COVID-19 has ushered in significant changes to the provision of telemedicine-based healthcare, fueled in part by emergency suspension of HIPAA restrictions on video chat applications like Zoom, relaxed provider licensing requirements, evolving reimbursement structure, and other factors. Many of these changes are expected to persist past the early months of pandemic, even as distancing protocols and other early-day protections are relaxed and rolled back. Given the new reality, investment in telemedicine has taken on a wider scope, and optimality becomes a consideration.
In this work we discuss a principled approach to remote care/telemedicine facility investment given realities of capacity constraints, costs both fixed and variable in nature, and service needs across underserved populations who may not have other facilities within easy distance. We show that a regionally-sensitive modeling approach combining optimization with machine learning exceeds central-planning, regionally-insensitive allocation. We also outline companion practical ideas for teaching these ideas in the classroom