Date of Award

January 2012

Document Type


Degree Name

Medical Doctor (MD)



First Advisor

Richard J. Gusberg

Subject Area(s)

Medicine, Public policy, Economics


Few published studies have investigated physician knowledge of the cost of diagnostic tests. These reports have been limited by ambiguous wording (cost, reimbursement, or price), small sample sizes (mostly under 100, largest under 500), and narrow study populations (i.e. only family medicine residents or only emergency medicine physicians). We sought to characterize physician knowledge of the cost of diagnostic tests and test several hypotheses, including the relationship between accurate knowledge of cost and: stage of training (student vs. resident vs. fellow vs. attending physician), specialty, clinician degree (MD vs. PA vs. NP), ownership interest in practice, beliefs of the importance of cost consideration, and estimates of payer reimbursement ratios. We sent a survey assessing knowledge of costs, prices, beliefs, and payer reimbursement ratios to 3,400 clinicians in a major academic medical center. 1,700 completed surveys were returned (50% response rate). Specific tests included: CHEM-7, CBC, EKG, hypercoagulability panel, chest X-ray, chest CT, and chest MRI.Subjects were not able to accurately estimate the costs or prices of diagnostic tests and almost uniformly overestimated. Mean relative errors in cost ranged from a low of 204% in nurse-midwives to a high of 695% in PA students. Along the physician training pathway, performance improves only slightly, with medical students, residents, clinical fellows, and attending physicians showing mean relative errors of 393%, 289%, 261%, and 250%, respectively. Subjects overestimated the cost and price of more expensive tests to a substantially greater degree than they did for less expensive tests. For each $100 of diagnostic test cost, mean estimates of reimbursement from Medicare, commercial insurers, and self-pay patients were $83, $122, and $216, respectively. On the whole, subjects did not believe that their estimates were accurate (78%) and did not believe that their estimates were better than those of their peers (75%). Subjects do consider costs in their clinical decision-making (66%). Our study contributes to the literature by including subjects in multiple stages of training, various specialties, and three categories of licensed independent practitioners, allowing greater generalizability and between-group comparisons. The large sample size (1,700) and good response rate (50%) allow us the statistical strength to draw conclusions about the relationships under investigation. Physicians have very poor knowledge of the cost of diagnostic tests and the current educational pathway does not provide adequate education on this topic. We recommend formal education on costs and prices be incorporated into physician training at multiple levels.


This thesis is restricted to Yale network users only. This thesis is permanently embargoed from public release.