Association of frailty with clinical and financial outcomes of allogeneic hematopoietic stem cell transplant.

Abstract

7046 Background: Considered the only curative therapy for many hematological conditions, allogeneic hematopoietic stem cell transplantation (alloHSCT) is increasingly used in older patients. Yet, chronological age alone has been found to be an inaccurate predictor of posttransplant sequelae. More recently, the concept of frailty has emerged as a better marker of physiologic reserve. While formal frailty instruments have not been widely adopted due to their resource intensive nature, we applied a previously-validated coding-based approach to examine the association of frailty with in-hospital outcomes of alloHSCT. Methods: All adults (≥ 18 years) undergoing alloHSCT for hematological malignancies were identified in the 2010-2019 Nationwide Readmissions Database. The previously-validated binary Johns Hopkins Adjusted Clinical Groups indicator was used to classify patients as frail. In this algorithm, the presence of diagnoses across various domains, including malnutrition, dementia, and decubitus ulcer, constitute frailty. Regression models were developed to evaluate the independent association of frailty with in-hospital mortality, periprocedural complications, length of stay, hospitalization costs, and 30-day non-elective readmissions. Results: Of an estimated 48,161 patients, 9.8% were considered frail. Frail patients were not significantly older (51.8 ± 15.0 vs 51.6 ± 14.2, p = 0.685), but had higher Elixhauser scores (2.8 ± 1.3 vs 2.0 ± 1.3, p < 0.001). After adjustment, frailty was associated with increased odds of in-hospital mortality (AOR 2.22; 95% CI 1.68-2.94; P < 0.001), as well as infectious (AOR 1.44, 95% CI 1.18-1.76; P < 0.001) and respiratory (AOR 1.69, 95% CI 1.41-2.02; P < 0.001) complications. Frailty was also associated with greater risk of acute graft-vs-host disease (AOR 1.39, 95% CI 1.12-1.72; P = 0.003) and CMV (AOR 1.72, 95% CI 1.27-2.33; P < 0.001) or invasive fungal infection (AOR 1.49, 95% CI 1.07-2.06; P = 0.017). Further, frailty was linked to increased likelihood of non-home discharge (AOR 1.29, 95% CI 1.03-1.61; P = 0.026), 30-day non-elective readmissions (AOR 1.29, 95% CI 1.11-1.50; P = 0.001), and significant increases in length of stay (+6.05 days, 95% CI 4.41-7.69; P < 0.001) and costs (+$36,660, 95% CI 27,413-45,907; P < 0.001). Conclusions: Among patients undergoing alloHSCT, frailty, as measured by an administrative tool, was independently associated with increased in-hospital mortality, complications and overall resource use. Inclusion of frailty in risk models may better aid benchmarking efforts and inform shared-decision making. [Table: see text]