Unplanned 30-day readmission in patients with diabetic foot wounds treated in a multidisciplinary setting

Abstract

Objective: Readmission rates are known to be high for vascular surgery patients in general, but there are limited data describing the risk of surgical and nonsurgical readmission among patients with diabetic foot ulcers (DFUs). Our aim was to identify factors associated with unplanned readmission in DFU patients treated in a multidisciplinary setting.

Methods: We studied a single-center cohort of patients enrolled in a multidisciplinary diabetic foot service (July 2012-June 2017). Readmissions were stratified by planned vs unplanned and related vs unrelated to the wound and vascular status. Predictors of unplanned 30-day readmission were examined with univariable and multivariable logistic regression models including all covariates with P ≤ .10.

Results: There were 460 admissions in 206 patients during the study period, including 99 total readmissions (21.5%). Readmissions were categorized as planned (n = 18 [18.2%]) or unplanned (n = 81 [81.8%]) and as related (n = 67 [67.7%]) or unrelated (n = 32 [32.3%]) to the wound and vascular status. The most frequent reasons for unplanned 30-day readmission were deterioration of the foot wound (41%), vascular complications (15%), gastrointestinal complications (10%), cardiac complications (8%), and acute kidney injury (8%). The average length of stay for the initial admission was 9.0 ± 7.1 days, whereas the average unplanned readmission length of stay was 8.6 ± 9.1 days (P = .38). On univariable analysis, hypertension (odds ratio [OR], 2.80; 95% confidence interval [CI], 1.19-6.59), peripheral arterial disease (OR, 1.80; 95% CI, 1.09-2.99), and exposure to an open vascular operation (OR, 2.64; 95% CI, 1.34-5.17) were associated with a higher risk of 30-day unplanned readmission (P ≤ .02). Private, military, or self-pay insurance was protective (OR, 0.52; 95% CI, 0.28-0.97). Wound duration, location, and Wound, Ischemia, and foot Infection (WIfI) classification were not associated with readmission (P ≥ .22). After risk adjustment, only hypertension (OR, 2.80; 95% CI, 1.19-6.59) and current smoking (OR, 1.95; 95% CI, 1.02-3.73) were independently associated with 30-day unplanned readmission, but the predictive accuracy of the model was weak (C statistic = 0.69).

Conclusions: We found a 17% unplanned 30-day readmission rate in this prospective cohort of DFU patients enrolled in a multidisciplinary diabetic foot service. Only current smoking and hypertension were independent predictors of readmission after risk adjustment. These findings suggest that implementation of a smoking cessation program may be beneficial to reduce unplanned readmissions in DFU patients. They also highlight the complexity involved in achieving comprehensive DFU care and the unpredictability of readmissions in this unique population of patients.