Adaptive Appointment Systems with Patient Preferences

点击翻译按钮获取中文摘要

Health care systems operate in a dynamic environment. Health service providers need techniques to first understand this environment, without increasing the data-collection burden, and subsequently techniques to respond to the dynamic inputs in an optimal fashion. The key objective of this project is to determine if time-stamp data can be used to reliably estimate dynamically evolving patient preferences, and to determine the extent to which an appointment system with dynamic preference information can outperform a similar system with static information. Intellectual merits of this project include the development of mathematical techniques for inferring and updating patient-preference categories from computerized appointments data, and developing booking control policies that utilize this information to improve access as well as clinic revenue. The project will enhance existing methods and develop new techniques for incorporating dynamic learning in non-stationary environments.Broader impacts include (a) practical tools for identifying and updating patient preferences from data routinely collected by clinics, (b) engineering education, and (c) improved health care management practices. This project will serve as a proof of concept of the benefits of using dynamic and adaptive methods in health care management. These techniques can be applied in other similar endeavors, for example, in developing individualized treatment and wellness programs based on partially-observed patient data and physician inputs. The project will help to partially fund doctoral studies of a student and provide opportunities for involving undergraduate students in research. Moreover, the data set obtained during the course of this project will be used to train health care professionals through post-professional courses. In addition to archival journal publications and conference presentations, results of this research will be disseminated by postings on the PI's research web page at http://www.me.umn.edu/labs/scorlab/, and by interactions with local health service providers.