[Detectig errors in chemotherapy prescription].

Abstract

OBJECTIVE: To analyse errors detected in the prescription of cytostatic drugs. MATERIAL AND METHODS: Prospective study (February 1st to April 15th, 2002). All medical orders containing cytostatic agents were checked and 13 variables were studied: identification, anthropometric data, diagnosis, protocol, number of cycles, cytostatic drug, length of treatment, dose, route of administration, volume of infusion and/or final concentration, infusion time, diluent, and doctor's signature. Several possible errors were identified for each variable. Information on whether it was the first cycle, prescribing service, and prescription format was also recorded. RESULTS: In all, 618 medical orders were reviewed for a total of 1178 lines of cytostatic agents and 2,171 doses. The possible number of errors was 12.101 and the total number of errors actually found was 2,706 (22,03%). Not all these errors had the same impact on patients. Errors by omission were 2,340 (87,77%). Those which nursing staff found difficult to check and/or administer stood at 281 (10,54%). Potentially serious errors numbered 60 (2.06%): wrong body area > 10%: 5; wrong body area 5-10%: 2; erroneus protocol: 2; incorrect volume: 16; wrong dosage: 23 (difference < 25%: 5; difference 10-25%: 14; difference < 10%: 4). Reductions indicated but not implemented: 9; Inappropiate diluent: 3. There were no incorrect or omitted medications. CONCLUSION: The description of such errors and their uantification is a useful method of quality asurance in order to establish appropiate corrective measures. The most common error was the omission of information and therefore this issue, along with the development of computerised prescriptions, should be addressed.