Changes in Primary Care Telehealth Use and Impact on Acute Care Visits for Ambulatory Care‐Sensitive Conditions during COVID ‐19

Research Objective The COVID‐19 pandemic forced a dramatic shift from in‐person care to telehealth, with an overall decrease in outpatient utilization. These changes particularly stressed outpatient care delivery and may have resulted in decreased availability and access to primary care for patients, potentially increasing otherwise avoidable emergency department (ED) visits and hospitalizations. One commonly used indicator of primary care access and quality is acute care visits for ambulatory care‐sensitive conditions (ACSCs), such as for pneumonia, uncontrolled diabetes, or congestive heart failure exacerbations. Our study examined differences in telehealth adoption across practices and evaluated the association between telehealth adoption and ACSC visits. 1 TABLE Differences‐in‐differences model of practice telemedicine conversion rate on acute care visits for acute and chronic ACSCsAcute ACSC aOR (95% CI) Chronic ACSC aOR (95% CI)Telemedicine tertile (mean telemedicine conversion rate)Low (9%) ref ref Medium (30%) 0.98 (0.91–1.07) 0.98 (0.76–1.26) High (66%) 1.07 (0.98–1.17) 1.14 (0.86–1.50)Study Design We conducted a retrospective study of claims data from a large commercial insurer in Michigan. We first profiled telehealth adoption by primary care practices during March–July 2020. We defined a practice's “telehealth conversion rate” as the proportion of visits conducted via telehealth during this period compared to the total number of visits during the same period in 2019. Then, to enable comparison between groups at a time when both outpatient and acute care visits were in flux, we used a differences‐in‐differences (DID) model to determine whether varying levels of primary care telehealth conversion were associated with differences in acute care visits (ED visits and hospitalizations) for ACSCs from June–September 2020. We examined visit rates for acute and chronic ACSCs separately, controlling for practice size, in‐person visit volume, and zip code‐level attributes as well as patient characteristics (age, gender, comorbidities). We performed sensitivity analyses using varying definitions of telemedicine conversion rates and multiple model specifications. Population Studied Six million Blue Cross Blue Shield of Michigan beneficiaries across 3780 primary care practices from January 2019 to September 2020. Principal Findings Average primary care practice telehealth conversion rate was 25% (median 10%), and 29% of practices had no telehealth claims identified. Practices that did not adopt telehealth tended to be smaller and were more likely to be in rural areas. We found no significant differences in the rate of ED visits and hospitalizations for ACSCs by practice‐level telemedicine conversion tertile after adjusting for practice case‐mix, as shown in Table 1. Sensitivity analyses showed similar results. Conclusions Beneficiaries within a large commercial payer experienced rapid shifts from in‐person to telehealth for their primary care, though telehealth adoption was not evenly distributed, with smaller and more rural practices being less likely to adopt telemedicine. These changes did not seem to obviously help or harm patients as ED visits and hospitalizations for ACSCs were similar across groups. Implications for Policy or Practice Widespread substitution of telehealth for in‐person care had little impact on cost of care with respect to avoidable ED visits and hospitalizations in the near‐term. Additional research should continue to monitor this trend as health care utilization stabilizes beyond the pandemic. Primary Funding Source University of Michigan Institute for Health Policy and Innovation.