Delivery of Health Services: Do Technological Features on Online Healthcare Platforms Matter?

Ms Zhou Fang
Dr Teo Hock Hai, Provost'S Chair Professor, School of Computing

  06 Dec 2018 Thursday, 02:00 PM to 03:30 PM

 Executive Classroom, COM2-04-02


The rapid development of information technology has facilitated the emergence of some novel healthcare-related platforms with the aim to improve the provision of health services. Despite the increasing efforts of the government in supporting healthcare-related platforms, there has been little research work investigating the effectiveness of these platforms in boosting better provision of health services. In this research, we focus on two types of online health care platforms that have burgeoned in recent years. One is health information exchange (HIE) which shares patient-level electronic health information across organization boundary; and another type is the online health consultation platform that allows patients to make online doctor consultation remotely.

According to the discussions about health service provision in the health economic literature, the principal-agent relationship between the physician and the patient inevitably leads to problems such as induced demand or unsatisfying input (i.e., efforts) from physician. In our research, we intend to examine the how the technological features on these platforms could influence health service provision and explore the role of physician agency in the process.

In study 1, we investigate the influence that HIE have on medical costs as a result of changes in the physician's provision of health services. We posit that information sharing through HIE could reduce redundant tests and thus leads to reductions in test costs. However, physicians acting as agents would induce more medicine demand to make up for the loss of income due to the reductions in test costs. Using hospitalization data from a general hospital in China that have joined a health information exchange platform, we construct a dataset whereby we can track the medical costs for each physician before and after the implementation of the HIE. Moreover, only some patients could use the platform after the implementation of HIE during the time span of our data, thus generating the treatment group and comparison group. In our empirical analysis, we combine difference-in-difference and propensity matching method to examine the impact of HIE and meanwhile control for time constant confounding factors and compositional changes in the treatment groups and the comparison groups. Our findings suggest that HIE could reduce test costs but increase medicine costs. Specifically, physician would reduce test costs only for patients who can use the HIE. Whereas, physician would induce more medicine costs for all patients no matter whether the patient uses the HIE platform.

In study 2, we examine how the voluntary incentives from patients in online health consultation platforms would influence the physician's efforts made to serve the patients in the platform. We postulate that voluntary incentives would facilitate the reciprocal actions (i.e., more efforts) taken by the physician. The reciprocal actions may reside not only at the dyad level but also at the macro level. At the dyad level, the physician receiving voluntary incentives would make more efforts in his interactions with the giver of incentives. At the macro level, general reciprocity would take place where physicians receiving incentives would be more willing to contribute toward other patients in general. We depict the physician's willingness to contribute as dynamic states which is theoretically updated gradually by pervious interactions with patients. We crawled the data from a leading Chinese online health consultation platform consisting of the consultation texts as well as the records of incentive giving actions. A content analysis method will be used to measure the physician's efforts in each consultation session. We develop a hidden Markov model (HMM) with latent states of willingness to contribute to model the physician efforts. Using the data from the platform, we will estimate the HMM to unravel the effects of voluntary incentives on physician efforts through two reciprocity mechanisms.

This research contributes to the literature on the value of online platforms in addressing problems related to the physician's provision of health care and offers important practical guidance on utilization of health information sharing and voluntary incentives in online healthcare platforms.