A COMPARISON OF 3 METHODS OF SETTING PRESCRIBING BUDGETS, USING DATA DERIVED FROM DEFINED DAILY DOSE ANALYSES OF HISTORIC PATTERNS OF USE

Background. Prescribing matters (particularly budget setting and resea rch into prescribing variation between doctors) have been handicapped by the absence of credible measures of the volume of drugs prescribed. Aim. To use the defined daily dose (DDD) method to study variation in the volume and cost of drugs prescribed across the seven main British National Formulary (BNF) chapters with a view to comparing different methods of setting prescribing budgets. Method. Study of one year of p rescribing statistics from all 129 general practices in Lothian, cover ing 808 059 patients: analyses of prescribing statistics for 1995 to d efine volume and cost/volume of prescribing for one year for 10 groups of practices defined by the age and deprivation status of their patie nts, for seven BNF chapters; creation of prescribing budgets for 1996 for each individual practice based on the use of target volume and cos t statistics; comparison of 1996 DDD-based budgets with those set usin g the conventional historical approach; and comparison of DDD-based bu dgets with budgets set using a capitation-based formula derived from l ocal cost/patient information. Results. The volume of drugs prescribed was affected by the age structure of the practices in BNF Chapters 1 (gastrointestinal), 2 (cardiovascular), and 6 (endocrine), and by depr ivation structure for BNF Chapters 3 (respiratory) and 4 (central nerv ous system). Costs per DDD in the major BNF chapters were largely inde pendent of age, deprivation structure, or fundholding status. Capitati on and DDD-based budgets were similar to each other, but both differed substantially from historic budgets. One practice in seven gained or lost more than pound 100 000 per annum using DDD or capitation budgets compared with historic budgets. The DDD-based budget, but not the cap itation-based budget, can be used to set volume-specific prescribing t argets. Conclusions. DDD-based and capitation-based prescribing budget s can be set using a simple explanatory model and generalizable method s. In this study, both differed substantially from historic budgets. D DD budgets could be created to accommodate new prescribing strategies and raised or lowered to reflect local intentions to alter overall pre scribing volume or cost targets. We recommend that future work on sett ing budgets and researching prescribing variations should be based on DDD statistics.