Peer Effects in the Emergency Department on Productivity and Patient Choice

描述（由申请人提供）：医生可以通过同行的存在而受到社会激励的概念是直观的。然而，医生之间这些“同伴效应”的性质以前没有被研究过。该项目的目的是研究社会网络对个别医生的行为和对卫生保健组织的整体生产力的影响。具体来说，我们将评估急诊科医生在一起工作时对彼此的影响。具体目标有三个方面。首先，我们将衡量同伴对个人生产力的影响——以及护理速度、护理质量和教学方面的影响。我们预测在更可观察的维度（如护理速度）和更不可观察的维度（如护理质量）之间的权衡，我们预计同伴效应在社会联系更紧密的同伴中会更强。其次，我们将研究同伴效应对医生选择患者的影响。个别医生可能想要选择容易的病人（即“精挑细选”），但社会网络可能会通过社会压力和适当的工作协调来缓解这种情况。第三，我们将研究在不允许医生选择自己的病人的新制度下对急诊科整体生产力的影响。初步结果表明，急诊医生的工作效率不同，而且他们通常对某些类型的病人有强烈的偏好，这些偏好可能与同行相同，也可能与同行不同。鉴于此，我们将使用我们的研究设计的两个互补的特点来完成我们的具体目标。首先，我们有详细的，纵向的数据，关于个别医生和他们所见的病人，包括在医生团队中相互合作的足够的变化。其次，被研究的急诊科将在2011年5月进行一项自然实验，在该实验中，病人分配规则将禁止医生选择自己的病人。这项研究将揭示医疗机构应该如何设计隐含在团队结构中的社会网络，以及管理团队行为的规则。具体的影响包括组织应该如何让同事之间相互匹配，他们应该在牺牲不太透明的措施的情况下，在同事之间制定透明的生产力衡量标准，以及他们应该在多大程度上约束隐含的社会行为，比如在医生中选择病人。

DESCRIPTION (provided by applicant): The notion that physicians can be socially incentivized by the presence of their peers is intuitive. However, the nature of these "peer effects" among physicians has not previously been studied. The objective of this project is to study the effect of social networks on the behavior of individual physicians and on the overall productivity of health care organizations. Specifically, we will evaluate the effects of emergency department physicians on each other as they work together. Specific aims are threefold. First, we will measure peer effects on individual productivity - along dimensions of speed of care, quality of care, and teaching. We predict tradeoffs between more observable dimensions (e.g., speed of care) and less observable ones (e.g., quality of care), and we expect that peer effects will be stronger among more socially connected peers. Second, we will study peer effects on the choice of patients by physicians. Individual physicians may want to choose easy patients (i.e., "cherry pick"), but social networks may mitigate this by social pressure and appropriate coordination of work. Third, we will study the effect on overall emergency department productivity under a new system that disallows physicians from choosing their own patients. Preliminary results show that emergency physicians differ in their productivity and that they often have strong reported preferences for certain types of patients, which they may share with or differ from their peers. Given this, we will use two complementary features of our research design to accomplish our specific aims. First, we have detailed, longitudinal data on individual physicians and the patients that they see, involving sufficient variation in the teams of physicians working with each other. Second, the emergency department being studied is undertaking a natural experiment in May of 2011 in which the rules of patient assignment will disallow physicians from choosing their own patients. This research will shed light on how health care organizations should design their social networks implicit in team structures and the rules that govern behavior within teams. Specific implications include how organizations should match peers with each other, how transparent they should make productivity measures among peers at the expense of less transparent measures, and how much they should constrain implicitly social behaviors such as the choice of patients among physicians. PUBLIC HEALTH RELEVANCE: Like any other type of worker, physicians may be influenced in their behavior by the presence of their colleagues or "peers." This research project studies such "peer effects" on productivity and choice of patients. In addition to studying the behavior of physicians as they work among different colleagues, this project will study the overall impact on peer effects in a natural experiment in which the same emergency physicians are no longer allowed to choose their own patients.