Identifying Barriers to Telehealth Home-Based Palliative Care for Patients with Serious Illness—The HomePal Study

BACKGROUND: The HomePal study was designed to provide rigorous evidence and policy guidance for providing more-accessible, more-effective, more-affordable, high-quality, home-based palliative care (HBPC) through telehealth tools. The comparative effectiveness design was framed to assess whether remote physician video consults can substitute for the initial physician home visit to improve program efficiency and enhance care coordination and team communications to promote patient trust and confidence in the care team. Patient and family stakeholders felt that the comparators were appropriate and important because they preferred a timely video visit with the physician, facilitated by a nurse in the patient's home, vs seeing the physician much later, not seeing the physician in person at all, or having to manage the video technology themselves. OBJECTIVES: 1. Determine if a tech-supported HBPC model that promotes early remote access to a provider while the nurse was in the patient's home was noninferior to a standard HBPC model that included routine home visits by a nurse and provider on the coprimary outcomes of symptom improvement (proximal) and days alive and at home in the last 180 days of life (distal). 2. Compare the effects of tech-supported HBPC and standard HBPC on the caregiver primary outcome of caregiver perception of preparedness for caregiving. 3. Examine the effectiveness of HBPC across highly relevant clinical and sociodemographic subgroups—for example, cancer, organ failure, neurodegenerative diseases, multiple chronic conditions (3 or more diseases/conditions), participation of a caregiver in the patient's care, racial/ethnic minorities, and low income. 4. Describe the longitudinal pattern of palliative care service exposure, intensity of acute health service use, and quality of end-of-life care in patients with serious illness who have been referred to or outreached for HBPC services but did not enter the HBPC pragmatic trial (observational cohort). 5. Describe barriers and facilitators to the implementation of tech-supported HBPC. METHODS: Study design: We conducted a pragmatic, cluster-randomized trial to compare tech-supported HBPC with standard HBPC across 2 Kaiser Permanente regions (Southern California and Northwest). Registered nurse case managers (n = 111) across the sites were randomized to 1 of the 2 arms so that approximately half of the patient-caregiver dyads received routine in-home nurse and physician visits (standard; patients, n = 1781; caregivers, n = 208), and the other half received in-person nurse visits coupled with remote physician consultation-supervision (tech supported; patients, n = 1752; caregivers, n = 233). The participant enrollment period was January 7, 2019, through January 25, 2020, with follow-up through October 31, 2020. HBPC: Both arms received standard palliative care components: a comprehensive assessment of bio-psychosocial-spiritual needs and care planning, aggressive pain and symptom management, patient and caregiver education/skills training, medication review and management, emotional and spiritual support, care coordination, referral to community resources and other home health services as needed, and 24/7 triage services as required. Nurses in the tech-supported HBPC arm could initiate remote physician video consultations at any point when they were in patients' homes, whereas nurses in the standard arm did not have access to video technology. Study population: Patients who were 18 years and older, were living with a serious illness and had an expected survival of 1 to 2 years, were homebound, and had been admitted to HBPC were included. Adult (at least 18 years old) family or other caregivers were identified by the patient during the phone screening as someone “who helps you with your care.”. Study outcomes: The coprimary patient outcomes included symptom improvement as measured by the Edmonton Symptom Assessment System (at 1 month) and days alive and at home in the last 6 months of life. Secondary outcomes included quality of life (QOL), satisfaction with care, health care use, and survival. Caregiver-centered outcomes included perception preparedness for caregiving (primary), QOL, burden, anxiety and depression, satisfaction with care, and health services use (Kaiser Permanente members). Analytic methods: Because of challenges with the trial, we were unable to complete the originally planned noninferiority analysis. The analytical approach for the descoped papers varied but included thematic analysis, descriptive statistics, correlations, linear mixed models, and/or Fine and Gray proportional hazards, and competing risk models. RESULTS: We stopped the study in January 2020 after the first year of participant enrollment and before the start of the COVID-19 pandemic. Several implementation challenges informed our decision to terminate the study, as follows: (1) the video-based visit frequency was low because of persistent HBPC nurse and physician workforce shortages, which contributed to scheduling and coordination difficulties and were further exacerbated by the restrictions imposed by the randomized nurse cluster design; (2) technology and connectivity were unreliable; and (3) when clinicians used video-based visits, they preferred to increase follow-up contact instead of using it as a substitute for the initial physician home visit because they had, among other things, concerns regarding the quality of care for a population with such complex needs. Ultimately, the health system's decision to expand telehealth use more broadly in response to external market pressures and changes to telehealth payment policies conflicted with and took precedence over the randomized study design. The findings described in this report reflect the following proposed “descoped” analyses of data collected up to the study's termination: (1) lessons learned from implementing video visits; (2) evaluation of the tradeoffs of using clinician-collected, patient-reported outcomes (PROs) within pragmatic trials of serious illness and the validity of the days-at-home outcome metric; and (3) examination of the associations between financial and spiritual distress on patient and caregiver outcomes. CONCLUSIONS: Our study highlights challenges to conducting rigorous comparative effectiveness research with patients who have complex, serious illness and their overburdened caregivers in the chaotic and chronically underresourced home care environment using unreliable video technology. The success of large, multisite pragmatic trials testing complex interventions embedded within health systems is highly dependent on ongoing engagement of multiple changing stakeholders within and across health systems, from frontline clinicians to executives. Multiyear palliative care trials must contend with the ever-evolving care delivery landscape caused by changing population needs, market pressures, and/or regulatory and payment policy changes. LIMITATIONS: There are several general limitations from the descoped analyses, including the generalizability of the findings to other HBPC programs that are not embedded within integrated health systems, the observational nature of the analyses and residual confounding, the heavy reliance on proxy responses for PROs, and the high rate of missingness in follow-up PRO data because of death and other circumstances related to serious illness.