Screening: principles, lead-time and length bias, overdiagnosis, benefits and harms

At a glance

• Screening tests asymptomatic people to find disease or risk earlier, when treatment works better. It differs from case-finding in someone with symptoms.
• Screening always harms some people (false positives, overdiagnosis, complications of work-up). A programme is justified only when benefit clearly exceeds harm at the population level.
• Survival from diagnosis is a misleading measure. Lead-time, length and overdiagnosis bias all inflate it. Only an RCT showing reduced mortality in the invited group proves benefit.