{
  "id": "medical-screening",
  "title": "Medical Screening",
  "lang": "en",
  "source_lang": "en",
  "summary": "Medical screening identifies unrecognized disease or elevated risk in people without relevant symptoms, using organized testing pathways whose benefits must outweigh their harms.",
  "category": "health",
  "subcategory": "medicine",
  "type": "concept",
  "aliases": [
    "Health screening",
    "screening in medicine"
  ],
  "generated_at": "2026-11-02",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/medical-screening/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/medical-screening/",
      "json": "https://aiwiki.page/zh/medical-screening.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "Medical screening is the systematic use of tests, examinations, or questionnaires to identify people who may have an unrecognized disease, a precursor lesion, or an elevated risk of a future health problem. Its defining feature is that the people tested have not presented with symptoms of the condition being sought. Screening separates an apparently healthy population into groups requiring different next steps; it does not ordinarily establish a definitive diagnosis. A positive result usually requires further assessment, while a negative result reduces—but does not eliminate—the possibility of the target condition.\n\nScreening is an important instrument of public health because some conditions can be treated more effectively before symptoms appear, and some precursor abnormalities can be treated before disease develops. However, earlier detection is not automatically beneficial. Testing can produce false results, identify abnormalities that would never cause illness, and expose people to unnecessary procedures. The value of screening therefore depends on an entire pathway: selecting an appropriate population, using a reliable test, confirming abnormal results, providing effective intervention, and measuring outcomes that matter to patients.\n\nScreening encompasses diverse activities, including newborn blood-spot testing, cervical screening, blood-pressure measurement, and testing for selected infections during pregnancy. These activities differ in purpose, evidence, and organization. Their common scientific question is whether offering the pathway to a defined population improves health sufficiently to justify its physical, psychological, financial, and social consequences."
    },
    {
      "heading": "Historical development and principles",
      "text": "Screening developed from several overlapping traditions: examinations of workers and military recruits, campaigns against communicable diseases, laboratory testing, and the expansion of preventive clinical care. Mass chest radiography became an important tool in twentieth-century efforts to detect tuberculosis, particularly before more targeted approaches and effective treatment systems changed its role. Cervical cytology, associated especially with George Papanicolaou’s work during the first half of the twentieth century, demonstrated how examining apparently healthy people could reveal precancerous cellular changes.\n\nA decisive development was the ability to connect an early laboratory finding with an effective intervention. In the early 1960s, Robert Guthrie’s bacterial inhibition assay made large-scale screening of newborn blood spots for phenylketonuria practical. This inherited disorder can cause severe neurological injury if untreated, yet affected newborns may initially appear healthy. Early dietary treatment provided a strong rationale for detecting the condition before symptoms. The example established a durable model: the screening test is useful because it opens a time-sensitive opportunity to alter the disease’s course.\n\nIn 1968, James Maxwell Glover Wilson and Gunnar Jungner published Principles and Practice of Screening for Disease for the World Health Organization. Their framework emphasized that the target condition should be an important health problem, its natural history should be understood, an acceptable test should exist, and appropriate diagnostic and treatment facilities should be available. It also called for an agreed policy on whom to treat, an economically balanced programme, and continuing rather than one-time case finding. These principles were a response to the growing technical ability to test large populations without equivalent assurance that testing improved health.\n\nLater frameworks retained these foundations while placing greater emphasis on programme quality, informed choice, equity, and evidence of benefit. Screening became an important subject within evidence-based medicine, particularly as trials revealed that some plausible early-detection strategies did not reduce deaths. The distinction between a technically impressive test and a beneficial programme became central.\n\nTerminology also evolved. Screening is generally classified as secondary prevention, because it seeks disease or its precursors before clinical presentation. Some programmes, such as removal of colorectal adenomas, also prevent future invasive disease. Screening differs from public-health surveillance, which primarily tracks population patterns, and from diagnostic investigation, which examines a clinical problem already suspected because of symptoms or signs. The same instrument can serve any of these purposes; the purpose and population, rather than the device alone, determine whether its use constitutes screening."
    },
    {
      "heading": "Target conditions, natural history, and screening strategies",
      "text": "The biological basis of screening is an interval during which a condition is detectable but has not yet produced recognizable symptoms. This preclinical detectable phase may last days, years, or decades. Its length depends on disease biology and on the test used. A molecular assay may detect an infection before an antibody test becomes positive, while an imaging method may reveal a lesion only after it reaches a certain size. Screening is useful only when detection during this interval enables an intervention that produces a better outcome than intervention after ordinary clinical presentation.\n\nThe natural history of the target condition is therefore fundamental. Some abnormalities progress predictably; others remain stable or regress. Persistent infection with certain types of human papillomavirus can lead to cervical precancer and eventually invasive cancer, but most infections do not follow that course. Similarly, not every colorectal adenoma becomes malignant. Programmes must distinguish, as far as possible, abnormalities whose detection creates a meaningful opportunity for prevention from findings that mainly create additional monitoring and treatment.\n\nPopulation selection affects both benefit and harm. Universal screening offers testing to everyone within a defined group, such as newborns in a jurisdiction. Selective screening targets people with a particular exposure, family history, or other risk factor. Screening for lung cancer with low-dose computed tomography, for example, has been evaluated chiefly in populations with substantial smoking exposure, not indiscriminately in all adults. Genetic predispositions such as Lynch syndrome can justify surveillance pathways that differ substantially from those used for average-risk populations.\n\nOrganized programmes identify eligible people, issue invitations, specify testing intervals, track results, and audit outcomes. Opportunistic screening occurs when testing is offered during contact with health services for another reason. Opportunistic approaches can reach people efficiently, but may repeatedly test frequent service users while missing those with limited access. A programme may also use sequential testing: a relatively inexpensive first test identifies a smaller group for a more specific or invasive second test.\n\nThe interval between rounds is a biological and operational decision, not simply a matter of convenience. Rapidly developing disease may arise between scheduled screens, producing an interval case. Very frequent testing can increase false alarms and resource use without proportionate benefit. Starting and stopping ages also reflect the changing frequency of disease, the delay before benefit appears, competing causes of illness, and treatment burden. Screening schedules consequently differ by condition, risk group, and health system rather than following a single general rule."
    },
    {
      "heading": "Test accuracy and interpretation",
      "text": "Screening tests include blood tests, imaging, physiological measurements, questionnaires, and examination of cells or tissues. Their performance has several dimensions. Analytical validity concerns whether an assay measures its intended substance reliably. Clinical accuracy concerns how well the result distinguishes people with and without the target condition. Clinical utility concerns whether using the test improves outcomes. These dimensions are related but not interchangeable: a highly precise measurement of a biomarker may have little value if the marker does not reliably identify actionable disease.\n\nSensitivity and specificity describe accuracy relative to a reference standard. Sensitivity is the proportion of people with the condition who test positive; specificity is the proportion without it who test negative. If TP, FN, TN, and FP denote true-positive, false-negative, true-negative, and false-positive results, respectively, then\n\n\\[\\text{Sensitivity}=\\frac{TP}{TP+FN}, \\qquad \\text{Specificity}=\\frac{TN}{TN+FP}.\\]\n\nA threshold determines how a continuous result becomes “positive” or “negative.” Lowering the threshold commonly increases sensitivity at the cost of more false positives. The appropriate trade-off depends partly on the consequences of missed disease and partly on the consequences of further investigation.\n\nPredictive values answer a different question: given the result, how likely is the condition to be present or absent? They depend strongly on the condition’s prevalence in the tested population. For example, consider 10,000 people among whom 1% have the target condition. A test with 90% sensitivity and 95% specificity would identify approximately 90 true positives but also 495 false positives. Of 585 positive results, only about 15% would represent the condition. This is not a contradiction of the test’s stated accuracy; it is the mathematical consequence of applying it where disease is uncommon.\n\nMore generally, if \\(p\\) is prevalence, \\(Se\\) sensitivity, and \\(Sp\\) specificity, the positive predictive value is\n\n\\[PPV=\\frac{Se\\,p}{Se\\,p+(1-Sp)(1-p)}.\\]\n\nA negative result can likewise be reassuring without being conclusive. Its meaning depends on underlying risk, the timing of testing, and whether the condition was detectable at that point.\n\nAccuracy estimates can also be distorted. Evaluating a test only in people with advanced disease and clearly healthy controls may exaggerate performance in an asymptomatic population. If only people with positive screens receive definitive investigation, verification bias can occur. For these reasons, accuracy studies require representative participants, appropriate reference standards, and transparent handling of uncertain results. Confirmatory testing may use a different biological target or method to reduce correlated errors, but even a completed diagnostic pathway is not infallible."
    },
    {
      "heading": "Demonstrating benefit and avoiding misleading evidence",
      "text": "A screening programme is evaluated by comparing outcomes among people offered screening with outcomes under an appropriate alternative, often usual care. A randomized controlled trial is particularly valuable because random allocation reduces systematic differences between groups. Depending on the condition, relevant outcomes include disease-specific death, all-cause death, severe complications, functional ability, and quality of life. Finding more cases, detecting smaller tumours, or increasing the proportion diagnosed at an early stage does not by itself establish benefit.\n\nLead-time bias occurs when diagnosis is advanced without postponing death or disability. Suppose a cancer would ordinarily be diagnosed at age 67 and cause death at age 70. If screening detects it at 62 but death still occurs at 70, survival measured from diagnosis rises from three to eight years despite no extension of life. For this reason, comparisons of survival among screen-detected and symptom-detected patients are unreliable measures of screening effectiveness.\n\nLength-time bias arises because periodic screening is more likely to detect slowly progressing disease than rapidly progressing disease. A slowly growing tumour remains in the detectable preclinical phase for longer and therefore has more opportunities to be found. Its comparatively favorable prognosis may reflect its biology rather than an effect of screening. Participation introduces another problem: people who attend screening may differ from nonparticipants in smoking, income, other preventive care, and general health. Comparing these groups directly can therefore exaggerate benefit.\n\nTrials usually analyze people according to their original allocation, an approach called intention-to-treat analysis. This estimates the effect of an invitation or screening policy under the attendance and follow-up conditions of the study. Nonattendance in the screening group and testing in the comparison group can reduce the observed difference. Analyses restricted to attenders require additional assumptions and can reintroduce selection bias.\n\nEffects also need to be expressed in absolute terms. A hypothetical reduction in disease-specific deaths from 5 to 4 per 1,000 people over ten years is a 20% relative reduction but an absolute reduction of 1 per 1,000. Roughly 1,000 people would need to be offered the evaluated pathway to prevent one such death over that period, assuming the estimate is valid. Harms must be measured over comparable periods.\n\nLong follow-up is often necessary because screening can cause immediate investigations while benefits appear much later. Observational studies and models can supplement trials, especially when technologies or treatment change, but depend on assumptions about attendance, disease progression, and treatment effectiveness. A change in stage distribution is a surrogate endpoint, not an automatic substitute for evidence of improved health."
    },
    {
      "heading": "False results, overdiagnosis, and other harms",
      "text": "A false-positive result indicates a possible condition that subsequent investigation does not confirm. Its consequences vary widely. Repeating a blood sample may impose little physical risk, whereas an abnormal scan can lead to invasive procedures, repeated imaging, or prolonged surveillance. A biopsy may cause bleeding, infection, or pain. Colonoscopic investigation can rarely cause perforation or significant bleeding. Radiation exposure from repeated imaging is another consideration, especially when many people without disease undergo testing.\n\nFalse-negative results have different consequences. They may delay recognition of disease or create false reassurance. Some negative results arise because disease was present but missed; others occur because the disease developed after testing or was below the method’s detection threshold. An interval cancer therefore does not always indicate a technical error. Programme evaluation distinguishes failures of test performance from limitations inherent in disease biology and testing schedules.\n\nOverdiagnosis is the detection of a condition that would not have caused symptoms or death during the person’s lifetime. It is not the same as a false-positive result: the abnormality is genuinely present, and may satisfy accepted diagnostic criteria, but its detection provides no health benefit. Overdiagnosis can occur when a lesion grows very slowly, stops growing, regresses, or is overtaken by death from another cause. Once detected, it may nevertheless lead to treatment because its harmless future cannot be identified with certainty in an individual.\n\nOvertreatment is a potential consequence. Treatment of overdiagnosed prostate cancer, for example, can expose people to urinary, sexual, and bowel complications without extending their lives. The size of overdiagnosis estimates varies with the condition, follow-up duration, and method of calculation. A sustained excess of diagnoses in a screened group after adequate follow-up may provide evidence, but interpretation must account for earlier diagnosis and changes in underlying disease occurrence. Estimates should specify whether their denominator is all screened people, all detected cases, or screen-detected cases.\n\nPsychological and social harms can also matter. An indeterminate finding may generate anxiety for months or years. Diagnostic labels can alter self-perception and family relationships, while genetic or infectious-disease information can raise confidentiality concerns. Not every participant experiences lasting distress, and psychological effects differ between conditions and programmes.\n\nThere are also opportunity costs. Staff, laboratory capacity, and clinical appointments devoted to low-value screening cannot simultaneously provide other care. A programme may appear inexpensive at the initial testing stage while creating substantial downstream expenditure. A complete assessment therefore includes consequences for participants, families, and services, rather than counting only the number of abnormalities found."
    },
    {
      "heading": "Screening programmes and quality assurance",
      "text": "A screening programme is a coordinated service rather than a collection of tests. Its pathway commonly includes defining eligibility, issuing invitations, obtaining an informed decision, collecting samples or performing examinations, interpreting results, communicating findings, arranging diagnostic assessment, and providing treatment or surveillance. Failure at any stage can erase the benefit demonstrated under research conditions. A positive result without accessible follow-up is not a completed preventive intervention.\n\nProgramme organization begins with a reliable way to identify eligible people. Population registers, clinical records, and recall systems can support invitations and repeat testing. Records must distinguish initial screening from diagnostic investigations and surveillance after a known abnormality. This distinction matters because the populations have different risks and because combining their results can produce misleading performance statistics. Secure information systems also allow programmes to identify people whose specimens were inadequate or whose diagnostic appointments were not completed.\n\nQuality assurance examines both technical performance and the pathway as a whole. Laboratories use standardized procedures, internal controls, and external assessment. Imaging programmes may audit image quality, interpretation, recall rates, and detection rates. Cytology and pathology services may review disagreements and correlate screening findings with later tissue diagnoses. Thresholds and categories must be sufficiently consistent that the same biological finding does not lead to markedly different management solely because it was processed in a different centre.\n\nImportant indicators include invitation coverage, participation, sample adequacy, time to results, completion of diagnostic assessment, complications, and treatment timeliness. Longer-term indicators include interval disease, advanced disease, and mortality. High participation alone is not proof of success: a programme can achieve excellent uptake while using an ineffective test or generating excessive harm. Conversely, poor follow-up can undermine an otherwise effective strategy. Performance indicators therefore require interpretation together rather than as isolated targets.\n\nResources and financing shape implementation. Cost-effectiveness analysis may compare additional expenditure with gains in quality-adjusted life years, while budget-impact analysis considers whether a programme is affordable at population scale. These are distinct questions. A programme can be cost-effective over a lifetime yet require substantial immediate expenditure on laboratories, information systems, and diagnostic capacity. Costs also fall on participants through travel, time away from work, or caregiving arrangements.\n\nResponsibility is commonly distributed across laboratories, primary care, hospitals, and public-health agencies. Clear accountability is therefore essential when a person moves between services. Integration with primary health care can support continuity, but specialist capacity remains necessary for many abnormal findings. Programme changes—such as a new assay, altered interval, or expanded age range—require renewed evaluation because they can change both the number of people referred and the balance of benefits and harms."
    },
    {
      "heading": "Cancer screening: contrasting biological and clinical models",
      "text": "Cancer screening includes several distinct strategies. Some tests seek invasive malignancy at a treatable stage; others identify precursor lesions whose treatment can prevent invasive cancer. The evidence cannot be transferred automatically from one organ or technology to another. Cancer is biologically heterogeneous, and detecting a tumour earlier is useful only if the resulting management improves outcomes enough to outweigh overdiagnosis and investigation-related harm.\n\nCervical screening illustrates prevention through precursor detection. Cytology identifies abnormal cells, while tests for high-risk human papillomavirus identify infection associated with cervical carcinogenesis. Because many infections resolve, HPV positivity is usually part of a multistep pathway rather than a diagnosis of cancer. Triage can involve cytology, HPV genotyping, repeat testing, or examination of the cervix. Vaccination changes the background distribution of infection and precancer, so programmes must adapt their methods and intervals as vaccinated populations enter screening ages.\n\nScreening for colorectal cancer uses stool-based tests, endoscopic examinations, or combinations of methods. Fecal immunochemical testing detects human blood in stool; a positive result generally requires colonoscopy to identify the source. Endoscopy can detect cancer and permit removal of some precancerous polyps. This creates both a detection effect and a prevention effect. However, bowel preparation, procedural complications, incomplete examinations, and access to follow-up influence the balance of benefit and harm.\n\nMammography uses low-dose X-rays to detect suspicious breast abnormalities. Trials and population programmes provide evidence of reduced breast cancer mortality in certain age groups, alongside false-positive recalls and overdiagnosis. The balance varies with age, baseline risk, screening interval, breast density, and the treatments available. An abnormal mammogram is not equivalent to a cancer diagnosis; additional imaging and sometimes tissue sampling are required.\n\nLung screening provides an example of evidence tied to a selected high-risk population. The National Lung Screening Trial reported in 2011 that low-dose CT reduced lung-cancer mortality by about 20% relative to chest radiography in its enrolled population of older people with substantial smoking histories. Subsequent studies, including the European NELSON trial, further informed the evidence. These findings do not establish a comparable benefit for population-wide chest CT, where lower underlying risk would alter predictive values and increase the relative burden of incidental findings.\n\nProstate-specific antigen testing illustrates a particularly close trade-off between possible mortality benefit and downstream harm. Elevated values can reflect benign enlargement or other noncancerous processes, and detected cancers vary greatly in aggressiveness. Conversely, large trials of ovarian-cancer screening have not established a mortality benefit sufficient to support routine population screening. Such contrasts show why an ability to detect cancer is only the beginning of programme evaluation."
    },
    {
      "heading": "Screening during pregnancy and the newborn period",
      "text": "Screening during pregnancy has several purposes: identifying maternal disease, detecting infections that can affect the fetus or newborn, estimating the probability of fetal abnormalities, and identifying pregnancy-related complications. These purposes differ ethically and clinically. A test that enables treatment to prevent transmission of an infection is not equivalent to a test that provides information about a fetal chromosomal condition. The intended outcome, possible next steps, and limits of interpretation must therefore be specified for each pathway.\n\nAntenatal screening for infections may include HIV, syphilis, and hepatitis B, depending on the programme. Their value depends on linked interventions: maternal treatment, appropriately planned perinatal care, and newborn prophylaxis or immunization where applicable. Screening for gestational diabetes uses glucose-based strategies whose thresholds and testing arrangements differ between health systems. These differences reflect variations in evidence interpretation, population characteristics, and service organization rather than a single universally implemented protocol.\n\nScreening for fetal chromosomal conditions can combine maternal characteristics, serum measurements, and ultrasound. Cell-free DNA testing analyzes fragments circulating in maternal blood, a pregnancy-related fraction of which originates mainly from the placenta. It has high screening accuracy for several common trisomies, including trisomy 21, which causes Down syndrome. It remains a screening method rather than a definitive fetal diagnosis. Confined placental mosaicism, a vanished twin, low fetal fraction, and maternal biological factors can affect results. Diagnostic assessment may involve chorionic villus sampling or amniocentesis when confirmation is sought.\n\nNewborn screening commonly combines blood-spot testing with hearing screening and, in some systems, pulse-oximetry screening for critical congenital heart defects. Blood-spot panels vary across jurisdictions. Conditions frequently included encompass phenylketonuria and congenital hypothyroidism; some programmes also screen for sickle cell disease, cystic fibrosis, and severe combined immunodeficiency. Inclusion depends not only on the availability of an assay but also on evidence that early identification permits effective intervention.\n\nTiming is especially important in newborn programmes. A sample collected too early may be difficult to interpret, whereas delayed reporting may miss a critical treatment window. Prematurity, transfusion, nutrition, and maternal influences can affect some measurements, requiring condition-specific repeat-testing rules. A positive blood-spot result usually requires urgent confirmatory assessment rather than being treated as a final diagnosis.\n\nNewborn screening also raises questions about residual sample storage and secondary use. Retained blood spots can support quality assurance and research, but their governance is distinct from the immediate clinical purpose of screening. Programmes differ in consent arrangements, storage periods, access rules, and the handling of findings not directly related to the original target condition."
    },
    {
      "heading": "Cardiometabolic, infectious, and questionnaire-based screening",
      "text": "Many screening activities occur during routine clinical encounters rather than through centrally organized invitations. Blood-pressure measurement is a prominent example. Hypertension is often asymptomatic, yet sustained elevation contributes to stroke, heart disease, and kidney damage. A single elevated office reading does not necessarily establish the diagnosis because measurement error, short-term physiological variation, and the clinical setting can influence results. Repeated standardized measurements or out-of-office monitoring help distinguish persistent elevation from transient or setting-related changes.\n\nScreening for diabetes mellitus may use fasting plasma glucose, glycated hemoglobin, or an oral glucose tolerance test. These measure different aspects of glucose regulation and do not identify exactly the same individuals. Factors affecting red-cell lifespan can alter the interpretation of glycated hemoglobin. In asymptomatic people, diagnostic confirmation is generally required unless accepted criteria for an unequivocal presentation are met. The value of detecting abnormal glycemia depends on effective subsequent management, not simply on assigning a label.\n\nCardiovascular prevention often combines several measurements rather than screening for one hidden lesion. Age, smoking, blood pressure, and lipid values can be incorporated into models estimating future cardiovascular disease risk. These models require validation in the populations where they are used. A risk score is a prediction, not proof of existing disease, and the same measured cholesterol concentration can imply different absolute risks in people of different ages or with different accompanying factors.\n\nInfectious-disease screening can serve both individual and population purposes. Identifying an asymptomatic infection may enable treatment and reduce transmission, although the magnitude of each effect depends on the organism and intervention. Tests for hepatitis C commonly distinguish evidence of prior exposure from current infection through sequential antibody and viral-RNA testing. Testing during an early window period may fail to detect infection. Frequency, exposure patterns, and access to treatment therefore influence programme performance.\n\nQuestionnaires are used to screen for conditions such as depression, harmful alcohol use, and developmental difficulties. Their scores indicate a need for further assessment; they do not independently establish a psychiatric or developmental diagnosis. Performance depends on language, cultural context, age, and the setting in which the questionnaire was validated. A low threshold can increase detection while also identifying transient distress or problems that require a different clinical explanation.\n\nScreening tools in these domains are most informative when embedded in a defined care pathway. Detecting elevated blood pressure, depressive symptoms, or an infection without ensuring diagnostic assessment and access to effective care may produce little health gain. This is particularly important where staff and treatment capacity are limited and the number of newly identified cases may exceed available services."
    },
    {
      "heading": "Ethics, equity, and emerging technologies",
      "text": "Screening is offered to people who may feel well and may never develop the target disease. This creates a distinctive ethical obligation to explain both possible benefit and possible harm. Informed consent requires more than agreement to specimen collection: the relevant information includes the purpose of testing, uncertain or incidental results, further procedures, and the possibility of overdiagnosis. Shared decision-making is especially relevant when reasonable people may value the same trade-off differently.\n\nCommunication can distort choices even when the numerical statements are technically correct. Presenting a relative reduction without the baseline risk can exaggerate the apparent magnitude of benefit. Describing only “lives saved” without a time horizon can conceal uncertainty about delayed deaths and competing causes of mortality. Balanced information uses comparable denominators and periods for benefits and harms. Declining screening after an informed decision is distinct from being unable to obtain it.\n\nHealth equity concerns arise at every stage. Invitation letters may not reach people without stable addresses; written materials may be inaccessible because of language or literacy; appointments may be difficult for people with inflexible work or caring responsibilities. Costs can occur even where the initial test is free. Screening may therefore widen disparities if better-resourced groups complete the pathway more readily. Equitable evaluation examines diagnostic follow-up and treatment as well as initial participation, and interprets differences in relation to social determinants of health.\n\nDigital systems and artificial intelligence are being used to prioritize images, identify eligible populations, interpret findings, and support risk-based intervals. Their performance can deteriorate when equipment, population characteristics, or clinical practice differ from the development setting. Retrospective accuracy does not establish improved patient outcomes. Automation can also change workload by increasing recalls or incidental findings. Prospective evaluation must therefore examine the complete pathway, including errors, treatment consequences, and unequal performance between groups.\n\nMulti-cancer early detection tests seek signals associated with several cancers, often using circulating DNA methylation patterns or other blood-based markers. Their appeal is the possibility of detecting malignancies for which no established population screening programme exists. Important unresolved questions include sensitivity for early-stage disease, identification of the tissue of origin, management of a positive result when imaging finds no lesion, and effects on cancer mortality. Detecting additional cancers is not by itself evidence that the strategy extends life.\n\nExpanded genomic screening raises related issues. Sequencing can reveal uncertain variants, adult-onset risks in children, or information relevant to relatives. Its clinical value depends on variant interpretation, penetrance, and the availability of effective action. At the same time, linked genomic and screening records create substantial data privacy obligations. Governance must address access, secondary use, retention, and the consequences of reinterpreting results as scientific knowledge changes."
    }
  ],
  "links": [
    "diagnosis",
    "public-health",
    "tuberculosis",
    "world-health-organization",
    "evidence-based-medicine",
    "disease-prevention",
    "human-papillomavirus-infection",
    "lung-cancer",
    "x-ray-computed-tomography",
    "lynch-syndrome",
    "blood-test",
    "biomarker",
    "prevalence",
    "randomized-controlled-trial",
    "quality-of-life",
    "intention-to-treat-analysis",
    "surrogate-endpoint",
    "biopsy",
    "overdiagnosis",
    "prostate-cancer",
    "mortality-rate",
    "quality-adjusted-life-year",
    "primary-health-care",
    "colorectal-cancer",
    "breast-cancer",
    "pregnancy",
    "hiv-aids",
    "syphilis",
    "hepatitis-b",
    "gestational-diabetes",
    "ultrasonography",
    "down-syndrome",
    "hypothyroidism",
    "sickle-cell-disease",
    "cystic-fibrosis",
    "severe-combined-immunodeficiency",
    "hypertension",
    "diabetes-mellitus",
    "glycated-hemoglobin",
    "cardiovascular-disease",
    "hepatitis-c",
    "depression",
    "informed-consent",
    "shared-decision-making",
    "health-equity",
    "social-determinants-of-health",
    "artificial-intelligence",
    "data-privacy"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}