Objective: To determine patterns of use of dilatation and curettage in Britain as compared with those in the United States; to examine variations in utilisation rates within one regional health authority.

Design: Analysis of routinely collected hospital inpatient statistics.

Setting: Statistics for England, Scotland, and the United States; local statistics for Oxford region.

Subjects: All inpatient episodes in which dilatation and curettage was performed but excluding those related to pregnancy.

Results: Dilatation and curettage rates remained stable in Britain between 1977 and 1990, whereas in the United States they declined dramatically. In 1989-90 the rate was 71.1 per 10,000 women in England as compared with only 10.8 per 10,000 in America. In 1989, 6936 women underwent diagnostic dilatation and curettage in the Oxford region, making it the most common elective operation. A total of 2726 (39%) of these women were under 40. There was a more than twofold variation in usage of the procedure among district health authorities within the region and even greater variation in rates in women under 40. The proportion of patients treated as day cases in the district general hospitals ranged from 22% to 82%.

Conclusions: Dilatation and curettage may frequently be used inappropriately. The considerable variations in usage of dilatation and curettage internationally and nationally indicate differences in clinical perception of its appropriateness. This makes it suitable for audit. In developing guidelines it will be important to agree on the most appropriate patients and the relative merits of alternative methods of endometrial sampling. Probably this could result in considerable cost savings at no risk and possibly some benefit to patients.