|
Back to 2016 Karmody Posters
Predictors of High Costs for the Most Expensive Vascular Surgery Admissions at a Large Academic Medical Center
Joshua Herb, Margaret Tracci, MD JD, Kenneth Cherry, MD, Gilbert Upchurch, MD, Amy Tucker, MD.
University of Virginia, Charlottesville, VA, USA.
OBJECTIVES:
Vascular disorders have a large economic impact on the healthcare system. This study sought to determine significant predictors of hospital admission cost, length of stay (LOS), and number of days in the ICU for the most expensive vascular surgery admissions over a fiscal year at a large academic medical center
METHODS:
Retrospective cost and clinical data were gathered from diagnosis related group reports on the most expensive admissions (85th percentile and above) from fiscal year 2014 at a large academic center. Admission data included the type of procedure, urgency, comorbidities, number of medical complications, and several different modifiable risk factors. A stepwise linear regression model was performed to identify significant predictors of total cost, total LOS, and length of ICU stay
RESULTS:
Eighty admissions were included in the final analysis. The median cost per admission was \,097, median LOS was 10 days, and median ICU stay was 2 days. Ninety-day mortality was 5%. Stepwise regression showed that LOS (p<.001), length of ICU stay (p<.001), number of operations during the admission (p<.001), and endovascular operations (p=.001) were predictive of high admission costs (adjusted R-squared, .757). The number of operations performed (p=.001), open operations (p=.037), number of medical complications (p<.001), and those with end-stage renal disease (p=.001) were predictive of longer total lengths of stay (adjusted R-squared, .547). More days in the ICU were predicted (adjusted R-squared, .485) by the number of medical complications (p<.001), admission for aortic disease (p=.002), open surgery (p=.013), and the absence of diabetes (p=.014).
CONCLUSIONS:
In a retrospective cost-prediction analysis, total LOS, number of ICU days, number of operations performed, and endovascular procedures were predictive of higher total admission costs. This suggests that efforts to minimize total length of stay and number of days in the ICU would be the most effective cost-reduction strategy in this subset of patients. The regression results suggest that reducing medical complications and number of operations performed would be the best strategy to reduce length of stay and the number of days in the ICU. Importantly, these results show that modifiable patient risk factors are not predictive of total admission costs, LOS, and number of ICU days for these admissions.
Back to 2016 Karmody Posters
|
