Key Facts

Screening aims to reduce morbidity and mortality by detecting disease at an early, treatable stage. However, screening programmes also cause harm and must be carefully evaluated. The balance of benefits and harms must favour the screened population.

Wilson-Jungner Criteria (1968)

1. Important health problem
2. Accepted treatment available
3. Facilities for diagnosis and treatment available
4. Recognisable latent or early symptomatic stage
5. Suitable test available
6. Test acceptable to population
7. Natural history of condition understood
8. Agreed policy on who to treat
9. Cost-effective
10. Screening should be a continuous process

Types of Screening

• Population screening: offered to all individuals in a defined population (e.g. NHS Breast Screening)
• Targeted screening: offered to high-risk groups (e.g. TB screening in new entrants from high-incidence countries)
• Opportunistic screening: offered during clinical encounters (e.g. blood pressure check at GP visit)
• Cascade screening: testing relatives of identified cases (e.g. familial hypercholesterolaemia)

Screening Test Properties

• Sensitivity: ability to detect true positives (high sensitivity = few false negatives)
• Specificity: ability to detect true negatives (high specificity = few false positives)
• PPV: proportion of positive results that are true positives (highly influenced by prevalence)
• NPV: proportion of negative results that are true negatives

UK NHS Screening Programmes

Antenatal and Newborn:

• Fetal anomaly scan (18-20+6 weeks)
• Combined screening for trisomies (11-13+6 weeks)
• Infectious disease screening (HIV, hepatitis B, syphilis, rubella immunity)
• Newborn blood spot (PKU, CHT, sickle cell, CF, MCADD + 5 others)
• Newborn hearing screening
• Newborn and infant physical examination (NIPE)

Cancer Screening:

• Breast: mammography every 3 years, women aged 50-71
• Cervical: primary HPV testing, 25-64 (3-yearly 25-49, 5-yearly 50-64)
• Bowel: FIT every 2 years, 50-74

Other:

• Abdominal aortic aneurysm: USS, men aged 65 (one-off)
• Diabetic eye screening: annual retinal photography for all diabetics aged ≥12
• NHS Health Check: CVD risk assessment, 40-74, every 5 years

Screening NOT Currently Recommended in UK

• Prostate cancer (PSA) — population screening; high false positive rate and overdiagnosis
• Lung cancer — NLST showed benefit but not yet implemented as national programme
• Ovarian cancer — UKCTOCS showed no mortality benefit

Evaluating Screening Programme Effectiveness

• RCT: gold standard for evaluating screening (e.g. HIP trial for breast screening, UKCTOCS for ovarian)
• Mortality reduction: primary outcome measure (disease-specific and all-cause)
• Lead-time bias: apparent survival improvement from earlier detection without actual prolongation of life
• Length-time bias: screening preferentially detects slower-growing disease
• Overdiagnosis bias: detecting disease that would never have caused symptoms or death
• Volunteer bias: those who attend screening may be healthier than non-attenders

Test Performance Metrics

• Sensitivity and specificity (intrinsic to test)
• PPV and NPV (influenced by prevalence)
• ROC curve: plots sensitivity vs (1-specificity) at different cut-off points
• Area under ROC curve (AUC): overall test performance (1.0 = perfect, 0.5 = useless)

Maximising Screening Programme Effectiveness

• High uptake/coverage (target >80%)
• Quality assurance: standardised protocols, training, audit
• Minimise interval cancers (cancers arising between screening rounds)
• Prompt follow-up of abnormal results
• Informed consent: explain benefits AND harms

Communicating Screening Results

• Use absolute numbers rather than percentages (e.g. '3 in 1,000' rather than '0.3%')
• Explain that screening is not diagnostic; positive results require further investigation
• Address false positive and false negative possibilities
• Provide written information before screening (informed choice)

Harms of Screening

• False positives: anxiety, unnecessary investigations, potential complications of investigation
• False negatives: false reassurance, delayed presentation
• Overdiagnosis: detecting conditions that would never cause harm (e.g. screen-detected DCIS, low-risk prostate cancer)
• Overtreatment: treating overdiagnosed conditions with associated morbidity
• Psychological harm: anxiety from screening invitation, results, further investigation
• Opportunity cost: resources used for screening could be used elsewhere

• Breast screening: estimated to prevent 1,300 deaths/year in UK; but causes overdiagnosis in approximately 4,000 women/year
• Cervical screening: prevents approximately 5,000 deaths/year in UK; HPV vaccination will further reduce need over time
• Bowel screening: reduces CRC mortality by 16-25% in screened populations
• AAA screening: reduces AAA-related mortality in men by approximately 42%
• Newborn blood spot: early detection of PKU prevents severe intellectual disability; early treatment of CHT prevents cretinism
• Effective screening programmes represent excellent value for money when Wilson-Jungner criteria are met

Lead-Time Bias Illustration

Overdiagnosis Examples