BiostatisticsTier 1Medical Sciences concept

Evidence for population screening programs (mortality benefit)

Core concept

A screening test finding disease earlier does not by itself prove it saves lives -- three biases inflate apparent benefit and must be excluded before accepting mortality benefit
Lead-time bias: earlier diagnosis lengthens the known survival time without changing the actual death date -- looks like benefit, is only earlier labelling
Length-time bias: screening preferentially detects slow-growing, indolent disease (present longer, easier to catch) while fast, aggressive disease is missed between rounds -- screened group looks to do better because it is enriched for less lethal disease

Overdiagnosis: detecting disease that would never have caused symptoms or death in the patient's lifetime -- a further step beyond length-time bias, at the extreme end
The only evidence that overcomes all three biases: randomised controlled trial of screening vs no screening, with disease-specific or all-cause mortality as the endpoint -- not survival time, not stage-shift, not detection rate
Trap: "5-year survival improved after screening was introduced" is exactly the pattern lead-time and length-time bias predict, and proves nothing about mortality reduction

Evaluating whether to support/recommend a screening programme: ask has it shown a mortality reduction in an RCT, not just improved detection or survival statistics
Established mortality-benefit screening (RCT evidence): mammography (breast), FOBT/colonoscopy (bowel), cervical screening (HPV/cytology)
Not similarly established / contested: PSA screening for prostate cancer -- detects large amounts of indolent disease (overdiagnosis), mortality benefit small/contested despite clear increase in detection

Wilson-Jungner criteria for a good screening programme (disease detectable in latent phase, effective treatment available, test acceptable/accurate, cost-effective) are the companion framework -- causal evidence (this note) is the piece those criteria assume
Overdiagnosis is the same phenomenon underlying incidental findings on imaging (e.g. thyroid nodules, small renal masses) driving unnecessary treatment
Same bias logic applies when critiquing any observational comparison of "screened" vs "unscreened" survival outside a trial

3 more sections, plus exam facts

Premium unlocks every note across every specialty, and the full exam fact library behind it.