Three Studies of Variation in Medical Decision-Making and Care Delivery by Physicians.

Abstract

In this dissertation, I examine the effects of select non-financial factors on the organizational and medical decisions of primary care practice managers, specialist consultants, and Veterans Health Administration (VA) clinicians. In my study of regulations governing non-physician clinicians’ (NPCs) practices and how they affect Medicaid participation in physician practices, I find heterogeneous effects across non-physician clinician types and across practices. Relaxing regulations on physician assistants leads to small increases in Medicaid participation among many primary care practices. However, relaxing regulations on nurse practitioners leads to reduced Medicaid participation among practices seeing few Medicaid patients and increased Medicaid participation among practices seeing more Medicaid patients. Thus policymakers should not overlook that relaxed NPC regulations will affect current Medicaid participants’ willingness to see Medicaid patients and may not lead to improved access overall. In my study of medical consultations, I find significant effects of previous consultants’ care decisions on the next consultant’s. Specifically, my findings identify medical consultants’ dual-roles: they provide both active management care support and specialty-specific insights. Moreover, patterns of consultants’ care decisions indicate the consultants understand their roles as active managers and calibrate their care to meet what they perceive to be the patient’s needs. The appropriateness of consultants’ care decisions may be increased by ensuring the case information they receive is as complete as possible. Finally, in my study of variation in care quality for veterans with diabetes, I find moderate variation in blood pressure control effectiveness across facilities. Available facility resources and coordination structures do not appear to affect quality measure performance, yet physicians’ care quality is significantly influenced by their peers and local care environments. These findings offer valuable insights to VA administrators developing policies and protocols to improve diabetes care quality both within individual facilities and across the VA. By integrating my key findings into policy and organizational structures at different levels of the health care system, meaningful steps can be taken to modify the environments in which physicians practice, thereby facilitating the decisions that promote health and efficient use of medical care resources.