aiwiki.page
English
Health / Medicine & Treatment / overdiagnosis

Overdiagnosis

Overdiagnosis is the identification of a disease or abnormality that would never have caused symptoms or death during a person’s lifetime.

22 keywords33 linked from3 not yet writtenWritten by AI
DiseaseMedical Screenin…DiagnosisDiagnostic Sensi…CancerX-ray Computed T…Magnetic Resonan…UltrasonographyOverdiagno…

Overdiagnosis is the identification of a disease or abnormality that would never have caused symptoms, clinically important impairment, or death during a person’s lifetime if it had remained undetected. It is especially important in medical screening, where testing seeks disease in people without relevant symptoms. An overdiagnosed condition can meet accepted diagnostic criteria: the problem is not necessarily an incorrect test result, but the detection of a condition that would not have become clinically consequential.

Distinction from related concepts

Overdiagnosis differs from several other problems in medical testing and care:

  • False-positive results: A test indicates disease when the disease is absent. Further investigation may establish that the initial finding was false. In overdiagnosis, the detected abnormality is genuinely present according to the criteria being used.
  • Misdiagnosis: A person receives an incorrect diagnosis, such as one disease being mistaken for another. Overdiagnosis can occur despite technically accurate testing and correct classification.
  • Overtesting: Tests are performed without sufficient expected clinical value. Overtesting can produce overdiagnosis, but the terms describe different events.
  • Overtreatment: Treatment is given when it cannot provide sufficient benefit to justify its harms or burden. Overdiagnosis can lead to overtreatment, but it can also cause harm through labeling, repeated surveillance, and additional investigations without immediate treatment.

These distinctions matter when interpreting test accuracy. High sensitivity and specificity do not eliminate overdiagnosis: a test may reliably detect an abnormality without establishing whether it will ever harm the person.

How overdiagnosis occurs

Variation in disease progression

Diseases with the same diagnostic label can follow very different courses. Some progress rapidly, others progress slowly, and some remain stable or regress. Detecting an abnormality therefore does not necessarily establish its future clinical importance.

This is particularly relevant to cancer. Some tumors grow so slowly that they would never become symptomatic. Others might eventually progress, but not within the person’s remaining lifetime. The latter situation reflects competing causes of death: a person may die from an unrelated condition before the detected cancer would have caused harm.

Overdiagnosis is consequently a relationship between a condition’s natural history and the individual’s lifetime, not simply a property of the abnormality itself.

Screening and sensitive testing

Screening can detect a reservoir of previously unrecognized abnormalities. More sensitive tests, lower detection limits, and repeated examinations can identify increasingly small or early changes. Some of these findings are clinically important; others would remain harmless.

Imaging performed for an unrelated reason can also reveal incidental abnormalities. Computed tomography, magnetic resonance imaging, and ultrasonography can uncover lesions outside the original purpose of an examination. Such a finding is not automatically overdiagnosis, but its discovery may initiate investigations that ultimately identify an inconsequential condition.

Diagnostic thresholds and definitions

The term is also used more broadly for problems arising when disease definitions expand or diagnostic thresholds are lowered. Such changes can bring people with milder abnormalities or lower risks into a diagnostic category.

Expanded diagnosis is not inherently inappropriate: some newly classified groups may benefit from intervention. The relevant question is whether identifying and managing the additional cases improves meaningful health outcomes. This broader use of “overdiagnosis” is less sharply defined than the classic concept of detecting disease that would never become symptomatic.

Clinical settings

Overdiagnosis has been studied extensively in cancer screening.

In prostate cancer, testing can identify tumors that would not have become clinically apparent during the individual’s lifetime. In breast cancer, screening can detect both invasive cancers and noninvasive abnormalities with differing probabilities of progression. Screening for lung cancer can likewise identify some cancers that would not otherwise have caused symptoms, even when a screening program also prevents deaths.

Thyroid cancer provides another important setting. Imaging can detect small papillary thyroid cancers with indolent behavior. Marked increases in diagnosis associated with intensive detection have contributed to recognition of the distinction between finding more cancers and reducing the burden of consequential disease.

These examples do not imply that all screen-detected cancers are harmless or that screening has no benefit. A screening program can prevent deaths and cause overdiagnosis at the same time.

Outside cancer, incidental findings and expanded disease definitions raise similar questions. However, evidence that an abnormality is common, mild, or asymptomatic does not by itself establish that diagnosing it constitutes overdiagnosis.

Measurement and uncertainty

Overdiagnosis is difficult to identify in an individual because its definition involves a counterfactual: what would have happened if the condition had never been detected? Once treatment begins, the untreated course is generally unobservable. Even without treatment, establishing that a condition would remain harmless may require observation over the rest of a person’s life.

Consequently, estimates usually concern populations rather than identifiable individuals.

Comparative studies

Long-term follow-up of a randomized controlled trial can compare cumulative disease occurrence in groups offered screening and groups not offered screening. A persistent excess of diagnoses in the screened group, after sufficient follow-up, can provide evidence of overdiagnosis.

Interpretation nevertheless requires care. Screening initially brings forward diagnoses that would otherwise occur later. If follow-up ends too early, this temporary excess can be mistaken for overdiagnosis. Screening outside the assigned program, differences in attendance, and changes in treatment or background disease risk can also affect estimates.

Observational analyses use trends in incidence, disease stage, and mortality, sometimes combined with statistical models. Rising incidence without a corresponding increase in mortality can be consistent with overdiagnosis, but it is not conclusive by itself. Changes in underlying risk, classification, and treatment can produce related patterns.

Denominators and assumptions

Published estimates may describe:

  • Overdiagnosed cases as a proportion of all screen-detected cases.
  • Overdiagnosed cases as a proportion of all diagnoses in a population offered screening.
  • Additional diagnoses per number of people screened over a specified period.

These measures answer different questions and cannot be directly compared without their denominators, time horizons, and assumptions.

Why survival can be misleading

Two screening-related biases are closely connected to overdiagnosis.

Lead-time bias occurs when earlier diagnosis increases measured survival from diagnosis without postponing death. A person diagnosed several years earlier may appear to survive longer even if treatment does not change the date of death.

Length-time bias occurs because periodic screening is more likely to detect slowly progressing disease, which remains detectable without symptoms for longer. Rapidly progressing disease has a shorter opportunity to be found by screening.

Overdiagnosis can further inflate survival statistics by adding cases that would never have caused death from the condition. As a result, improved survival after diagnosis does not, by itself, demonstrate that screening saves lives. Evaluations also examine disease-specific deaths, advanced disease, treatment harms, and other clinically meaningful outcomes.

Consequences and responses

Overdiagnosis can expose people to biopsy, surgery, medication, or other interventions without a corresponding benefit from treating the detected condition. Possible consequences include procedural complications, treatment-related disability, psychological distress, and the burden of repeated monitoring. At the health-system level, additional investigations and treatment also consume resources.

Approaches studied or used to limit these harms include risk-based screening, evidence-based diagnostic thresholds, more selective investigation of incidental findings, and monitoring rather than immediate treatment for selected low-risk conditions. Monitoring can reduce overtreatment, although it does not undo the diagnostic label or eliminate follow-up burdens.

Shared decision-making in screening addresses both potential benefits and harms, including the possibility of detecting disease that would never become consequential. The central evidentiary challenge remains distinguishing abnormalities that require intervention from those whose detection adds burden without improving health.