High-Intensity End-of-Life Care of Allogeneic Hematopoietic Cell Transplantation Recipients (HCT) — a Population-Based Study

BACKGROUND: Intensity of end-of-life care has received much attention in oncology due to concerns that high intensity care is likely expensive and may be inconsistent with patient goals. For these reasons, the National Quality Forum endorsed quality measures around intensity of end-of-life care for oncology patients and many studies have examined rates of and disparities in intensity of end-of-life care in this population treated with conventional therapy. However, little is known about intensity of end-of-life care in allogeneic HCT recipients, a population that is at very high risk for morbidity and mortality.

METHODS: Using the California Office of Statewide Health Planning and Development administrative discharge database linked to death certificates, we performed a population-based analysis of allogeneic HCT recipients who died between 2000-2011 and had received their transplant within one year of death. Previously-defined and validated end-of-life care intensity indicators were used and included an intense medical intervention (CPR, intubation, ICU admission, or hemodialysis) within 30 days of death, discharge and subsequent readmission within 30 days of death, and hospital death. A patient was considered to have a given indicator if it was coded during an admission that took place entirely within 30 days of death and/or coded during a terminal admission. Multivariable logistic regression models were used to evaluate associations between clinical and sociodemographic characteristics and: i) hospital death, ii) any medically-intense intervention, iii) ≥2 intensity indicators, iv) readmission within 30 days of death. Regression analyses were done for the whole group and then stratified by age at death (pediatrics: 0-21y; adolescent and young adult [AYA]: 22-39y; adult: 40+y). Sociodemographic and clinical characteristics were considered associated with high intensity care if they were associated with an intensity indicator in at least 2 of the 4 regression analyses. Estimated costs incurred in the last year of life were calculated using each hospital's charge-to-cost ratio adjusted for 2016 dollars.

RESULTS: A total of 1978 patients were included in the analysis (390 pediatric, 439 AYA, and 1149 adult) with a median age at death of 40 years old; 57% were non-Hispanic white and 28% were Hispanic; 58% were male; 94% had an oncologic diagnosis (AML [38%]; ALL [22%] were most common). Eighty-four percent of the population died in the hospital, 71% of the patients had ≥2 intensity indicators and 52% had a medically-intense intervention. During their terminal admission or admission within 30 days of death, 48% of the cohort was admitted to the ICU, 44% was intubated, 21% received hemodialysis, and 8% received CPR. Entire Cohort: patient age (15-21y, 30-39y, and 40-49y; referent grp: age >60y) and presence of comorbidities at HCT were associated with a medically-intense intervention, ≥2 interventions, and hospital death. Pediatric Cohort: Presence of ≥2 comorbidities during the admission for HCT was associated with high-intensity end-of-life care (a medically-intense intervention, ≥2 interventions, and hospital death). AYA Cohort: Presence of ≥2 comorbidities during the admission for HCT was associated with high-intensity end-of-life care (a medically-intense intervention, ≥2 interventions, and hospital death). Older Adult Cohort: Patient age (40-49y; referent grp: age >60y) and presence of ≥2 comorbidities at HCT were associated with a medically-intense intervention and >2 intensity indicators. Cost of Care: HCT recipients in this study were admitted a median of 106 days during their last year of life, costing the system $558,136 in inpatient expenses per patient.

CONCLUSIONS: Allogeneic HCT recipients who die within a year of transplantation frequently receive a number of intensive end-of-life care interventions, as evidenced by the fact that 84% of the patients die in the hospital and 48% are admitted to the ICU at the end-of-life. Further work is needed to determine if these rates are consistent with patients' goals and whether interventions like earlier end-of-life conversations and/or palliative care consults may decrease intensity of care in this population as in non-HCT oncology patients.