Prescribing at the Primary Care Group level: census data and prescribing indicators

Objective: To examine the relationship between prescribing and morbidity, mortality and deprivation for the 481 Primary Care groups set up on 1 April 1999, by examining the prescribing of their constituent practices in the year 1998/99, i.e. the year before. Design: Cross-sectional study. Outcome measures: Set of 11 prescribing indicators taken from the Prescription Pricing Authority 'Toolkit' system; census measures of morbidity and deprivation and Standardized Mortality Ratio for those aged 75 and under. Setting: All practices in England belonging to a Primary Care Group. Results: Several of the indicators showed strong correlation with morbidity, mortality and, to a weaker extent, with deprivation. There was a negative correlation between the census-based measures and choice of more expensive alternatives or greater duration prescriptions for antibiotic prescribing and with the use of hormone replacement therapy (HRT). Conversely, there was a positive correlation with use of premium products and drugs of limited therapeutic value. Conclusion: Practices in areas of greater need (as measured by permanent sickness and Standardized Mortality Ratio) seem to be trying to hold down costs by the way they prescribe antibiotics. Conversely, their use of premium price products pushes up their costs in the absence of clear evidence concerning compliance. The use of HRT is low in areas of high overall need.