Aims/hypothesis: The study aimed to explore the variation in recorded incidence of lower limb amputation in England.
Methods: The incidences of amputations in adults with and without diabetes were determined from hospital episode statistics over 3 years to 31 March 2010 and compared between the 151 Primary Care Trusts (PCTs) in England.
Results: There were 34,109 amputations, including 16,693 (48.9%) in people with diabetes. The incidence was 2.51 per 1,000 person-years in people with diabetes and 0.11 per 1,000 person-years in people without (relative diabetes risk 23.3). Incidence varied eightfold across PCTs in people both with diabetes (range 0.64-5.25 per 1,000 person-years) and without (0.03-0.24 per 1,000 person-years). Amputations in people with diabetes varied tenfold--both major (range 0.22-2.20 per 1,000 person-years) and minor (range 0.30-3.25 per 1,000 person-years). The incidences of minor and major amputations were positively correlated both in those with (r = 0.537, p < 0.0005) and without (r = 0.611, p < 0.0005) diabetes. Incidences of amputations were also correlated between people with and without diabetes (total amputations r = 0.433, p < 0.0005; major amputations r = 0.528, p < 0.0005). There was a negative correlation between the incidence of amputation and estimated prevalence of ethnic Asians. No association was found between the PCT incidence of either total amputations and general population prevalence of social deprivation (r = -0.138, p = 0.092) or smoking (r = 0.137, p = 0.096).
Conclusions/interpretation: Variation in amputation incidence occurs across England. Because of the similarity in amputation variation between people with and without diabetes the variation may reflect generic differences in local healthcare delivery, although racial factors may also contribute.