Abstract: Estimating the Impact of Social Work on Palliative Care Outcomes: A Meta-Analysis (Society for Social Work and Research 26th Annual Conference - Social Work Science for Racial, Social, and Political Justice)

Schedule:

Sunday, January 16, 2022

Marquis BR Salon 10, ML 2 (Marriott Marquis Washington, DC)

* noted as presenting author

John Cagle, PhD, Associate Professor, University of Maryland, Baltimore, Baltimore, MD

Todd Becker, LMSW, PhD Student, University of Maryland, Baltimore, Baltimore, MD

Background and Purpose

Palliative care (PC) is an interdisciplinary approach to care that seeks to maximize comfort for individuals with serious illness. Common outcomes evaluated in PC include patient quality of life (QOL), symptom relief (SR), home death (HD), and cost savings (CS). Research has suggested that the composition of the interdisciplinary team may be associated with PC outcomes. As core members of PC teams and stewards of the psychosocial domains of health, social workers directly target these evaluative criteria. Yet, comprehensive research into the contribution of social work on PC outcomes is lacking. The objective of this study was to synthesize the impact of social work involvement on commonly evaluated PC outcomes (QOL, SR, HD, CS) in the published literature.

Methods

We used a meta-analytic approach. Article searches were limited to recently published (2000–2016) controlled trials.

Outcomes were standardized to facilitate comparison. QOL was assessed through the Functional Assessment of Chronic Illness Therapy-Palliative Care Scale. Higher scores indicate better QOL. SR was standardized on the Edmonton Symptom Assessment Scale. Lower scores indicate greater SR. We converted the raw short-term (1–3 months) and long-term (4–6 months) QOL and SR results into aggregate mean differences (palliative care vs. control) in studies with (ES_SW) and without (ES_NSW) social work involvement. HD was analyzed through averaging the odds ratios in studies with (OR_SW) and without (OR_NSW) social work involvement. CS was standardized against U.S. dollars adjusted for the 2020 inflation rate and averaged across studies. We weighted all outcomes at the study-level to accommodate differing sample sizes.

Results

Twenty-three (82%) of the retained clinical trials (N = 28) used individual random assignment. Based on 15 studies (n_SW = 6; n_NSW = 9), short-term QOL outcomes were substantially better when social work was involved (ES_SW = 1.75 vs. ES_NSW = 0.1). Long-term QOL outcomes based on 12 clinical trials (n_SW = 2; n_NSW = 10) were comparable yet slightly lower in social work-involved studies (ES_SW = −0.01 vs. ES_NSW = 0.09). Based on 10 studies (n_SW = 3; n_NSW = 7), PC approaches with social work involvement (ES_SW = −2.71) demonstrated superior SR compared to those without social work involvement (ES_NSW = −0.15). No studies involving social work examined long-term SR (ES_NSW = −0.18). Based on three studies (n_SW = 2; n_NSW = 1), the average odds of HD were greater when social work was involved (OR_SW = 2.07 vs. OR_NSW = 1.36). An average CS of $7,017.92 was calculated across studies. Although only four studies out of five studies reflected statistically significant CS, all five studies included social workers. Thus, there is currently no scientifically rigorous evidence showing that PC without social work saves costs.

Conclusions and Implications

These results affirm social work as a critical member of the interdisciplinary PC team by suggesting that social work involvement benefits patients, families, and health systems alike. Future trials should explicitly describe the elements of the PC model being tested, including composition of the interdisciplinary team, fidelity, and dose.