Abstract

AIMS/BACKGROUND: Geographical variations in health care are common. There is, however, no simple explanation for why they arise. Variations in rates of ophthalmic surgery in the population aged 65 and over were investigated, with the aim of determining their cause. METHODS: Routine data sources were used to obtain the 1991-2 age and sex standardised rates for English health districts with an ophthalmic unit. Weighted least squares regression was used to study the relation between these rates and various factors describing the population and the provision of care. RESULTS: Surgery rates varied more than threefold. High rates of surgery were associated with high throughput and bed numbers (both p < 0.001), high proportions of day case surgery (p < 0.001), long waiting lists (p < 0.001), and a high number of theatre sessions (p = 0.002). Conversely, a high percentage of emergency admissions was associated with lower rates of surgery (p = 0.004). These six variables accounted for 58% of the variation. CONCLUSION: Geographical variations were found to exist, less than two thirds being explained by differences in the provision of care. The remaining variation may partly be attributed to private practice and the lack of consensus for many ophthalmic procedures (the 'surgical signature'), including a lowering of the threshold for surgery. These findings have implications for the planning of ophthalmic services.