Methodological issues affect variability in reported incidence of lower extremity amputations due to diabetes.

The aim of this study is to evaluate the influence of different methodological techniques commonly utilized to identify the incidence of diabetes related lower extremity amputations. Medical records for each hospitalization for an amputation in 1993 in six metropolitan statistical areas in South Texas were abstracted. Every hospitalization, amputation and amputee was identified to allow separate analysis. Furthermore, data was categorized by ethnicity, level and age. Diabetes was verified using WHO criteria. Incidence rates were calculated per 10,000 diabetic patients at risk per year, both diagnosed only (DO) and diagnosed and undiagnosed combined (DUC). In total 1922 amputations were carried out during 1228 hospitalizations for 1043 amputees. The incidence rates per 10,000 diabetic patients (DO) were: 157.6 amputations, 101.2 hospitalizations and 87.0 amputees. When calculated using the DUC population at risk the rates were: 92.8 amputations, 59.6 hospitalizations and 51.2 amputees. Trends were found to be similar when analyzed by gender and ethnicity. The variability detected using of different methodological techniques to determine incidence rates is considerable and may have significant consequences when rates from different studies are compared.