Assessment of Determinants of Value in Carotid Endarterectomy

Abstract

Background: Over 150,000 carotid endarterectomies (CEA) are performed annually worldwide, accounting for $900 million in the United States alone. How cost/spending and quality are related is not well understood but remain essential components in maximizing value. We sought to identify determinants of variability in hospital 90-day episode value for CEA.

Methods: Medicare and private-payer admissions for CEA from January 2, 2014 to August 28, 2020 were linked to retrospective clinical registry data for hospitals in Michigan performing vascular surgery. Hospital-specific, risk-adjusted, 30-day composite complications (defined as reoperation, new neurologic deficit, myocardial infarction, additional procedure including CEA or carotid artery stenting, readmission, or mortality) and 30-day risk-adjusted, price-standardized total episode payments were used to categorize hospitals into low or high value by defining the intersection between complications and spending.

Results: A total of 6,595 patients across 39 hospitals were identified across both datasets. Patients at low-value hospitals had a higher rate of 30-day composite complications (17.9% vs. 10.1%, P < 0.001) driven by a significantly higher rate of reoperation (3.0% vs. 1.4%, P = 0.016), readmission (10.7% vs. 6.2%, P = 0.012), new neurologic deficit (4.6% vs. 2.3%, P = 0.017), and mortality (1.6% vs. 0.6%, P < 0.049). Mean total episode payments were $19,635 at low-value hospitals compared to $15,709 at high-value hospitals driven by index hospitalization ($10,800 vs. $9,587, P = 0.002), professional ($3,421 vs. $2,827, P < 0.001), readmission ($3,011 vs. $1,826, P < 0.001), and post-acute care payments ($2,335 vs. $1,486, P < 0.001). Findings were similar when only including patients who did not suffer a complication.

Conclusions: There is tremendous variation in both quality and payments across hospitals included for CEA. Importantly, costs were higher at low-value hospitals independent of postoperative complication. There appears to be little to no relationship between total episode spending and surgical quality, suggesting that improvements in value may be possible by decreasing total episode cost without affecting surgical outcomes.