Change in ranking order of prescribing patterns by age and sex standardization of the practice population - audit may be misleading

Abstract

Objective — To demonstrate whether standardization of practice populations by age and sex changes the internal prescription ranking order of a group of practices.

Design — Data on the prescribing of cardiovascular drugs in a group of practices were obtained from a county-based database. Information on the age, sex, and numbers of patients per practice was also obtained. The direct standardization method was used to adjust practice populations for age and sex.

Setting — The town of Randers, Aarhus County, Denmark.

Subjects — 35 practices, 41 GPs.

Main outcome measures — Ranking of the 35 practices with respect to number of prescribed daily doses of cardiovascular drugs in crude utilization index and standardized utilization index. The slope of the regression curve of variation between the practices.

Results — The change from crude to standardized utilization index gave an entirely new ranking order of the practices. Only four practices did not change ranking position, while four moved more than ten places. The slope between highest and lowest ranked practice did not diminish after standardization.

Conclusion — Care should be taken when comparing peer prescribing patterns from crude utilization data, and we recommend prior age and sex standardization. Our results are of value in creating a more reliable basis for a debate on variation in prescribing patterns between peers.

Key Words:

• demography
• Denmark
• epidemiology
• family practice
• peer review
• physician's practice patterns
• prescriptions
• drug
• medical audit
• standardization