The Palliative Care in Assisted Living (PCAL) pilot study: successes, shortfalls, and methodological implications.

Troubling deficits exist in palliative care (PC) of older adults under the prevailing "terminal care"-oriented model. We previously described a PC model--TLC--that provides a blueprint for remedying these shortfalls. In this model, PC is envisioned as Timely and Team-oriented, Longitudinal, and Collaborative and Comprehensive. We present results of the Palliative Care in Assisted Living pilot, comparing two TLC model-based, facility delivered interventions for improving the PC of elderly assisted living residents in Sacramento, California, a growing and under-researched population. The less intensive intervention involved one assessment followed by a PC improvement recommendation letter to the resident, family member, primary provider, and facility staff, while the more intensive intervention involved assessments and letters every three months. Primary outcomes were SF-36 Physical (PCS) and Mental (MCS) Component scores and recommendation adherence. Eighty-one subjects enrolled (mean age 85), 58 in the more and 23 in the less intensive group. A loved one attended 56% of baseline assessments. Most subjects expressed a preference for maintaining current quality of life over prolonging life at reduced quality. None were eligible for hospice care. A total of 418 recommendations (mean 5.1 per subject) were generated concerning symptoms, mood, functional impairments, and advance directives. We found no significant differences in recommendation adherence between more (42%) and less (44%) intensive groups, and no significant changes in PCS and MCS scores within or between groups. However, a loved one's attendance of the baseline assessment was associated with improved PCS scores (p=0.04). Our pilot study had methodological limitations that could account for the lack of significant outcome effects. In this context, and given the myriad unmet PC needs we detected, interventions based on the TLC model might allow delivery of timely PC to assisted living residents not eligible for hospice care. Further studies exploring the TLC model appear warranted.