Preventing overdiagnosis
Overdiagnosis is the diagnosis of a condition that would never have caused symptoms or harm in a person’s lifetime, so that finding it and treating it can only do harm, not good. It is different from a wrong diagnosis: the label may be technically correct, yet detecting it brings no benefit because the condition was never going to matter. Preventing Overdiagnosis is the name of both an international research conference and the wider movement built around this problem. This page describes that movement neutrally as a reference; it is not medical advice, and it is not affiliated with any conference organizer or prior owner of this domain.
What overdiagnosis means, and why it happens
The paradox of overdiagnosis is that it grows out of good intentions. Better scanners, more sensitive tests and lower thresholds for what counts as abnormal all find more. Some of what they find is real disease that will progress. But some is disease so slow, so mild or so static that the person would have lived out their life untroubled by it. Once it is found, though, it is hard to ignore, and the machinery of treatment engages. The healthy person becomes a patient, exposed to the side effects and anxiety of managing something that would never have surfaced.
This is why the movement argues we should not use more medicine than is warranted. Overdiagnosis is not a reason to avoid genuinely useful care; it is a reason to be careful about where the line of “disease” is drawn, because drawing it too wide converts healthy people into patients without making anyone better.
Overdiagnosis bias and its cousins
A recurring source of confusion is that screening statistics can make a test look effective when it is only finding harmless disease earlier. Three related effects explain how this happens, and telling them apart is central to the field.
| Effect | What it is | Why it misleads |
|---|---|---|
| Lead time bias | Diagnosis is made earlier, but the moment of death is unchanged | Survival “from diagnosis” looks longer only because the clock started sooner, not because life was extended |
| Length time bias | Screening preferentially catches slow-growing cases that were always going to do well | The screened group looks healthier because fast, aggressive cases surface between screens |
| Overdiagnosis bias | An extreme of length time bias: cases found that would never have caused harm at all | Every such case is counted as a “cure”, inflating the apparent benefit of screening |
Overdiagnosis is best understood as the far end of length time bias: not merely slow disease, but disease that would never have progressed. Distinguishing these effects requires long-term studies rather than survival figures alone, which is much of what the research community works on.
The conference and the “Too Much Medicine” alliance
The Preventing Overdiagnosis conference began in 2013 and has been held in a rotating set of countries since, drawing researchers, clinicians, patient advocates and policymakers to study how overdiagnosis arises and how to wind it back. Its early meetings were associated with academic partners including the University of Oxford, and it has run annually into the 2020s. The conference is deliberately international, because overdiagnosis is a feature of well-resourced health systems everywhere, not one country’s problem.
Running alongside it is the medical journal BMJ and its “Too Much Medicine” initiative, which publishes evidence on where diagnosis and treatment have expanded beyond benefit. Together these form a loose alliance with campaigns such as Choosing Wisely and researchers of quaternary prevention, sharing the same core claim: that a portion of modern medicine is doing more than is helpful, and that healthy people are being harmed as a result. The clinician and researcher H. Gilbert Welch has written widely on the subject, and his book Overdiagnosed is among the most cited popular accounts, making the case that early detection is not an unqualified good.
Is ADHD overdiagnosed?
Attention-deficit/hyperactivity disorder is one of the most debated examples, and it deserves careful handling. ADHD is a real condition with real impairment for many people, and effective help exists. At the same time, diagnosis rates have risen sharply in many countries, and researchers disagree about how much of that rise reflects better recognition of people who were previously missed versus a broadening of the boundary that now captures milder presentations, and normal variation in attention and energy, as disorder.
The honest position, and the one the overdiagnosis literature takes, is that both can be true at once. Some people are genuinely underdiagnosed and helped by treatment, while at the margins others may be labeled and medicated for traits within the normal range. Online discussion, including large communities on sites such as Reddit, has amplified both awareness and concern. None of this implies that any individual’s diagnosis is wrong, and nothing here is guidance about anyone’s own care; the point is only that the boundary of the diagnosis is contested and shaped in part by non-medical forces.
Winding it back safely
Research on reversing overdiagnosis focuses less on blame and more on practical fixes: narrowing overly broad disease definitions, being honest about the limits of screening, renaming very low-risk lesions so they sound less like cancer, and giving patients real numbers about their chances. These connect directly to the case for less medicine, to the specialty lists of Choosing Wisely: fewer tests, better care, and above all to shared decision-making, which lets an informed patient decide whether early detection is worth its downsides for them.
The movement also intersects with the marketing of illness, where the promotion of lifestyle drugs and the widening of diagnostic thresholds can turn features of ordinary life into treatable conditions. Preventing overdiagnosis, in the end, is not about diagnosing less for its own sake. It is about diagnosing the people who will actually benefit, and sparing everyone else the burden of being treated for something that was never going to harm them.