For decades, the medical community and the general public have grappled with a complex trade-off inherent in breast cancer screening programs: the promise of early detection versus the risk of overdiagnosis. Overdiagnosis occurs when mammography identifies a cancer that, if left undetected, would never have caused symptoms or threatened a woman’s health during her lifetime. Because this phenomenon leads to unnecessary medical interventions and psychological distress, it has long been viewed as a significant drawback of population-based screening. However, a comprehensive new study, which reanalyzes data from all major randomized controlled trials, suggests that the scale of this problem has been significantly overestimated.
For years, researchers have debated the frequency of overdiagnosis, with estimates from various randomized trials producing starkly different figures. Some of these earlier studies suggested that between 30% and 50% of breast cancers detected through screening might fall into the category of overdiagnosis. These high figures have heavily influenced international public health policy, shaping the way governments and medical organizations communicate the potential harms of screening programs to the public.
A team of researchers, led by Sisse Helle Njor, a professor at the University of Southern Denmark and Lillebælt Hospital, set out to bring clarity to these conflicting figures. "The aim of our study was to bring together the evidence from all randomized controlled trials to get a clearer picture of the extent of overdiagnosis in breast cancer screening," says Njor. "Randomized trials have often been cited as evidence that overdiagnosis is a substantial problem. Our study shows that this interpretation is not as straightforward as it may seem."
A New Look at Mammography Trials
To investigate the issue, the research team performed a rigorous reanalysis of all available randomized mammography screening trials, comparing the findings against real-world data from Denmark. Denmark serves as a particularly useful reference for epidemiological studies because the country’s organized breast cancer screening program was introduced in different regions at different times, creating a 17-year gap between the earliest and latest rollouts. This staggered introduction allowed researchers to track how breast cancer incidence rates shifted immediately following the implementation of screening and how those trends evolved over several decades.
The researchers found that when they accounted for the timing of these diagnoses, the additional breast cancer cases identified in the randomized trials closely mirrored the patterns observed in Denmark, where the estimated rate of overdiagnosis is below 5%. This finding directly challenges the widely cited estimates that suggested nearly half of screen-detected cancers could be considered overdiagnosed.
Elsebeth Lynge, professor emerita at the Department of Public Health at the University of Copenhagen, explains the importance of timing in these statistical models. "When screening is introduced, the number of breast cancer diagnoses initially rises because cancers are detected earlier than they would have been without screening. Over time, this should be followed by a drop, as some of these cancers would otherwise have been diagnosed later. This pattern can also be affected if women in either group continue to undergo screening after the trials have ended, which was common. If researchers do not take these factors into account, the initial increase can be mistaken for overdiagnosis."
By comparing breast cancer incidence at matching points in time across the trials and the Danish screening programs, the researchers were able to assess whether the temporal patterns were consistent. This approach revealed that the perceived high rates of overdiagnosis were likely an artifact of insufficient follow-up time and a failure to account for how screening shifts the diagnosis timeline.
Matejka Rebolj, a Senior Epidemiologist at Queen Mary University of London, notes that the disparity between their findings and earlier estimates is largely due to the maturity of the data. "Taken together, we believe some previous high estimates of overdiagnosis, which influenced screening guidelines and communication, were based on evidence before trial data had fully matured," Rebolj says. "When interpreted in their full temporal context, randomized trial data are consistent with overdiagnosis of less than five percent, rather than with estimates nearing 50%."
Understanding Overdiagnosis
To understand why these statistics matter, it is necessary to define exactly what constitutes overdiagnosis. At its core, overdiagnosis is the identification of a breast cancer that would have remained indolent or asymptomatic for the remainder of a woman’s life. Without the screening, the woman would have remained unaware of the cancer, and it would not have resulted in death or significant health complications.
The definition also extends to women who may be diagnosed with a breast cancer but who pass away from an unrelated cause shortly thereafter. In these scenarios, the screening process offers little clinical benefit because the presence of other, more pressing health issues or a limited life expectancy means that treating the breast cancer is unlikely to extend the woman’s life or improve her overall health.
The core challenge for epidemiologists is that screening fundamentally alters the timeline of cancer progression. When a population begins regular mammography, there is an immediate, artificial spike in the number of cases detected. This is because the screening identifies tumors that would have eventually been diagnosed through symptoms later on, as well as those that might never have become problematic. If a study concludes before enough time has elapsed to observe the expected subsequent decline in later-stage diagnoses, researchers may incorrectly label that initial surge as a high rate of overdiagnosis.
Furthermore, many of the early randomized trials were complicated by the fact that women in control groups often gained access to screening once the trials had officially ended. This "crossover" effect, where the distinction between the control and intervention groups becomes blurred, can further distort estimates if not meticulously adjusted for in the final analysis.
What This Means for Women
The implications of this study are significant for both clinicians and the women they advise. Choosing whether or not to participate in a breast cancer screening program is a personal decision that requires a clear understanding of the balance between benefits and risks. For many years, the fear of overdiagnosis has been a prominent part of that conversation.
"Most women will not develop breast cancer, but with this study we can now be reassured that the benefits of detecting breast cancer early and preventing premature death will outweigh the small risk of unnecessary treatment," says Njor. The research team hopes that by providing a more accurate, evidence-based framework, they can assist medical professionals in better informing women when they are invited to participate in screening. The goal is to shift the dialogue toward a more realistic interpretation of the evidence, ensuring that women can make decisions based on the most robust data available rather than outdated, high-end estimates.
About the Study
The researchers conducted a comprehensive reanalysis of existing mammography screening literature, focusing on eight major randomized trials that have defined the field for decades. These include the New York Health Insurance Plan (HIP) trial, the Malmö mammographic screening trial, the Two-County trial in Sweden, the Edinburgh trial, the Canadian National Breast Screening Study, the Stockholm trial, the Gothenburg trial, and the UK Age trial.
By using the regional screening programs in Denmark as a real-world benchmark, the team was able to look at both invasive breast cancer and ductal carcinoma in situ (DCIS) with a high degree of precision. In their reassessment, the researchers systematically adjusted for three primary factors: the length of follow-up time, the specific timing of the initial screening, and the contamination of control groups. After meticulously accounting for these differences in exposure and follow-up, the team reached the conclusion that overdiagnosis is substantially less common than previous, less context-aware studies had suggested.
The research was supported by key health organizations and foundations. Casper Urth Pedersen’s work was supported by the Novo Nordisk Foundation (reference: NNF22OC0076184), while Matejka Rebolj’s contributions were supported by Cancer Research UK (reference: C8162/A29083). By refining these estimates, the study provides a vital update to the public health narrative, offering a more nuanced view of the long-term effectiveness of breast cancer screening.
