A short paper on overdiagnosis and lead-time was recently published in the Journal of Internal Medicine (JGIM) to show how the statistical use of the term lead-time’ is very different from what doctors may think about when they use the term ‘lead-time’.
The authors argue that if overdiagnosis is adjusted for model-based lead-time, then all overdiagnosis is removed, whereas the incidence increase when screening should only be adjusted for clinically relevant tumours, and suggest these two different approaches are the main cause of the discrepancy on the level of overdiagnosis of breast cancer when screening with mammography in the literature.
They also compare model-based lead-time to lead-time for clinical relevant tumours when screening with PSA.
‘Lead-time models should not be used to estimate overdiagnosis in cancer screening’ – Per-Henrik Zahl, Norwegian Institute of Public Health, Oslo, Norway; Karsten Juhl Jorgensen, Peter C Gotzsche, both from the Nordic Cochrane Centre, Denmark. J Gen Intern Med 2014 Mar 4. Epub 2014 Mar 4.