Natural Language Processing to Assess Palliative Care Processes and Health Care Utilization in Seriously Ill Older Adults with Severe Trauma

Comprehensive Summary

Hoffman et al. aimed to assess the prevalence of inpatient palliative care documentation in elderly patients with severe trauma and determine associations with post-discharge outcomes. Five key palliative care processes were measured using natural language processing and tested for associations with post-discharge health care utilization via multivariable regression: code status limitations, goals of care conversations, hospice discussions, palliative care consultations, and health care proxy designations. Of the 1,267 patients, 87% survived hospitalization, and among survivors, one-year mortality was 26%. The presence of one or more palliative care processes did not significantly affect mean hospital days, home days, emergency department visits, or intensive care unit days. However, palliative care processes were significantly associated with increased future hospice designation.

Outcomes and Implications

Palliative care has been recommended as a treatment option for older adult patients with severe trauma. Despite emphasis on palliative care utilization alongside life-sustaining treatment, outcomes associated with palliative care interventions and their effectiveness in this population remain poorly defined. This study aimed to address this knowledge gap by examining the prevalence of key palliative care processes and their associations with post-discharge outcomes. The findings suggest that goals of care conversations, specialty palliative care consultations, and inpatient hospice discussions were under-utilized in this population, highlighting opportunities for health care providers to improve integration of palliative care processes in elderly trauma patients.