This retrospective cohort study evaluated the effect of diabetes on early postoperative outcomes following below-knee amputation (BKA). Prior work has described risk factors for early complications and reoperation after BKA in the overall population, but there is limited large-scale literature focused on outcomes specific to diabetic patients. The authors aimed to compare 30-day rates of complications, readmission, reoperation, discharge disposition, and mortality in patients with and without diabetes.
The analysis used the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) database for the years 2005 through 2019. Current procedural terminology (CPT) codes 27880, 27881, and 27882 were used to identify patients undergoing BKA. The final cohort comprised 28,776 patients, of whom 20,310 (70.6%) were recorded as having diabetes.
Outcomes compared between diabetic and nondiabetic patients included 30-day overall complications, severe complications, specific adverse events (including bleeding requiring transfusion and sepsis), readmission, reoperation, 30-day mortality, length of stay, and discharge destination. Multivariate analysis was used to identify independent associations between diabetes and selected outcomes, and a subgroup analysis compared insulin-dependent and noninsulin-dependent diabetic patients.
In unadjusted comparisons, patients with diabetes experienced higher rates of 30-day overall complications than nondiabetic patients: 39.0% versus 34.8% (P < .001). Severe complications were also more frequent among diabetic patients (9.12% vs 8.20%; P = .012).
The diabetic cohort had notably higher rates of specific adverse events. Rates of bleeding requiring transfusion were 20.4% in diabetic patients versus 15.0% in nondiabetic patients (P < .001). Reported rates of sepsis were 6.28% among diabetics and 5.23% among nondiabetics (P < .001).
These unadjusted results indicate greater early morbidity following BKA among patients with diabetes, encompassing both systemic infectious complications and perioperative bleeding requiring transfusion.
On multivariate analysis, the presence of diabetes independently predicted increased odds of adverse discharge to a facility (defined by the authors as discharge to a skilled nursing or rehabilitation facility rather than home). The odds ratio (OR) for discharge to a facility associated with diabetes was 1.184 (P < .001).
Paradoxically, diabetes was independently associated with decreased odds of reoperation within 30 days (OR 0.785; P < .001) and decreased odds of 30-day mortality (OR 0.809; P = .002). These adjusted findings contrast with the higher unadjusted complication rates observed in diabetic patients and suggest that, after accounting for measured confounders, diabetes remained linked to selected outcome differences in both directions.
Among patients with diabetes, the investigators compared outcomes for those labeled insulin-dependent versus noninsulin-dependent. Insulin-dependent diabetic patients had higher rates of any postoperative complication (39.7% vs 36.4%; P < .001) and higher readmission rates (14.1% vs 12.1%; P = .005) compared with other diabetic patients.
Length of stay was also greater in insulin-dependent patients: 12.4 days versus 11.6 days for noninsulin-dependent diabetics (P = .007). These subgroup findings identify insulin-dependent diabetes as a marker of further increased perioperative resource use and early morbidity within the diabetic population undergoing BKA.
The authors conclude that patients with diabetes undergoing BKA experienced significantly increased rates of 30-day complications, including severe complications, and were more likely to be discharged to a facility rather than home. However, after adjustment, diabetes was associated with lower odds of reoperation and lower odds of 30-day mortality.
Given the observed higher rates of bleeding requiring transfusion and sepsis among diabetic patients, the authors recommend that targeted perioperative optimization strategies be considered for this population to address modifiable risks and potentially reduce early morbidity and resource utilization.
The report is a Level III retrospective cohort study based on the ACS NSQIP database from 2005–2019. The source provides cohort size, timeframe, CPT codes used, and the outcomes and statistical results summarized above. Details beyond those reported in the abstract (for example, specifics of covariates included in multivariate models, institutional practices, or granular perioperative management protocols) were not provided in the source abstract and therefore are not detailed here.
Clinicians and hospital teams should interpret the findings with consideration of retrospective design and database-derived limitations, and consider these results as hypothesis-generating when planning perioperative optimization for patients with diabetes undergoing BKA.