Wednesday, September 26, 2018 10:35 AM – 12:00 PM Understanding Lumbar Stenosis/Spondylolisthesis: 12. A preoperative score for predicting 30-day readmission after elective 1- or 2-level posterior lumbar fusion

Abstract

BACKGROUND CONTEXT Posterior lumbar fusion is one of the most commonly performed procedures in the United States. Readmission rates are an important metric of quality of care and directly tied to reimbursement. Identification of factors that influence readmission rates is valuable for preoperative optimization and for appropriate risk stratification of patients. PURPOSE To develop a model to predict 30-day readmission in elective 1- or 2-level posterior lumbar fusion patients. STUDY DESIGN/SETTING Case control study. PATIENT SAMPLE Patients undergoing elective 1- or 2-level posterior lumbar spine fusion in the state inpatient databases (SID) of New York, Florida, North Carolina, Utah, and California. OUTCOME MEASURES Creation of the readmission after posterior spinal fusion (RAPSF) scoring system. METHODS Data were queried for 30-day readmission, as well as demographic and surgical data. Patients were randomly assigned to a derivation cohort and a validation cohort. A stepwise multivariate analysis was conducted to create a model to predict 30-day readmission. Variables with p 1.1 and p RESULTS There were 92,262 patients in the derivation cohort and 90,257 in the validation cohort. Thirty-day readmission rates were 10.9% and 11.1% respectively. The following variables were significant in the multivariate model and included in the RAPSF (point value in parentheses): age (50-59: 4, 60-69: 4, 70-79:6,>90: 8), gender (female: 3), race (Hispanic: 4, black: 5), insurance (Medicare: 4, Medicaid: 5, other–non-commercial: 3), anterior approach (4), cerebrovascular disease (3), chronic pulmonary disease (3), congestive heart failure (4), diabetes without chronic complications (4), diabetes with chronic complication (4), hemiplegia/paraplegia (7), rheumatic disease 3), drug abuse (4), electrolyte disorder (4), osteoporosis (3), depression (3), obesity (4), morbid obesity (5). Linear regression between readmission rate and RAPSF fit the derivation cohort with an adjusted r2 of 0.92 and a coefficient of 0.011 (p CONCLUSIONS We developed an easy-to-use tool, the RAPSF, to accurately predict readmission rates in patients undergoing elective 1-2 level posterior lumbar fusion. The RAPSF may be useful to empower informed choice regarding the risks of surgery and to guide an evidence-based approach to preoperative optimization and risk adjustment within alternative payment models for elective spine surgery. FDA DEVICE/DRUG STATUS This abstract does not discuss or include any applicable devices or drugs.