223 Acceptability and Reliability of a Novel Palliative Care Screening Tool Among Emergency Department Providers During Pre-Implementation Testing

Abstract

Physical and psychosocial suffering are highly prevalent among patients with advanced illness. Most patients who could benefit from inpatient palliative care (PC) consultation are never identified or referred to this service. The emergency department (ED) is an ideal place to screen for unmet PC needs and initiate early PC consultation in order to mitigate unnecessary suffering and reduce costs from invasive interventions that are incongruent with patient wishes. We developed the Palliative Care and Rapid Emergency Screening Project (P-CaRES): a multi-phase project facilitating early referral to inpatient PC consultation among ED patients with significant palliative care needs. In the first stage of this project, we derived a novel, content-validated screening tool for patient screening. We previously developed the P-CaRES tool. In this study we report on a pilot trial to test the reliability and acceptability of the P-CaRES tool among ED providers, and assess how both vary based on a provider’s role (attendings, residents, and nurses) and level of experience. A two-part electronic survey was distributed to ED providers across multiple sites and across the United States. A “gold standard,” against which to compare respondents’ use of the P-CaRES tool, was created based on input from expert palliative care physicians on 10 written case vignettes featuring ED patients with PC needs. The experts’ input was validated using Gwet’s AC1 coefficient for inter-rater reliability. To test external reliability of the screening tool, we then distributed the case vignettes to ED providers at each site, and recorded their responses. These were compared both against the gold standard and against different subsets of ED providers (divided both by role and level of experience). Acceptability of both ED-PC screening in general and of our PC screening tool specifically was assessed using a modified Ottawa Acceptability Scale based on a 1-5 Likert rating. Finally, descriptive statistics were used to report all outcomes. Two hundred thirteen ED providers employed in three different regions across the country responded to the survey, and 185 completed it. Overall, 78.5% of ED providers self-reported that currently they only screen patients with advanced illness for palliative care needs 10% of the time, or less. Only 10.75% of patients agreed that they already use an effective strategy to screen patients for such needs. The majority of providers felt the P-CaRES tool would be useful in their practice (80.5%); 87.1% agreed that the tool was clear and unambiguous and 87.5% agreed that use of it would likely benefit patients. Using the screening tool, ED providers generated a PC referral in concordance with PC experts at a rate of 88.7% (95% CI 86.4% - 90.6%). Efficacy of the tool was similar across all provider roles and levels of experience. ED screening for unmet palliative care needs using a brief, novel, content-validated screening tool is both acceptable and reliable in pilot testing. Implementation of the tool and clinical trial is warranted.