End of Life Healthcare and Hospice Utilization Among Patients with Diffuse Large B-Cell Lymphoma

Introduction: Although diffuse large B-cell lymphoma (DLBCL) is considered a highly curable malignancy, 5-year overall survival rates have ranged between 60-65% in population-based studies. There is a lack of data on predictive factors that can guide therapy in DLBCL in the frontline setting or at relapse. Therefore, most patients are treated in a similar fashion with the goal of achieving cure, which may lead to increased times spent in the healthcare setting and decreased rates of hospice enrollment toward the end of life in DLBCL patients. In this study, we sought to examine the patterns of healthcare and hospice utilization in the last 90 days of life among DLBCL patients.

Methods: We conducted a retrospective analysis using a registry of DLBCL patients who died between 2009 and 2020 from a single-center. We quantified the total time spent in the last 90 days in a medical setting, amassing time spent in the emergency ward, inpatient wards, oncology clinics, infusions, and transfusions. We computed each patient’s round trip time to and from the clinic, and used pre-specified standardized appointment times to one hour for clinic visit and two hours for infusion/transfusion visits. Hospital days were counted as 24 hour periods. We combined the total time spent with healthcare and reported it as a percentage of last 90 days of a patient’s life. Additionally, we captured other metrics such as time from diagnosis to death, percentage of ICU admissions, and hospice enrollment. We conducted univariate and multivariate analyses using age, stage, International Prognostic Index (IPI score), sex, insurance status, double/triple hit status to ascertain factors associated with increased hospitalizations (≥15 days), hospice enrollment, and short interval from hospice enrollment to death (≤5 days).

Results: A total of 83 patients were included in the study. The median age at diagnosis was 64 years, and 47% were males. Median IPI score was 3, and 14/53 (28%) were patients with high grade lymphomas (double/triple hit). Median time from diagnosis to death was 9 months, and 77% patients died during or shortly after completion of first line therapy. The proportion of time spent by patients with inpatient or outpatient healthcare in the entire cohort during their last 90 days of life was 23%. The patients spent 16% of their last 90 days in the hospital (median 14 days, interquartile range or IQR 7-26). The median number of electronic portal messages or calls to the oncology clinic was 3 (IQR 0-5). 54% patients were enrolled in hospice within the last 90 days, with a median interval from hospice enrollment to death of 5 days. Overall, 54% of the 83 patients died in the hospital itself, and this comprised 84% of deaths among patients who were not enrolled on hospice. Also, 56% percent of the patients had at least one intensive care unit (ICU) admission in their last 90 days of life. On univariate analysis, the only factor associated with ≥15 days hospitalization was increasing age (p=0.013), while no significant associations were found on multivariate analysis for this outcome measure. Higher IPI score was found to be positively correlated with enrollment on hospice [RR 2.7; 95% CI 1.38-5.45]. No factors were found to be significantly predictive of a short interval from hospice enrollment to death (≤5 days) on univariate or multivariate analysis.

Conclusion: In our study, DLBCL patient spent 23% of their last 90 days of life with healthcare services, which predominantly included hospital stays. While utilization of hospice was slightly over 50%, most patients were on hospice for less than one week before death. Over half the patients had at least one ICU admission during last 90 days of life. Chemotherapy use was prevalent in the last days of life, with 7% of patients having received greater than 2 cycles of therapy within the last 30 days of life. Advanced IPI was associated with increased enrollment on hospice. Our study shows high burden of healthcare utilization in DLBCL patients, with most deaths occurring in the hospital. There is a need to develop strategies to optimize utilization of hospice services, which may include earlier goals of care discussion and palliative care referrals. Even in a highly curable malignancy such as DLBCL, our study highlights an area of unmet need to improve patient/caregiver experience toward end of life.