Lower extremity amputation in diabetes: An exploratory study of the factors in amputation type and patient outcomes

Background: Lower extremity amputation among patients with diabetes is a rarely studied topic. Providers' decisions regarding a below-knee Diabetes-Related Lower Extremity Amputation (DRLEA) single-step vs. repeat major DRLEA terminating in a below-knee final amputation status are important for patient outcomes. Major DRLEA is any amputation at or above the metatarsophalangeal joint.This research will profile DRLEA procedures among diabetic patients of all age groups, discharged between January 1 2006 through December 31 2010 from over 600 member-hospitals of the largest consulting and group purchasing hospital consortium in the US. We hypothesize that patients' insurance source and plan type (HMO vs fee for service) influences the amputation approach, single step, vs. incremental, consecutive procedures. Unlike most health services studies that focus on Medicare data, this study will be unusual in providing a nationwide (albeit convenience) sample of working age adults together with Medicare patients served by these hospitals. Objective:To describe the potential role of non-clinical drivers of diabetes- related lower limb amputation decisions Methods: We conduct a retrospective analysis of secondary (claims) data on all adult cases of below-knee amputations during 2006 through 2010 (the study population). All these patients' discharges during the year prior to the index discharge will be identified to characterize the DRLEA as a single-procedure (one-step) amputation or as the final step following one or more repeat major amputations during the prior year. Independent variables include diabetes severity, co-morbidities (Elixhauser index), geographic region, payer source, plan type, hospital teaching status, and hospital ownership. Bi-variate statistics, multiple and multinomial logistic regression will be used, including fixed effects for hospital/ region, as appropriate. We will test for associations between insurance source and plan type (HMO vs fee for service) and the amputation approach (single vs. multiple step), adjusting for the variables mentioned above. The association of amputation approach with patient outcomes (mortality likelihood, length of stay, and in-hospital infection likelihood) will be presented. Results: Data cleaning and analyses are in progress. Conclusion: As the first study to explore this issue, our findings will serve to inform the medical professional and health policy community of the possible factors influencing DRLEA decisions.