Variations in clinical and regional practice patterns affect length of stay after elective carotid endarterectomy
Elsie Ross, MD, MSc, Matthew Mell, MD, MS.
Stanford Health Care, Stanford, CA, USA.
OBJECTIVES: To determine the effect of potentially modifiable practice patterns and regional variations in length of stay (LOS) after elective, uncomplicated carotid endarterectomy (CEA).
METHODS: Data were obtained from the Vascular Quality Initiative database and included patients with complete data who received elective, uncomplicated CEA between 2012 and 2015 across 16 regions in the United States. The main outcome measure was LOS greater than 1 day after surgery. Using generalized linear modeling and Least Absolute Shrinkage and Selection Operator (LASSO) regression and controlling for important clinical variables, we evaluated what modifiable clinical practices and regional differences contributed to variance in postoperative LOS. RESULTS: A total of 24,233 patients were included. Overall, 25% of patients had a LOS greater than 1 day, with mean post-procedure LOS of 1.7 ± 8 days. After adjusting for clinical factors contributing to LOS > 1 day (gender, history of hypertension, diabetes, CAD, CHF, COPD, ESRD, stress test results, discharge to SNF), modifiable clinical practices significantly associated with LOS > 1 day included weekend operation (Odds Ratio 1.1, Confidence Interval 1-1.2), preoperative aspirin (OR 0.85, CI 0.8-0.9) and statin use (OR 0.89, CI 0.8-0.96), intra-operative dextran (OR 1.4, CI 1.3-1.5), drain placement (OR 1.3, CI 1.2-1.4), intraoperative duplex (OR 0.89, CI 0.8-0.97) or arteriogram (OR 0.6, CI 0.6-0.8), general anesthesia (OR 1.5, CI 1-2), high volume surgeon (OR 0.7, CI 0.6-0.9), and need for continuous intravenous blood pressure management (IVBPM) (OR 3.1, CI 2.9-3.3). Use of IVBPM accounted for 28% of the variation in LOS. Further analysis demonstrated that high volume centers (OR 0.52, 0.4-0.6), eversion endarterectomy (OR 0.74, CI 0.7-0.8), regional anesthesia (OR 0.59, CI 0.5-0.7), and routine shunting (OR 0.89, CI 0.8-0.96) were significantly associated with less use of IVBPM. Patients receiving beta-blockers only on the day of surgery were more likely to need IVBPM (OR 1.5, CI 1.3-1.8). Surgeon volume did not significantly affect IVBPM use. After adjusting for demographic, clinical and practice variables, there remained significant regional variation in length of stay.
CONCLUSIONS: Modifiable clinical practices as well as unmeasured regional factors significantly contribute to variations in LOS after elective, uncomplicated CEA. Standard care pathways may reduce LOS variation, reduce overall LOS, and improve value.
Back to 2017 Karmody Posters
