Association of Adequacy of Broadband Internet Service With Access to Primary Care in the Veterans Health Administration Before and During the COVID-19 Pandemic.

Importance: Although telemedicine expanded rapidly during the COVID-19 pandemic and is widely available for primary care, required broadband internet speeds may limit access. Objective: To identify disparities in primary care access in the Veterans Health Administration based on the association between broadband availability and primary care visit modality. Design, Setting, and Participants: This cohort study used administrative data on veterans enrolled in Veterans Health Administration primary care to identify visits at 937 primary care clinics providing telemedicine and in-person clinical visits before the COVID-19 pandemic (October 1, 2016, to February 28, 2020) and after the onset of the pandemic (March 1, 2020, to June 30, 2021). Exposures: Federal Communications Commission-reported broadband availability was classified as inadequate (download speed, ≤25 MB/s; upload speed, ≤3 MB/s), adequate (download speed, ≥25 <100 MB/s; upload speed, ≥5 and <100 MB/s), or optimal (download and upload speeds, ≥100 MB/s) based on data reported at the census block by internet providers and was spatially merged to the latitude and longitude of each veteran's home address using US Census Bureau shapefiles. Main Outcomes and Measures: All visits were coded as in-person or virtual (ie, telephone or video) and counted for each patient, quarterly by visit modality. Poisson models with Huber-White robust errors clustered at the census block estimated the association between a patient's broadband availability category and the quarterly primary care visit count by visit type, adjusted for covariates. Results: In primary care, 6 995 545 veterans (91.8% men; mean [SD] age, 63.9 [17.2] years; 71.9% White; and 63.0% residing in an urban area) were seen. Adjusted regression analyses estimated the change after the onset of the pandemic vs before the pandemic in patients' quarterly primary care visit count; patients living in census blocks with optimal vs inadequate broadband had increased video visit use (incidence rate ratio [IRR], 1.33; 95% CI, 1.21-1.46; P < .001) and decreased in-person visits (IRR, 0.84; 95% CI, 0.84-0.84; P < .001). The increase in the rate of video visits before vs after the onset of the pandemic was greatest among patients in the lowest Area Deprivation Index category (indicating least social disadvantage) with availability of optimal vs inadequate broadband (IRR, 1.73; 95% CI, 1.42-2.09). Conclusions and Relevance: This cohort study found that patients with optimal vs inadequate broadband availability had more video-based primary care visits and fewer in-person primary care visits after the onset of the COVID-19 pandemic, suggesting that broadband availability was associated with video-based telemedicine use. Future work should assess the association of telemedicine access with clinical outcomes.