A Focus on Amputation Level: Factors Preventing Length Preservation in the National Inpatient Sample

Abstract

Background: Diabetic lower-extremity disease is the primary driver of mortality in patients with diabetes. Amputations at the forefoot or ankle preserve limb length, increase function, and, ultimately, reduce deconditioning and mortality compared with higher-level amputations, such as below-the-knee amputations (BKAs). We sought to identify risk factors associated with amputation level to understand barriers to length-preserving amputations (LPAs). Methods: Diabetic lower-extremity admissions were extracted from the 2012-2014 National Inpatient Survey using ICD-9-CM diagnosis codes. The main outcome was a two-level variable consisting of LPAs (transmetatarsal, Syme, and Chopart) versus BKAs. Logistic regression analysis was used to determine contributions of patient- and hospital-level factors to likelihood of undergoing LPA versus BKA. Results: The study cohort represented 110,355 admissions nationally: 42,375 LPAs and 67,980 BKAs. The population was predominantly white (56.85%), older than 50 years (82.55%), and male (70.38%). On multivariate analysis, living in an urban area (relative risk ratio [RRR] = 1.48; P < .0001) and having vascular intervention in the same hospital stay (RRR = 2.96; P < .0001) were predictive of LPA. Patients from rural locations but treated in urban centers were more likely to receive BKA. Minorities were more likely to present with severe disease, limiting delivery of LPAs. A high Elixhauser comorbidity score was related to BKA receipt. Conclusions: This study identifies delivery biases in amputation level for patients without access to large, urban hospitals. Rural patients seeking care in these centers are more likely to receive higher-level amputations. Further examination is required to determine whether earlier referral to multidisciplinary centers is more effective at reducing BKA rates versus satellite centers in rural localities.