in compounding i.v. admixtures at five hospitals.

Rates of errors in i.v. admixture compounding at five U.S. hospital pharmacies were studied. Pharmacy staff members at five hospitals representing each U.S. geographic region were observed as they compounded sterile products in order to record the medication, dose, base solution, and other details. Intravenous admixtures, antineoplastic preparations, parenteral nutrient solutions, and ready-to-use products were included. Observations took place for five days at each pharmacy. The observers' notes were checked against the labels used to prepare the doses; any deviation was considered an error. The clinical importance of each error was assessed for its potential to affect a patient adversely. The mean error rate for the five hospitals combined was 9% (145 errors for 1679 doses), excluding ready-to-use products. Mean error rates for individual pharmacies ranged from 6% to 10%. Wrong-dose errors were the most common type of error. Parenteral nutrient solutions had the highest error rates-37% for manual preparation and 22% for preparation that was partly automated. Of every 100 errors, 2 were judged to be potentially clinically important. In five U.S. hospital pharmacies, the observed error rate for compounding i.v. admixtures was 9%.