Impact of centralized care on aortic dissection outcomes in the United States.

Abstract

Over the past 30 years in the United States, hospitals have increasingly become incorporated into hospital systems, leading to organized care with more complex cases being managed at large urban-teaching hospitals. Over a similar period, changes in intervention guidelines for aortic dissection (AD) have occurred, with continually growing options for endovascular, minimally invasive treatment. Given these dynamic changes, we examined trends in AD hospitalizations, intervention approaches, and hospital characteristics over past 20 years to elucidate the effect of centralized care on outcomes. We identified all patients presenting with AD, both ascending and descending (International Classification of Diseases [ICD]-9-CM: 441.0; ICD-10-CM: I71.0) in the National Inpatient Sample between 2000 and 2021. We then examined the use of open repair, endovascular repair, and complex endovascular repair, as well as nonoperative or medical management. Stratified by hospital setting (urban-teaching, urban-nonteaching, and rural), we analyzed trends of interventions and in-hospital mortality over time and compared them with unadjusted Wald test. If an operation was performed, we were able to discern between ascending/arch or descending aorta after the 2017 ICD revision. We identified 553,030 patients with AD. The number of inpatients in the United States with ADs has increased, with an incidence of 26.7 cases per 100,000 in 2000 to 47.2 cases/100,000 in 2020 (P < .01). Overall, including all hospital settings, ADs were less frequently managed nonoperatively (2000-2021: 83%-71%) and more frequently managed endovascularly, with 85% of all descending and 16% of all ascending/arch ADs undergoing thoracic endovascular aortic repair in 2021. Over time, ADs have increasingly been managed at urban-teaching hospitals (2000-2021: 72%-92%; P < .01). Since 2016, urban-teaching hospitals more frequently intervened on ADs compared with their rural counterparts (21% vs 6%; P < .01), despite having similar rates of failed medical management (9.8% vs 8.2%; P = .30). Finally, comparing the last 5 years, urban-teaching hospitals have lower mortality rates when managing AD vs their rural counterparts (10.9% vs 11.7%; odds ratio, 1.10; P = .02) and if managed operatively, there was a lower associated risk of mortality at urban-teaching hospitals compared with urban-nonteaching hospitals (12.5% vs 17.3%; odds ratio, 1.46; P < .01). Both AD hospitalizations and interventions have significantly increased over the past two decades in the United States. The growth of large hospital systems and their absorption of smaller hospitals into integrated primary through quaternary care centers has resulted in an increase in "regionalization" of care, in which complex cases are transferred to larger urban-teaching centers. Our analysis suggests there is a mortality benefit from the centralization of aortic care to tertiary/urban-teaching centers, although further research into this question is required.