Identification of Risk Factors in Lymphatic Surgeries for Melanoma: A National Surgical Quality Improvement Program Review
Mona Ascha1, Mustafa S Ascha, Brian Gastman
1From the *Case Western Reserve University School of Medicine; †Center for Clinical Investigation, Department of Epidemiology and Biostatistics, Case Western Reserve University; and ‡Department of Plastic Surgery, Cleveland Clinic, Cleveland, OH.
Introduction:
Sentinel lymph node biopsy (SLNB) and lymphadenectomy (LAD) are commonly performed in the staging and care of patients with malignant melanoma. These procedures are accompanied by complications that may result in hospital readmission, negatively affecting patient outcomes and potentially affecting surgical procedure reimbursement. The National Surgical Quality Improvement Program (NSQIP) database offers a large data set allowing physicians to evaluate 30-day readmission for surgical complications. We used this database to explore predictors of 30-day hospital readmission for SLNB and LAD in the axillary, cervical, and inguinal regions.
Methods:
Data from the years 2005 to 2014 of the American College of Surgeons NSQIP database were used. Cohorts were constructed according to International Classification of Diseases, Ninth Revision, classification and current procedural terminology codes. The outcome of 30-day return to hospital was defined as patients who were readmitted to the hospital or the operating room within 30 days. Multiple logistic regression results are presented for a prespecified set of predictors and predictors that were significant on univariate logistic regression analysis. Odds ratios and confidence intervals were calculated using maximum likelihood estimates, along with Wald test P values.
Results:
A total of 3006 patients were included. Of those, 151 (5.0%) returned to the hospital. Among 1235 LAD patients, 65 (5.3%) returned; among 1771 SLNB patients, 86 (4.9%) returned. Smoking was a predictor of hospital readmission for overall SLNB and for cervical SLNB on multivariate analysis. Age was a significant predictor for cervical and inguinal LAD. Hypertension was significant for cervical LAD. Diabetes, preoperative hematocrit, and male sex were predictors for inguinal SLNB. There were no significant predictors for axillary SLNB and axillary LAD, as well as overall LAD procedures.
Conclusions:
This is the first and largest study using American College of Surgeons NSQIP to examine 30-day readmission after SLNB and LAD for melanoma in 3 commonly operated anatomical regions. We have found several significant risk factors associated with hospital readmission, which are now being used as a quality measure for hospital performance and reimbursement, that may help surgeons optimize patient selection for SLNB and LAD.
