🧪 This platform is in early beta. Features may change and you might encounter bugs. We appreciate your patience!
lead_time_bias
Lead-time bias occurs when earlier detection of a disease through screening appears to extend survival time even when it does not actually change the date of death. If a disease would have been detected at age 60 clinically but is detected by screening at age 55, the patient now appears to survive 10 years (to age 65) instead of 5, even if they die at exactly the same age 65.
A lung cancer screening study reports that screened patients live an average of 15 months after diagnosis compared to 9 months for unscreened patients. But if the disease was caught 6 months earlier through screening, the actual survival benefit may be zero.
A new blood test for pancreatic cancer is celebrated because patients who test positive live an average of 18 months after diagnosis, versus 8 months for those diagnosed after symptoms. However, autopsies and disease modeling suggest the cancer's biological course is identical in both groups — the test simply detects the disease 10 months earlier, advancing the diagnosis clock without extending actual life.
A dementia screening program reports that patients identified early live with the diagnosis for an average of 9 years, compared to 5 years for those diagnosed after cognitive decline becomes obvious to family members. Neurologists caution that the disease progression timeline appears unchanged and that the extra 4 years largely represent time the patient spent labeled as having dementia without any alteration in the disease's ultimate course.
Binary (yes/no) questions an LLM must answer to identify this aspect:
Is survival measured from time of diagnosis rather than from symptom onset or death?
Type: binaryDoes earlier detection via screening appear to extend survival time?
Type: binaryIs there evidence that the earlier diagnosis actually changed the date of death?
Type: binaryAre mortality rates (deaths per population per year) used alongside survival rates?
Type: binaryLead-time bias occurs when earlier detection of a disease through screening appears to extend survival time even when it does not actually change the date of death. If a disease would have been detected at age 60 clinically but is detected by screening at age 55, the patient now appears to survive 10 years (to age 65) instead of 5, even if they die at exactly the same age 65.
Survival time is measured from diagnosis, so moving the diagnosis earlier mechanically increases measured survival time without any change in mortality. Audiences conflate longer survival time with longer life.
Require mortality data (deaths per 100,000 population per year) rather than survival data as the primary outcome for screening studies. Use randomized trials that compare total mortality.
Lead-time bias contributed to overestimates of neuroblastoma screening benefits in Japan in the 1990s; subsequent randomized trials showed no mortality benefit.
Use these tools to detect, analyze, or train this aspect.