Employing Computational Linguistics In Patient-Provider Secure Exchanges for Comparative Effectiveness Research (ECLIPPSE-CER)

The digital communication revolution in health care is well underway. Doctors and patients are now communicating using secure messaging (email-like messages exchanged on a patient portal). This way of communicating is becoming a very popular form of interaction in health care and has become even more popular during the COVID-19 pandemic. However, despite this change, very little research has been carried out to understand secure messaging’s function in health care.

In addition, little research has informed how this platform can best be used to improvement quality, safety and equity. Finally, there has been no research to learn whether secure messaging is being used to provide the scientific evidence people need to make good health care decisions for themselves and their families. As the popularity of secure messaging grows, there is a real need and opportunity for researchers to create computerized methods to be able to analyze the conversations taking place in online exchanges. Doing so will allow researchers to discover how this platform can best be used to spread the scientific evidence people need to make informed decisions.

ECLIPPSE-CER is a research project to develop new methods to allow the study of huge numbers of secure messages without breaking doctor-patient confidentiality. The project will involve over 2 million messages exchanged between diverse patients with diabetes and their primary care doctors. First, the study team will use new computerized methods to remove all private identifiers. Then they will use other computerized methods that allow the rapid interpretation of written language so as to study whether doctors and patients are discussing and making important decisions on these platforms. This proposal has the potential to spark an entirely new generation of research in communication and health and will support efforts to get important scientific evidence about different treatment options into the hands of the patients who can most benefit from it.