Calling for Care? The Risky Proposition of Teletriage in Healthcare Demand Management

Many healthcare providers and payers offer teletriage, a service where concerned patients can get advice over the phone about what kind of medical care they should seek. In theory, teletriage should help patients choose an appropriate provider (or choose not to visit a provider at all), reducing unnecessary and duplicate provider visits. This decision support should benefit patients and improve demand management in the healthcare system, yet teletriage is not widely used despite a growing push for telemedicine in healthcare operations. We use a Markov decision process to model a patient’s choice of providers during an illness episode where the illness severity is only partially observable to the patient. Allowing teletriage improves the patients’ outcomes and reduces the rate of arrivals to the emergency department (ED), because patients with a high level of uncertainty in their state use the teletriage service. Whether teletriage is cost effective depends on the extent to which calling teletriage changes the action that the patient would otherwise take. However, when patients are biased about their own state transition probabilities, we show that they may actually be better off without the option to use teletriage. We show that when patients underestimate their probability of worsening, adding a teletriage option can increase the overall arrival rate to the ED. Making teletriage more accessible exacerbates this problem, which occurs when too many patients use teletriage as a substitute for visiting a primary care physician. Our results demonstrate that teletriage is a risky proposition and highlight the need for more personalized medicine, such as targeting teletriage services based on a patient’s risk profile.