Trends in drug use in intensive care units in Japan: determinants of drug utilization under an evidence-based approach to medicine.

Objective: To clarify actual drug expenditures used in ICUs, and to evaluate the relationship between these expenditures and therapeutic evidence. Test Sites: Ten university hospitals located in six different regions of Japan (Tohoku, Kanto, Chuubu, Kinki, Chuugoku and Kyuushuu areas).

Design: Cross-sectional surveys of drug expenditures in ICUs throughout Japan.

Method: The 10 leading pharmaceutical products, based on drug expenditures per month, were investigated. The ICUs at 10 facilities (national, public, and private university hospitals) were randomly selected from the attendees of the annual university training seminar. A postal questionnaire was sent to ten university hospitals located in six different regions (Tohoku, Kanto, Chuubu, Kinki, Chuugoku, and Kyuushuu areas) of Japan. The level of evidence was investigated to determine whether the use of the most expensive drug is rational or not.

Results: The response rate was 80% (8/10). Pharmacists in each hospital filled in the questionnaire to determine drug expenditure per ICU bed. The consumption of protease inhibitors based on drug expenditure per bed ranked the highest out of 14 different drugs used in the ICUs studied. The average drug expenditures ranked as follows: protease inhibitors > antithrombin III (AT III) > albumin-derived products > intravenous immunoglobulin > human haptoglobin. There was a 100-fold difference in the expenditures on protease inhibitor drugs between Nagoya university hospital and the hospital which consumed the most. It has been reported that there is little evidence of the effectiveness of protease inhibitors and AT-III for critically ill patients in the ICU.

Conclusion: We analyzed the data from a questionnaire about drug expenditures in ICUs, responded to by 8 Japanese university hospitals. Although protease inhibitors accounted for a large percentage of all drug expenditures in the ICUs, the evidence supporting their usage was insufficient. Therapeutic evidence is important for determining physicians' prescriptions.