Evolution of primary care telehealth disparities during COVID-19: a retrospective cohort study (Preprint)

Abstract

Telehealth has become widely utilized as a novel way to provide outpatient care during the coronavirus disease (COVID-19) pandemic, but data about telehealth use in primary care remains limited. Studies in other specialties raise concerns that telehealth may be widening existing healthcare disparities, requiring further scrutiny of trends in telehealth utilization.

Our study aims to further characterize sociodemographic differences in primary care utilization via telehealth compared to in-person office visits before and during the COVID-19 pandemic, and determine if these disparities changed throughout 2020.

We conducted a retrospective cohort study in a large U.S. academic center with 46 primary care practices from April-December 2019 and April-December 2020. Data were subdivided into calendar quarters and compared to determine evolving disparities throughout the year. We queried and compared billed outpatient encounters in General Internal Medicine and Family Medicine via binary logic mixed effects regression model and estimated odds ratios with 95% confidence intervals (OR, 95% CI). We used sex, race, and ethnicity of the patient attending each encounter as fixed effects. We analyzed socioeconomic status of patients in the institution’s primary county based on the zip code the patient was residing in.

A total of 81,822 encounters in the pre-COVID-19 timeframe and 47,994 encounters in the intra-COVID-19 timeframe were analyzed; in the intra-COVID-19 timeframe, a total of 5,322 (11.1%) were telehealth encounters. Patients living in zip codes with high supplemental nutritional assistance usage were less likely to utilize primary care in the intra-COVID-19 timeframe (OR 0.94, 95% CI 0.90-0.98, P= .006). Encounters with the following patients were less likely to be via telehealth compared to in-person office visits- patients who self-identified as Asian (OR 0.74, 95% CI 0.63-0.86) and Nepali (OR 0.37, 95% CI 0.19-0.72), patients insured by Medicare (OR 0.77, 95% CI 0.68-0.88), and patients living in zip codes with high supplemental nutritional assistance usage (OR 0.84, 95% CI 0.71-0.99). Many of these disparities persisted throughout the year. While there was no statistically significant difference in telehealth utilization for patients insured by Medicaid throughout the whole year, sub analysis of quarter 4 found encounters with patients insured by Medicaid were less likely to be via telehealth (OR 0.73, 95% CI 0.55-0.97, P=.03).

Telehealth was not utilized equally by all patients within primary care throughout the first year of the COVID-19 pandemic, specifically by patients who self-identified as Asian and Nepali, were insured by Medicare, and lived in zip codes with low socioeconomic status. As the COVID-19 pandemic and telehealth infrastructure change, it is critical we continue to reassess the utilization of telehealth. Institutions should continue to monitor disparities in telehealth access, and advocate for policy changes that may improve equity.