Enhancing Caregiver Empathy by Empowering Patients to Discuss Their Values

ENHANCE was a prospective cluster randomized controlled trial with patients admitted to four general medicine units at one hospital and their hospital Caregivers (nurses and hospitalists). We chose to focus on internal medicine units specifically, due to the documented need for hospitalists to establish rapport and maintain relationships with their patients. Two units were randomized to the control group and two units to the intervention group. The cluster randomized design was chosen to reduce spillover effect.

INTERVENTION The intervention was an 18"x18" poster board with a large blank surface for patients to share information about who they are as people and what is important to them. The Care to Share poster was co-designed with ten medicine inpatients during pilot testing. The goal of the poster was to equip Caregivers with humanizing information about their patients, thereby provoking more genuine empathetic interactions. Based on input from pilot testing, the researcher wrote on the poster for the patient to ensure ease of completion and legibility. The posters were displayed on the wall behind patients' beds, so Caregivers entering the room could easily see the information when facing the patient. Nurses (registered nurses, patient care nursing assistants, nurse managers and health unit coordinators) were designated as control or intervention Caregivers based on their unit, while all hospitalists (physicians and advanced practice providers) were treated as intervention Caregivers since their practice was not limited to specific medicine units.

DATA COLLECTION Data collection occurred between December 2018 and May 2019 and in three stages: Caregiver baseline survey, Intervention vs Control patient survey, and Caregiver post-intervention survey. Caregivers were recruited to the study using emails and in person surveys and were blinded to details of the intervention. Caregiver surveys contained questions about communication and experience, validated inventories on Caregiver-reported burnout and empathy, and demographic information. Caregiver surveys for all hospitalists and nurses in the intervention group, also contained questions to gather feedback on the Care to Share poster.

Patients were eligible for study recruitment when they were admitted to one of the study units (either from the ICU or as a new admission) within the past 48 hours for the intervention group, or within 3-5 days of admission for the control group, age 18 years or older, on the Internal Medicine service, and capable of making their own medical decisions. This timing was chosen so that all patient participants - both control and intervention - would complete surveys at about the same length of time into their admission.

Intervention patients were informed that the Care to Share poster was for them to share whatever they want their care team to know about them as a person. A member of the research team (CF) recorded content on the poster and immediately hung it on the wall behind the patient's bed. The intervention participants were then approached for survey administration 48 hours after the poster was hung (or after 24 hours if the patient was to be discharged the day after poster creation). Patients discharged before 24 hours after poster creation were not eligible for survey administration. The researcher administered the survey to patients unless the patient indicated they preferred to complete it on their own.

MEASURES Patient: The primary patient outcome variable was patients' perceived empathy of their hospitalist, measured by the Jefferson Scale of Patient Perceptions of Physician Empathy(JSPPPE). This 5 item 7-point scale, completed by patients, measures patient-physician empathetic engagement. Patient surveys also included covariates such as trust in their physician, measured by the Interpersonal Physician Trust Scale (IPTS), a 10 item 5-point scale, as well as a 3 item patient-reported measure of shared decision making, CollaboRATE, and demographic characteristics. The survey fielded to patients in the intervention group also contained items to gather feedback on the Care to Share poster, including two open-text questions for patients to comment in their own words. Additionally, specific medical details were gathered from the EMR, such as length of stay (LOS), and 30-day readmission rates.

Caregiver: The caregiver survey consisted of the Maslach Burnout Inventory (MBI)36, Jefferson Scale of Empathy (JCE), and demographic items. The MBI measures the frequency of job-related feelings of burnout using a scale of 0 (never) to 6 (every day). Burnout is measured on 3 dimensions- emotional exhaustion (EE), depersonalization (DP), and personal accomplishment (PA). We used the burnout criteria of EE>26 and DP>9 which was used in previously reported literature on internal medicine physicians. The JCE is a 20 item 7-point Likert scale completed by the Caregiver that measures a health professional's empathy. As with the patient survey, the surveys fielded to Caregivers on intervention units (i.e., some nurses and all hospitalists) contained items to gather feedback on the Care to Share poster.

ANALYSIS Based on a priori power calculations, we required 306 patient participants and approximately 40 caregivers to detect a difference in the JSPPPE. We computed descriptive statistics to describe patient and caregiver samples, and multivariate nested regression models to test if the intervention predicted patients' perceptions of clinician empathy, shared decision-making, and trust. Data analysis was conducted using JMP, SAS and R. Verbatim comments to the two open-text questions were analyzed using content analysis.