Risk Factors for Unplanned Readmission and Stump Complications following Major Lower Extremity Amputation
John Phair, Charles DeCarlo, Karan Garg, Evan Lipsitz
Montefiore Medical Center, Bronx, NY
OBJECTIVE – Unplanned thirty-day readmission rates – a marker of quality of patient care – after major lower extremity amputation are limited. We evaluated predictors of readmission at our institution.
METHODS – We conducted a retrospective review of all patients undergoing above-knee (AKA) or below-knee amputation (BKA) with closure between November 2007 and November 2014. Patient demographics were collected. Predictors of unplanned 30-day readmission and stump complications were determined by multivariable logistic regression.
RESULTS – We identified 811 patients (AKA 325; BKA 486). The overall 30-day readmission rate was 26.3% (AKA 24.3 %; BKA 27.6%). The cohort consisted of 54% men and had a mean age of 68.7 ± 14.9. Stump complications accounted for 25.8% of readmissions (AKA 20.0%; BKA 80.0%). Other common diagnoses included sepsis (27.7%), infection without the diagnosis of sepsis (9.9%), chf exacerbation (7.0%), and diabetes related complications (6.6%). Surgical intervention was performed on 61.8% of stump complications (AKA 11.8%; BKA 88.2%). BKA stump complications were converted to AKA’s in 34.1% of cases. None of the AKA stump complications required a higher level of amputation, i.e., hip disarticulation. Independent predictors of 30-day readmission included previous contralateral or ipsilateral major amputation, ASA class 4, ESRD, and gangrene as the indication for the index procedure. Independent predictors of 30-day readmission for stump complications included previous major amputation, below-knee amputation, Hispanic ethnicity, and chronic kidney disease (stages II-V). Multivariate odds ratios (95% confidence interval) and p-values are listed below (Table 1).
CONCLUSIONS – The 30-day readmission rate following major lower extremity amputation is high, with wound infections accounting for a significant proportion of these readmissions. There was no difference in readmission rates based on level of amputation. However, those undergoing a below-knee amputation were more likely to present with stump complications, require a surgical intervention, and often a higher level of amputation. Identification of high-risk patients may play a role in reducing post-operative readmissions, and stump complications.
|Independent Predictors of 30-day readmission||Multivariate Odds Ratio (95% Confidence Interval)||Multivariate P Value|
|Previous Major Amputation||2.91 (1.77-4.78)||<0.01|
|ASA 4||1.71 (1.15-2.51)||0.01|
|Independent Predictors of 30-day readmission for stump complications||Multivariate Odds Ratio (95% Confidence Interval)||Multivariate P Value|
|Previous Major Amputation||12.75 (6.34-26.0)||<0.01|
|Below-Knee Amputation||5.27 (2.55-11.95)||<0.01|
|Hispanic Ethnicity||1.99 (1.08-3.66)||0.03|
|Chronic Kidney Disease||2.40 (1.07-6.21)||0.05|
|Non-Insulin Dependent Diabetes||0.45 (0.24-0.83)||0.01|
|Independent Predictors of 30-day readmission for stump complications requiring surgical intervention||Multivariate Odds Ratio (95% Confidence Interval)||Multivariate P Value|
|Previous Major Amputation||27.89 (11.91-68.43)||<0.01|
|Below-Knee Amputation||13.93 (4.73-53.00)||<0.01|
|Rest Pain||3.81 (1.30-10.24)||0.01|
|Hispanic Ethnicity||2.57 (1.16-5.83)||0.02|
Table 1: Multivariate odds ratios (95% confidence interval) and p-values
Back to 2017 Annual Meeting Program