The business model of medicine: modern health care's awkward flirtation with the marketplace.

• Author: Whalen, James P.
• Position: Controversy

A current advertisement on the radio says that a new diagnostic tool, electron beam tomography (EBT, a scanner that is much more rapid and capable of showing more detail than the better-known CT scanner), can detect lung cancer in its early stages, thereby possibly increasing the survival rate for this common form of cancer from 14 percent to 80 percent over a period of five years (Heart Check America n.d.). If the claim is true, this new way of diagnosing lung cancer that improves the survival rate almost sixfold will be a dramatic breakthrough in the treatment of this devastating disease.

This ad is an example of one of the many approaches to marketing medical goods and services that has become a major and accepted practice of the health care industry in the past ten years or so. Indeed, the corporate approach in general is becoming such a prominent feature of health care that medicine may be losing its identity as a social service and slipping into the business model of providing service. Is this change a good thing?

To answer this question, we must decide whether it is good to sell as much product as possible in order to maximize profits. Undoubtedly it is if you are General Motors. Is it good to purvey as many medical services as possible, however? Not in every case--and there are many reasons for this conclusion. They include evidence that high-powered advertising techniques used in the marketing of medical goods and services have a serious defect and that the advisability of allowing Darwinian competition to determine the availability of these goods and services is debatable. Also at issue is whether increasingly significant financial ties between industry and academic medicine best serve the national research agenda. In fact, these issues and many others call into question the wisdom of allowing the business model of medicine to be the operating paradigm for modern health care in the United States.

Medical Screening Sold Directly to the Public

Consider the radio ad for the fancy scanner. In the best capitalist tradition, the organization (a major university medical center) that purchased the air time hopes to create work that is well reimbursed and therefore highly profitable. Prospective patients, most of them probably smokers, will get a scan for themselves (no need for a doctor's order), and if the scan indicates probable cancer, they will be referred for further diagnosis and treatment at the sponsoring medical center. The $400 price for the EBT is not the prize. In fact, that sum may barely cover the costs of performing and interpreting the scan. The real bonanza lies in the treatment. In contrast to many medical services that are of proven benefit and therefore important to the health of the population, surgery for lung cancer is well reimbursed. So health care organizations, always looking for ways to improve their precarious bottom lines, have every financial incentive to find and treat people with lung cancer.

This arrangement would be splendid if it helped people, but the science of screening for disease in the hope of improving therapeutic outcomes reveals that determining the value of the procedure is far trickier than the radio ad would have us believe. The typical consumer of health services, however, generally has limited means to weigh the complexities.

For example, proper evaluation of the usefulness of an early diagnostic intervention requires understanding of some crucial concepts in screening (which can be defined as some type of medical test administered with a view toward early detection of disease). The first concept is "lead-time" bias, which is the belief that early detection has bestowed a benefit on mortality, when in fact the benefit is only apparent (Patz, Goodman, and Bepler 2000). In lung cancer, the lead-time bias operates as follows.

A microscopic focus of cancer begins to grow in the lung at a point in time--say, the year 1996. There is presumably a period of time (the length of time in this cancer, and in most cancers, is generally unknown) when the cancer grows before it causes a symptom, such as coughing up blood, that prompts medical attention. Suppose this symptom occurs in the year 2000 (these intervals are presented purely for the purpose of illustration--they may bear no resemblance to reality), which would be the point at which diagnosis and treatment are conventionally undertaken. With current means of treating lung cancer, death may occur in 2002, typically after a great deal of Sturm und Drang. A patient with this experience would be counted as having a two-year survival.

An entirely equivalent patient may undergo screening with EBT in 1998, get treatment, and live until 2002. This latter patient, however, is considered to have a four-year survival rate from the time of diagnosis, even though he or she lives no longer than the patient whose symptoms prompted medical attention two years after the onset of the cancer. The screening test has resulted only in apparent benefit: the second patient lives no longer than if there had been no screening. If such experience is what tends to happen with early screening, therapeutic results that show an advantage to screening are biased. The beguiling notion that early detection and treatment of lung cancer results in favorable mortality rates is just not true. Presumably the lung cancer spreads prior to becoming radiologically evident, rendering local removal futile.

Untrained observers may conclude confidently that survival in the screened patient has been increased by 100 percent because life after the diagnosis has been increased from two to four years. This difference is a rather dramatic advance, in anyone's estimation. Until now, however, no one has been able to demonstrate convincingly that lead-time bias is not at play here. In scientifically rigorous studies that deal with this difficulty (randomized, controlled trials), screening smokers with yearly chest X-rays simply has not resulted in lower mortality (Patz, Goodman, and Bepler 2000).

The idea of early detection (find the cancer before it spreads and cut it out) is so compelling, however, that researchers continue to look for effective screening techniques that are more sensitive than conventional X-rays. So they are using technology such as EBT to find lung cancers to which routine radiography is blind. Results? Unknown. Even though EBT is advertised on the radio and has leaked into the current screening armamentarium of medicine, no one knows if it works to save lives. In the words of researchers at Duke University Medical Center, scientifically valid studies have to be completed and carefully analyzed "before CT screening for lung cancer can be accepted as the standard of care" (Patz, Goodman, and Bepler 2000). Neither these authors nor any other serious cancer investigators are calling for cessation of research into this potentially important diagnostic technique. They are simply saying that investing huge amounts of dollars in an unproved approach is not warranted until clinical trials demonstrate a true reduction in mortality from lung cancer.

Other subtleties also cloud the assessment of a screening procedure. The...