1053 — Prognosis-Based Palliative Care Management: Impact on End-of-Life Resource Use

Rosenfeld KE (COE-Sepulveda), Steckart MJ (COE-Speulveda), Riopelle DD (COE-Sepulveda), Magner-Perlin ML (COE-Sepulveda), Lorenz K (COE-Sepulveda), Lee ML (COE-Sepulveda), Wagner GJ (RAND Corporation)

Objectives:
VHA has widely promoted palliative care (PC) as a means to improve quality of care in seriously ill veterans and to reduce resource use through alternatives to costly, invasive end-of-life care. However, previous controlled trials of PC services have not clearly demonstrated such resource savings. We examined the resource impact of a conceptually-driven, longitudinal palliative care management intervention among a population of seriously ill veterans.

Methods:
Between August 2004 and November 2006 at the VA Greater Los Angeles, 400 eligible, consenting medical inpatients with a resident-estimated one-year mortality greater than 25% were randomized to usual care or to an intervention employing palliative care consultation followed by longitudinal nurse care management targeting care coordination, symptom management, treatment preference clarification, and patient and family practical support. Primary outcome measures included total hospital and ICU days, and for patients who died, hospital and ICU days overall, and in the final 30 and 60 days of life. Data were collected through patient and family interviews, chart review, and administrative sources. Statistical tests included Chi Square, F-test, and Mann -Whitney U.

Results:
Patients randomized to intervention (n = 200) and control (n = 200) were similar with regard to demographics, diagnostic category, resident-estimated mortality risk, baseline functional status and symptoms other than pain, and overall mortality and survival duration. There were no significant differences between intervention and usual care patients in total hospital days or ICU days following study entry (25.7 vs. 29.1 days and 3.3 vs. 3.0 days, p > .05). Among patients who died, intervention patients had fewer hospital and ICU days in both the final 30 days (6.2 days vs. 9.0 days and 1.2 days vs. 2.0 days respectively, p = .04 for both) and 60 days of life (8.9 days vs. 13.9 days, p = .02, and 1.0 days vs. 3.0 days, p < .05, respectively).

Implications:
A complex PC intervention reduced terminal, but not total, hospital-based resource use among seriously ill patients.

Impacts:
VHA may be able to enhance care quality while reducing terminal resource use by expanding PC services, particularly through mechanisms that promote active case-finding and longitudinal palliative support services.