A. Akhavan1, A. Verma1, A. Ng1, N. Chervu1, S. Bakhtiyar1, S. Sakowitz1, P. Benharash1 1David Geffen School Of Medicine, University Of California At Los Angeles, Cardiothoracic Surgery, Los Angeles, CA, USA
Introduction:
Heart transplantation is a lifesaving, yet resource intensive, therapy for patients with late stage heart failure. However, in the current era of value-based healthcare, contemporary studies evaluating patient and hospital characteristics associated with excess resource use are lacking. Therefore, we used a nationally representative database and robust statistical methods to evaluate drivers of and inter-hospital variation in costs following heart transplantation.
Methods:
All adults (≥18 years) undergoing heart transplantation within two days of admission were identified from the 2016-2019 National Inpatient Sample (NIS). Hospitalization costs were calculated via application of hospital specific cost-to-charge ratios and inflation adjusted to 2019. A mixed effects regression to model costs was developed, with patient characteristics as the first-level and hospital identifier as the second. Baseline hospitalization costs attributable to each institution were derived from random effects using Bayesian post estimation methodology. High-cost hospitals (HCH) were defined as centers with baseline costs greater than $60,000. The association of HCH status with in-hospital mortality, perioperative complications, length of stay and non-home discharge was assessed using multivariable regression.
Results:
Over the four-year study period, an estimated 4,945 patients met study criteria, with median cost of $143,000 [interquartile range 110,000-186,000]. The average age was 55 years, and 28% of the cohort was female. After multivariable adjustment, neurological disorders (+$42,000, 95% CI 23,300-61,200), liver disease (+$37,000, 95% CI 17,000-57,400) and ventricular assist device explant (+$20,000, 95% CI 10,600-29,500) were associated with increased costs. Moreover, approximately 31.7% of variation in costs was attributable to inter-hospital differences, and 5.9% of hospitals were classified as HCH. HCH status was not associated with altered odds of in-hospital mortality (adjusted odds ratio (AOR) 3.93, 95% CI 0.31-49.6). However, patients at HCH had increased odds of infectious complications (AOR 6.50, 95% CI 2.13-19.8), deep vein thrombosis (AOR 11.0, 95% CI 1.52-78.9), and acute kidney injury (AOR 2.70, 95% CI 1.01-7.22). HCH status was also associated with increased length of stay (LOS) (+8.2 days, 95% CI 2.4-14) and likelihood of nonhome discharge (AOR 3.6, 95% CI 1.3-9.9).
Conclusion:
The present study ascertained the presence of significant inter-hospital variation in hospitalization costs of heart transplantation. Management at an HCH was associated with greater odds of complications and longer LOS, factors which may be suitable targets for cost mitigation strategies. Nonetheless, future studies are necessary to identify specific care pathways and practice patterns that may reduce costs at HCH.
