Impact of Virtual vs. In-Person Palliative Care on Symptom Burden: A Multisite Cohort Study.

Abstract

Outpatient palliative care (OPC) is increasingly being delivered virtually but symptom-related outcomes associated with virtual OPC are unknown. Examine the association over time between symptom burden and visit modality among people with serious illnesses. In this retrospective cohort study, we included patients ≥18 years old with ≥2 visits at 58 OPC programs. We examined change in symptom burden (Edmonton Symptom Assessment Scale, ESAS; range 0-100) over time (first 4 visits) by visit modality (in-person vs. virtual vs. mixed). Using mixed effects multinomial regression, we examined the difference in the linear trend of ESAS scores over time by visit modality. We repeated analyses restricting to those with baseline moderate-to-severe symptom burden (ESAS ≥ 40/100) and stratifying by language (English vs. non-English). Among 3,896 patients, 86% had all in-person visits, 4% all virtual, 10% mixed; 51% were ≥65 years old; 84% were referred for symptoms; 21% had non-English preferred languages. ESAS scores decreased significantly more for the mixed (-5.9 points; P-trend vs. virtual = .007) and in-person groups (-3.6; P-trend vs. virtual = .049) compared to the virtual group (+0.03). Among those with ESAS≥40/100 at baseline, ESAS scores decreased most in the mixed group (-24.1; P-trend vs. in-person = .02; P-trend vs. virtual <.001) followed by in-person (-17.7; P-trend vs. virtual <.001), compared with stable symptoms in the virtual group (+0.03). Trends were similar after stratifying by language. Our results suggest that mixed virtual and in-person care is associated with a greater reduction in symptom burden compared to in-person or virtual care alone.