Delivering Healthcare Through Teleconsultations: Implications for Offline Healthcare Disparity

Teleconsultations allow patients to search for, receive, and pay for medical consultations virtually. With remote diagnosis and treatment capability, teleconsultations have been proposed as a potential solution to the long-standing social problem of geographic disparity in healthcare. Although this sounds promising, unforeseen frictions could suppress the virtual flow of healthcare. It is unclear, then, whether teleconsultations actually mobilize healthcare to underserved regions. To advance our understanding, we first empirically investigate whether teleconsultations generate a virtual flow of healthcare to mitigate geographic healthcare disparity. Second, we examine whether social, information, and geography frictions are present in the virtual healthcare flow. To this end, we curate unique data capturing regional offline health resources and various regional characteristics and match them with teleconsultation instances over 10 years (2006–2015). Our exponential random graph model analysis provides encouraging empirical evidence that teleconsultations connect physicians in resourceful regions and patients in underserved areas—a desirable direction that can alleviate geographic healthcare disparity. However, we also find that various frictions are present. For instance, social and information frictions, such as cultural and linguistic differences and limited media coverage, suppress the supposedly free flow of teleconsultations across regions. Furthermore, although teleconsultation is anticipated to spark long-distance healthcare, we find that teleconsultations are less likely as the regions between patient and physician become farther apart. We examine two plausible mechanisms that contribute to the observed geography friction: (1) a low-information bandwidth of a teleconsultation channel and (2) the financial constraint of rural patients. Supplementary analyses using granular data (fees, physician ranks, and illness types) provide corroborating evidence for the proposed mechanisms.