Palliative care referral timing and end-of-life outcomes in multiple myeloma.
Abstract
311 Background: Patients with multiple myeloma (MM) experience high symptom burden and intensive end-of-life (EOL) care, yet palliative care (PC) referral timing remains inconsistent. The optimal timing and clinical triggers for PC referral in MM are poorly defined. Methods: A retrospective study of adult MM decedents treated within a single health system from January 2020 through December 2024 was performed. Patients were categorized by timing of first PC consultation prior to death. The primary outcome was an EOL aggressiveness score incorporating hospitalization, ICU admission, ED utilization, chemotherapy near death, and hospital/ICU death. Secondary outcomes included hospice enrollment and length of stay (LOS), outpatient versus inpatient PC utilization, advance care planning (ACP) utilization, and acute-care use. Trigger-based analyses evaluated admissions and ED visits preceding consultation. Results: Among 50 MM decedents, 64% received PC consultation. PC involvement more than 30 days before death was associated with substantially longer hospice LOS, greater outpatient PC exposure, increased ACP utilization, and fewer ED visits near death compared with late or no PC involvement. Trigger-based analyses demonstrated that any recent admission was strongly associated with whether PC consultation occurred, suggesting that clinicians recognize clinical deterioration. However, these same triggers did not reliably distinguish late from earlier referral timing, supporting the finding that recognition of decline alone may be insufficient to optimize PC utilization. Patients referred between 31 and 90 days before death demonstrated the lowest observed EOL care intensity, although this subgroup was small. Conclusions: In MM, PC referral occurs in response to clinical deterioration but is often delayed. Referral before the final month of life was associated with longer hospice engagement and reduced acute-care utilization. These findings support development of structured referral pathways and trigger-based strategies to facilitate earlier outpatient PC involvement in MM. End-of-life outcomes by timing of palliative care consultation. Outcome No PC ≤30 Days 31–90 Days >90 Days p-value Median aggressiveness score 2.5 3.0 0.0 3.0 0.033 Hospitalization ≤30 days, % 72.2 100 25 91.7 0.003 ED visit ≤30 days, % 66.7 93.8 0 58.3 0.003 Hospice enrollment, % 27.8 62.5 100 16.7 0.005 Median hospice LOS, days 4 3 42 23.5 0.170 Hospice LOS (>30 vs ≤30/no PC), median days 3 42 0.030 Outpatient PC involvement, % 0 6.2 50 66.7 <0.001 ACP documentation, % 52.9 93.8 100 83.3 0.021 Aggressiveness score incorporated hospitalization, ICU admission, ED utilization, chemotherapy within 14 days of death, and hospital/ICU death. Hospice LOS was calculated only among hospice-enrolled patients.