How Often do Mammograms Overdiagnose Breast Cancer?

InpharmD™ Clinical Literature Summaries — S3E12

4 min readMar 25, 2022


Season 3 is available on all streaming platforms.

Like our script? Like our podcast? Like our study summary?

Give us a review!

Photo by Angiola Harry on Unsplash

InpharmD combines pharmacists’ intelligence with artificial intelligence to provide you custom, evidence-based responses to your clinical questions. Ask any question, anytime, from anywhere, on Android, iphone, Epic, and now, contactless with Siri and Alexa.

On behalf of all of us here at InpharmD, whether you’re listening to this from your home or on the front lines, we hope you’re well and we thank you for the important work you’re doing.

With this podcast, we recap innovative, practice changing studies in ten minutes or less. And remember, nothing here is medical advice; we just present the evidence and our (sometimes hot) takes.

How Often do Mammograms Overdiagnose Breast Cancer?

We all know that mammograms are an invaluable screening tool for diagnosing breast cancer. While there are various subtypes of breast cancer based on histology and gene expression, one common theme is that earlier diagnosis leads to better outcomes. But on the flip side, mammograms can also overdiagnose breast cancer, meaning they detect cancer that may not be clinically relevant.

While it may not seem like a big deal to catch too many cases as compared to missing cases, its important to remember both how anxiety provoking a possible cancer diagnosis can be and also the cost to the healthcare system that results from further invasive testing after a positive screening mammogram. This is why the US preventative services task force recommends screening every 2 years in those age 50–74, and shared decision making for screening in women age 40–49.

So today, we’re going to look at a study that attempted to quantify how often breast cancer is overdiagnosed from mammograms. This was a retrospective cohort study of over thirty five thousand women age 50–74 who were screened every 2 years with mammograms. The researchers attempted to identify what percent of these women ended up being “overdiagnosed” with breast cancer. In this study they defined overdiagnosis as either detecting an indolent pre-clinical cancer or detecting a progressive pre-clinical cancer in a woman who would have died from an unrelated cause before clinical diagnosis.

We won’t bog you down too much with the complex statistical methods the study used, but they essentially looked at past medical records from the year 2000 to 2018 to determine who was diagnosed with breast cancer. They then indirectly quantified how many individuals were overdiagnosed using an inference model as its unable to be directly quantified.

Believe it or not, they found that just over 15% of breast cancer diagnoses were considered overdiagnosed. About 6% of these were indolent pre-clinical cancers and just over 9% were from detection of a progressive pre-clinical cancer in a woman who would have died from an unrelated cause. While it’s widely accepted that breast cancer screening can lead to overdiagnosis, getting an actual quantifiable result gives us more perspective. In this case, this means that about one out of every seven cases of screen detected breast cancer are overdiagnosed!

This is pretty shocking and it shows us that no test is perfect. At the same time, we have to realize that this study wasn’t perfect either. It used a large breast cancer registry database, but unfortunately this database wasn’t able to distinguish between women who were lost to follow up and those who had not yet come back for their next screening exam. They did combat this however by censoring out any diagnoses that took place over 18 months after the last screening mammogram.

They also excluded any woman who had their first screening mammogram outside of the breast cancer registry. While it might not seem like a big deal, it did lower event numbers, taking away from the ability to fit models to different patient features such as race or ethnicity.

Despite all this, the study had a solid foundation and produced provocative results. This study is not meant to discourage breast cancer screening, but rather to provide context for how often the screening can lead to overdiagnosis even in the age range recommended for screening by the US preventative services task force.

If you liked the show, please write a review. It helps other listeners find us, and it means a lot when you do.

The InpharmD Podcast is an independent production of InpharmD. Check out more evidence-based information at




We take clinical questions from the world’s best healthcare providers and return custom, evidence-based answers, so they can make better decisions.