|
|
 |
Back to Annual Meeting Program
Factors that determine length of stay after carotid endarterectomy: Opportunities to avoid hospital financial losses.
Julia Glaser, MD1, David Kuwayama, MD1, David Stone, MD1, Phillip Goodney, MD1, Andres Schanzer, MD2, Marc Schermerhorn, MD3, Jens Eldrup Jorgenssen, MD4, Richard Powell, MD1, Andrew Stanley, MD5, Jack Cronenwett, MD1, Brian Nolan, MD1
1Dartmouth Hitchcock Medical Center, Lebanon, NH, 2University of Massachusetts Medical Center, Worcester, MA, 3Beth-Israel Deconess Medical Center, Boston, MA, 4Maine Medical Center, Portland, ME, 5Fletcher-Allen Medical Center, Burlington, VT
Introduction: Length of stay (LOS) > 1 day after elective carotid endarterectomy (CEA) incurs financial losses for hospitals, given fixed DRG-based reimbursement. We sought to identify factors leading to prolonged length of stay following carotid revascularization.
Methods: Patients undergoing CEA in 22 centers of the Vascular Study Group of New England database between 2003 and 2011 (n=8860) were analyzed. Only elective, primary, same-day admission CEAs performed on a weekday were included, leaving a study cohort of 7108. Hierarchical multivariate logistic regression analysis was performed to identify predictors of LOS > 1 day. A Knauss-Wagner Chi-Pie analysis was performed to determine the relative contributions of the covariates to the prediction model.
Results: Post-op LOS > 1 day occurred in 17.5% of the sample (n=1244). Average LOS was 1.4 days and ranged from 1 to 91 days (median=1). There was significant variation in rates of LOS > 1 day across centers (5% to 100%, p 1 day and their % contribution to the prediction model included: any major adverse post-op event (MAE) (45%), low volume surgeon (<15 CEA per year) (29%), increasing age (8%), female gender (5%), positive preoperative stress test (3%), pre-operative major stroke within 30 days (2%), medication dependent diabetes (1%), O2 or medication dependent COPD (1%), history of CHF (1%), and CEA performed on Friday (2%). The model exhibited good predictive capacity (ROC 0.75) and explained nearly all variation (97%) in LOS.
Conclusions: Certain patient characteristics predispose to LOS > 1 day after elective CEA. However, patient-level factors play only a modest (21%) role in determining LOS. MAEs are the biggest driver of LOS > I day, while other systematic processes of care associated with low operative volume contribute substantially to prolonged LOS, independent of MAEs. These findings can be used to guide quality improvement efforts designed to reduce LOS after elective CEA.
Back to Annual Meeting Program
|
