Breast‑cancer screening sometimes finds tumors that would never cause symptoms or danger during a woman's life. This situation is called overdiagnosis and is often listed as a drawback of screening programs.
Scientists have argued for decades about how common overdiagnosis really is. Some early trials reported that 30‑50 % of cancers detected by screening might be overdiagnosed, shaping global discussions about the benefits and harms of mass screening.
"Our aim was to gather all randomized‑trial data and get a clearer picture of overdiagnosis," explains Professor Sisse Helle Njor of the University of Southern Denmark. "Past studies have presented overdiagnosis as a big problem, but the picture is more nuanced. "
The new analysis shows that the extra breast‑cancer cases found in trials match the pattern seen in Denmark, where overdiagnosis is thought to be below 5 %.
A Fresh Look at Mammography Trials
Researchers combined results from every randomized mammography trial and compared them with real‑world data from Denmark. Denmark is useful because some regions began organized screening 17 years earlier than others, allowing a clear view of how cancer rates changed when screening started and later settled.
"When screening begins, the number of diagnoses jumps because cancers are found earlier. Over time, the rate should fall as those early detections replace later ones," notes Professor Elsebeth Lynge, emerita, Department of Public Health, University of Copenhagen. "If studies stop too soon, the early rise can be mistaken for overdiagnosis, especially when women in the control group later receive screening themselves. "
By matching time points in the trials with Denmark’s routine programs, the team could see whether the patterns were similar and what they meant for the true size of overdiagnosis.
"We think earlier high estimates were based on incomplete trial data. When the full time course is considered, the evidence points to less than five percent overdiagnosis, not the 30‑50 % that some guidelines have used," says Senior Epidemiologist Matejka Rebolj of Queen Mary University of London.
Fact: Overdiagnosis means a screening test finds a cancer that would never become life‑threatening or cause symptoms during a woman's lifetime. Without screening, the woman would never know the tumor existed.
It also includes cases where a woman dies from another cause soon after diagnosis, meaning treatment would not have improved her health or lengthened her life.
Why Timing Changes the Numbers
Screening shifts the moment a cancer is discovered. Early detection creates an initial spike in cases, followed by a later dip as the same cancers would have appeared later without screening. If a study ends before that dip shows up, the early spike can be wrongly counted as overdiagnosis. The picture also gets blurred when women in the control group later get screened.
The new analysis shows that correcting for these timing effects can dramatically lower the estimated rate of unnecessary cancer detection.
What This Means for Women
Knowing both the good and the possible downsides helps women decide about screening. "Most women will never get breast cancer, but this study reassures us that the life‑saving benefits of early detection outweigh the small chance of unnecessary treatment," says Professor Njor. "We hope this work gives doctors a clearer way to explain the evidence when inviting women for screening. "
About the Study
The researchers re‑examined data from all eight major randomized mammography trials: New York Health Insurance Plan, Malmö, Two‑County, Edinburgh, Canadian National Breast Screening Study, Stockholm, Gothenburg, and UK Age. They used two Danish regional programs as a real‑world reference and looked at both invasive cancers and ductal carcinoma in situ (DCIS).
Three key factors were considered:
- Whether women in the control groups later received screening.
- The number of screening rounds each group experienced.
- The length of follow‑up for both screened and unscreened women.
After adjusting for these variables, the team concluded that overdiagnosis is likely far less common than earlier estimates suggested.
Funding
Casper Urth Pedersen received support from the Novo Nordisk Foundation (NNF22OC0076184). Matejka Rebolj was funded by Cancer Research UK (C8162/A29083).