Independent Predictors of Readmission After Femoral-to-Popliteal Artery Bypass Grafting in Diabetics

Abstract

A database review was performed to determine which factors contribute to hospital readmission for diabetic patients within 30-days of femoral-to-popliteal artery bypass (FPB) surgery. The National Surgical Quality Improvement Program (NSQIP) database was queried from the years 2005 to 2011 to identify diabetic patients who underwent FPB. Patients groups by those who were readmitted (RA) ≤30 days of operation and those who were not readmitted (NRA). Patient demographics, comorbidities, perioperative data, and outcomes were compared. Of the 5523 patients identified, 334 were in the RA group and 5189 were in the NRA group. There were 208 vein bypasses in the RA group and 2949 vein bypasses in the NRA group. The rest were prosthetic conduit. Difference in bypass conduit approached but did not achieve statistical significance (P = .051). The RA group had a higher proportion of patients with medical comorbidities, including ventilator dependence (2.1% vs 0.1%; P < .01), congestive heart failure (8.7% vs 3.1%; P < .01), recent myocardial infarction (4.5% vs 2%; P < .01), and insulin dependence (59.6% vs 53.6%; P = .03) compared with the NRA group. Preoperative vascular disease was more severe in the RA group as well, with rest pain/gangrene (50.9% vs 39.1%; P < .01) and history of amputation (73.1% vs 49.8%; P < .01) both higher in the RA group. The RA group also had a higher proportion of emergency procedures (11.4% vs 4.4%; P < .01), longer operative time (226 ± 11.5 minutes vs 208.4 ± 2.6 minutes; P < .01) and longer prior hospital length of stay (16.5 ± 1.8 days vs 8 ± 0.3 days; P < .01). The RA group also was taken back to the operating room more frequently during the index admission (36.2% vs 15.3%; P < .01) than the NRA group. Preoperative hypertension (odds ratio, 46.7; 95% confidence interval, 34.6-62.9) and rest pain/gangrene (odds ratio, 50.1; 95% confidence interval, 36.4-69.1) were the only independent predictors of RA on multivariate analysis. Insulin dependence and preoperative cardiac comorbidities are not predictive of readmission rates after FPB in diabetic patients. The presence of critical limb ischemia is the most predictive factor for 30-day readmission after FPB in diabetic patients.