Exploring the Impact of Virtual Medicine on the Patient-Centered Medical Home

Abstract

The Patient-Centered Medical Home (PCMH) is a health care delivery model where a patient has an ongoing relationship with a personal physician who provides comprehensive and appropriate health care. Enhanced health care accessibility is one of the desired characteristics in a PCMH, which supports patient access to health care service 24/7, 365 days a year. Virtual medicine, such as secure messages, video diagnoses, and email communications, is one of the general tools used to realize this increased accessibility. Email communication was the form of virtual medicine studied in this thesis. This thesis explores the impact of email communication on patient visits to different health care services, such as primary care, specialty care, and urgent care. The data was provided by Group Health Cooperative (GHC) and was divided into two time periods: 2001 - 2004, before email communication implementation at Group Health, and 2009 - 2012, after email implementation. A Markov chain model was used to represent the transition data of patient visit encounters. Three metrics were used to study the impact of email communication: 1) number of patient visit encounters per member per month, 2) steady state probabilities of patient visit encounters, and 3) patient visit frequency to different health care services. The results indicated that, with email communication, the frequency of combined primary and specialty care visits decreased. This decrease in patient visits is consistent with the results of a study conducted at Kaiser Permanente Northwest (Zhou et al., 2007). The result in this thesis then showed that the decrease in overall visits was dominated by a decrease in primary care visits. The analysis showed an increase in specialty care visits with email communication. Possible causes for these observations are discussed in the thesis. A generalized, extendable PCMH simulation model was developed. It was used to validate the steady state probabilities calculated from the Markov chain model. A more important contribution of the simulation model is the capability to model limited availability of physician schedules, whereas the Markov chain model assumed infinite capacity of all the health care services.