Determinants of prescribing costs amongst single-handed general practitioners

Objective. To investigate the contribution of patient and doctor characteri stics in explaining observed variations in prescribing costs between indivi dual doctors. Method: Secondary analysis of data collected from general practitioners, Fa mily Health Services Authorities, 1991 Census data set and the Prescription Pricing Authority. Results: A multiple regression model with four variables (social class, tra ining status, generic prescribing and length of time in general practice) e xplained only 16.5% of the variation in costs/ASTRO-PU. Conclusion: This study highlights that very little of the variation in pres cribing costs can readily be explained. Further research is needed to docum ent contributing factors.