Coping with changes in the Defined Daily Dose in a longitudinal drug consumption database

Objective: To illustrate the best approach to accurate representation of temporal evolution of drug consumption, when confronted with changes over time in the Defined Daily Dose (DDD) measurement unit. Method: As an example, drug consumption (2001–2005) at the University Hospital Rijeka was expressed with the corresponding DDD in each year; and with the DDD value of the last year of the observation time window. By visual appraisal, results with both methods were compared with the graph of drug consumption trends expressed in physical units (mg, IU, etc.). Results: Of 20 drugs for which the DDD value has changed in the 5 year period, 5 were used in Rijeka hospital: erythropoietin parenteral, budesonide nasal spray, repaglinide oral, amoxicillin with clavulanic acid parenteral, and fentanyl transdermal. Graphs for individual drugs with the year by year method showed artificial jerks in the years proceeding the year of DDD change. Representation of total hospital consumption did not differ notably with the two methods. Conclusion: Drug consumption data should be calculated with the latest DDD value in the time window. The policy of the World Health Organisation to keep changes to the system of DDD to a minimum should be continued.