医疗决策情景下医生利他性测量的实验经济学研究

随着实验经济学、行为博弈论的发展，传统经济学中的自利假设受到利他倾向的挑战。利他性在医疗决策中体现为医生行为偏离个人利益最大化，关注患者健康福利，被认为是医生绩效和医疗服务质量的关键决定因素之一，对公益性导向和满足社会公众利益最大化目标的落实具有重要意义。但是利他性难以测量影响了对医生利他性的研究和有效激励机制的建立。本研究基于显示偏好原理并采用独裁者博弈为基础设计的受控实验室实验模拟医生医疗决策情景，根据医生决策结果，分别基于医生效用函数计算和运用随机效用模型估算医生的利他性。结果显示医生决策过程中对患者健康效益给予了正向权重。医生的利他性存在异质性。基于医生效用函数计算的利他性均值为0.82，表明医生对患者健康效益的权重为82%。96%的医生利他性大于0.5，50%的医生利他性大于0.85。运用随机效用模型估算的利他性(患者健康效益与医生利润的边际替代率)为1.078>1，49%的医生边际替代率大于1，说明相比个人利益医生更加重视患者健康效益。两种方法下测量的利他性显著正相关(Spearman’s ρ=0.715, p<0.001)。提示本研究应用的医疗决策情景下利他性的测量方法可实现医生利他性的量化，基于两种方法测量的医生利他性可以相互印证。

Measurement of Physician Altruism in Medical Decision-Making Scenarios: An Experimental Economic Study

With the development of experimental economics and behavioral game theory, the assumption of self-interest in traditional economics is challenged by altruism. Altruism reflected in the medical decisions is the behavior of physicians deviating from the maximization of personal interests and focusing on patients' health benefits. Altruism is considered to be one of the key determinants of physicians' performance and medical service quality, which is of great significance to the implementation of public welfare oriented policies and the goal of maximizing social public welfare. But the difficulty of measuring altruism has affected the research and intervention of physicians' altruism. Based on the revealed preference theory, this study used a controlled laboratory experiment of dictator game to simulate physicians' medical decision making. According to the experimental data, we can estimate altruism based on the utility function and the random utility model respectively. The results show that physicians attach a positive weight to patients' health benefit. There is substantial heterogeneity in physician altruism. The mean value of altruism based on the utility function is 0.82, indicating that physicians had 82% weight on patients' health benefit. 96% of physician was greater than 0.5, and 50% was greater than 0.85. Based on the random utility model, the marginal rate of substitution between patients' health benefit and physicians' profit is 1.078. And, that of 49% of physicians is greater than 1, means that they pay more attention to patients' health benefit than their own interests. There is a significant positive correlation between the altruistic parameter and the marginal rate of substitution (Spearman's ρ=0.715, p<0.001). It is suggested that the altruism measurement method applied in this study is practicable, and the two methods can confirm each other.