Historically, epidemiology has been directed toward confirming and controlling widely spread infectious diseases. In a research paper just published in the NewEnglandJournalofMedicine, my colleagues and I point out that epidemiologists have another task: to identify and control epidemics caused by the medical community. The setting for the study was South Korea. Over the past 20 years, the incidence of thyroid cancer here has increased to 15 times its initial rate. Nowhere else in the world has cancer increased so rapidly. We have been taught to look for biological explanations for the dramatic increase in incidence, perhaps a new pathogen, perhaps environmental exposure. In Korea, however, we see a different picture, an epidemic caused by diagnosis. In 1999, the Korean government launched a nationwide medical screening program with the primary goal of reducing cancer and common diseases. Although it did not include thyroid cancer screening, it required one simple step: an ultrasound of the neck. Ultrasound equipment is available in hospitals and in many doctors’ offices. Hospitals and doctors see it as an inexpensive supplemental test to government programs, and therefore encourage patients to get tested. And patients were receptive, especially because the government, the medical community, the news media and cancer “survivors” all praised the benefits of early detection. But it also unexpectedly highlighted the great danger of early diagnosis: an otherwise rare cancer became the most common cancer in Korea. Where are all the new thyroid cancer patients coming from? They have actually always existed. As early as 1947, pathologists found that, although rarely the cause of death, thyroid cancer was frequently found in the deceased during autopsies. Since then, studies have found that more than one-third of adults have thyroid cancer. Almost all of them have a tiny “papillary thyroid” cancer, and many have no symptoms during their lifetime. Unless the person undergoes an ultrasound. In fact, in Korea, almost all newly diagnosed thyroid cancers are papillary cancers. How can we know that this is not a real epidemic? Because the number of people dying from thyroid cancer in Korea has not changed. If ultrasound is saving patients’ lives, then the death rate should be decreasing, and if the epidemic is spreading, the death rate should be slowly increasing – yet the numbers don’t ebb and flow at all. The diagnostic presentation of the epidemic did no one’s health any good, and it needlessly diverted resources and needlessly frightened patients. The biggest problem, however, is that it has spawned overtreatment within the healthcare system. Most patients who are diagnosed with thyroid cancer have their thyroid gland removed. Yet the thyroid is an important gland that secretes hormones that control metabolism. If removed, patients may need thyroid replacement therapy for the rest of their lives, and it may take a while for doctors to find the right dose for each patient. In the meantime, patients suffer the consequences of low or high thyroxine levels, including fluctuations in energy and weight. This procedure can also produce some complications that are not very common. In Korea and the United States, about 10% of patients develop problems with calcium metabolism and about 2% develop vocal cord paralysis. In addition, like any type of surgery, it can have life-threatening consequences such as blood clots in the lungs, myocardial infarction and stroke. Approximately two patients die for every 1,000 thyroid cancer surgeries. It’s rare, but it does happen. Will what happens in Korea also happen in the United States? Absolutely. Thyroid cases in the U.S. have also tripled since 1975, despite no concerted effort to promote screening. To reverse this trend, we need to actively discourage early screening for thyroid cancer. The notion that there are benefits to early screening is so deeply rooted and appealing that many people believe that a little screening can only be beneficial. Yet this is not the case. The Korean experience shows that trying to diagnose cancer early has negative implications: overdiagnosis and overtreatment. Thyroid and prostate cancers are the most serious problems, yet they also exist for lung, breast, skin, and kidney cancers. And of course there is the anxiety about testing, which is not good for anyone’s health. Of course, screening makes sense in some cases, especially for patients at really high risk for cancer: patients who have had multiple deaths from cancer in their family. People who are at moderate risk for the disease but have a long enough life expectancy to experience potential benefits in the future – and those who are willing to accept the possibility of being compromised by a treatment they don’t need now – might also decide that it makes sense to get screened. Nonetheless, those interested in early screening should also consider the question of how early is too early. Of course, we would rather diagnose a small breast lump as cancer earlier than sit back and watch it develop into a large lump. However, I am afraid that it would be too much to infer that we should look for tiny lumps that can only be seen with a microscope. Many of the thyroid carcinomas found in Korea are less than a centimeter in size. If we go looking for cancerous lesions, we can always find more cases. But beyond a limit, we diagnose so many problems that it’s better to leave them there undetected. In short, it’s in the patient’s interest to make doctors less eager to find early cancerous lesions. This is exactly what epidemiology is about. Rather than focusing on controlling infectious diseases, too many epidemiologists want to look for small effects on health from environmental exposure factors, or worse, the possible effects of small genetic differences. Perhaps they should be focusing on the more important risk to human health: epidemics caused by the medical community.