What’s Really Behind Overtesting?

Ocular Imaging Room

Photo: Ocular Imaging Room at the University of Pittsburgh Medical Center (UPMC) Eye Center
Courtesy of Joshua A. Rheinbolt, MD, formerly of UPMC

Patient expectations, distorted by insurance structure, influence physician behavior, leading to overtesting, even in the absence of fee-for-service payments and malpractice concerns.

As we move into the second decade of the 21st century, the public is more convinced than ever that the US healthcare system suffers from an obsession with medical tests and procedures. The Institute of Medicine estimated up to $765 billion per year, or 6 percent of the nation’s GDP, is spent on low-value tests and treatments. In the case of medical testing, various studies estimate as much as 30 percent of it is not cost effective, a phenomenon commonly referred to as overtesting.

What causes overtesting? Popular media often blame two factors: a fee-for-service payment system and defensive medicine. Both factors originate from the supply side, with a view that physicians unilaterally order medical tests to enrich and protect themselves. Echoing this supply-side view, a group of health economists have developed a body of work known as the supplier-induced demand (SID) literature.

Another group of health economists argue the demand side also plays a crucial role. Joseph Newhouse, who, between 1971 and 1986, led a nation-wide, multi-million-dollar experiment known as the “RAND Health Insurance Experiment,” champions this group. The RAND experiment shows insurance coverage is behind the over-provision of health services and leads to welfare loss. This demand-side view has become more prominent over the past few years, due to broad public discussions surrounding healthcare reform, and was articulated in a September 2009 Atlantic article, “How American Health Care Killed My Father” by David Goldhill:

Every time you walk into a doctor’s office, it’s implicit that someone else will be paying most or all of your bill; for most of us, that means we give less attention to prices for medical services than we do to  prices for anything else. Most physicians, meanwhile, benefit financially from ordering diagnostic tests, doing procedures, and scheduling follow-up appointments. …[Y]ou have a system where physicians can, to some extent, generate demand at will.

According to the demand-side view, health insurance distorts the price signal crucial for efficient resource allocation in the health service market. Accordingly, one way to reduce overtesting would entail increasing patients’ out-of-pocket expenses (e.g., copayments) and encouraging more “responsible” spending behavior.

Which side do we take? We (Tinglong Dai of Johns Hopkins University; Mustafa Akan and Sridhar Tayur of Carnegie Mellon University) started our effort with in-depth observations of imaging-testing practice and interviews with medical doctors, because we believe, ultimately, doctors are the ones who make diagnostic-testing decisions. According to our medical collaborators, both sides—supply and demand—are rather “bizarre” and out of touch with the medical reality.

Motivated by our extensive conversations with them, our work brings together both sides. In our model, patients are strategic in choosing their frequency of medical visits, and physicians respond to patient demand by choosing the level of testing. In other words, we believe both sides have control over the demand for health services: the demand side (patients) controls the frequency of care, whereas the supply side (physicians) controls the intensity of care.

To our best knowledge, ours is the first paper to analytically investigate financial (insurance and reimbursement), operational (system throughput and congestion), and clinical (service quality) incentives behind physicians’ test-ordering behavior in an outpatient setting. Our paper, titled “Imaging Room and Beyond: The Underlying Economics Behind Physicians’ Test-Ordering Behavior in Outpatient Services,” reveals several interesting aspects that help us understand physicians’ decision making leading up to the phenomenon of overtesting.

First, health economists often speak of “demand” for health services as a single-dimensional quantity, and have characterized supply-demand curves often at highly abstract levels. They tend to treat patients’ waiting as the health provider’s unilateral, self-concocted decision variable. The strategic queuing framework allows us to treat patient waiting time as an output variable that depends on the physician and patients’ joint strategic decision-making.  As a result, our research breaks down patient demand for outpatient services into two dimensions: (1) frequency of visits and (2) intensity of care per visit.

Although this distinction may sound subtle, the implication is tremendous. Our research shows, for example, increasing patients’ out-of-pocket expenses may not necessarily help reduce overtesting, because as patients make decisions about their frequency of visits, a larger amount of the copayment creates a friction and induces them to pay a visit only when they have a sufficiently high expectation for service quality. The quality expectation, in turn, drives physicians to order more tests than deemed socially efficient.

Our research highlights the role of health insurance structure in influencing physicians’ service decisions. The implication is that focusing solely on insurance coverage may be counterproductive. Examining the structure of insurance plans is as important.

Second, although we derive most of our findings by assuming physicians have the power to influence health-service pricing, our key insights extend to the scenario in which all the prices are fixed. This result is rather surprising: physicians may choose to overtest even if they do not benefit financially from testing. This result holds also in the absence of malpractice concerns. As surprising as the result may seem, it is aligned with the situations we have observed at the University of Pittsburgh Medical Center (UPMC) and various other academic hospitals, as described by an ophthalmologist: “When we order three tests for a patient, we understand the insurance firm will reimburse only one test, and the other two will generate no additional revenue.”

What’s behind this counterintuitive finding? The answer lies in both the supply and demand sides: patients determine their visiting pattern based on their insurance structure; physicians, in turn, have to make testing decisions by incorporating patient-level decisions. The dynamics are present even when service fees are fixed. As long as the friction is strong enough to prevent patients from seeing their physicians at a socially efficient frequency, physicians would have to answer patients’ expectations for excessive quality through overtesting.

Third, our research also incorporates other aspects emerging out of an outpatient-care setting, such as malpractice concerns, patient heterogeneity in diagnostic precision, and disparities in health insurance coverage. These aspects validate the robustness of our findings.

In summary, what’s really behind overtesting? Our research suggests patient expectations, distorted by insurance structure, influences physician behavior leading to overtesting, even in the absence of a fee-for-service payment system and malpractice concerns. We tease out “why” using a strategic queuing framework, incorporating a finer insurance structure, and accounting for patient and insurance heterogeneity in our analysis.

At a broader level, our research, by examining the cost, quality, and access of healthcare using a strategic queueing approach, answers an urgent call for studying “big topics” in the US healthcare industry. Given that our community has invested so much effort to analyze and solve supply-demand mismatch problems of virtually any scale, we believe operations management researchers can address significant healthcare issues and contribute to the ongoing conversations about healthcare reform. Studying overtesting will turn out to be just one of many such efforts.



Dai T, Akan M, Tayur S (2016) Imaging Room and Beyond: The Underlying Economics Behind Physicians’ Test-Ordering Behavior in Outpatient Services. Manufacturing & Service Operations Management. Forthcoming. http://dx.doi.org/10.1287/msom.2016.0594