Customized optimal use of telehealth for chronic care management and disparity reduction: A machine learning and clustering approach

During the COVID-19 pandemic, many waivers and programs were implemented to allow telehealth to be delivered with few restrictions, leading to a rapid increase in use. Since then, public authorities, payers and providers are deciding to either end, extend, or make permanent these waivers or programs. To inform these major policy changes, this project aims to understand the role of telehealth and patterns of telehealth use for chronic care management and disparity reduction in care outcomes, as well as the effects of policies and policy rollbacks. Aim 1 will identify clusters of telehealth delivery patterns, balance and alternating sequence with in-person care by patient rural/urban advantage/disadvantage subgroup and care system groups. Aim 2 will estimate the causal effects of changes in telehealth policy (waivers, subsidies, etc.) and care system contextual factors on telehealth delivery patterns, chronic care outcome and variations over time. This project will provide critical evidence to inform national decisions on telehealth delivery and policies for chronic care management and disparity reduction.