False positives and overdiagnosis, by test
Every screening test sometimes raises a false alarm, and some find cancers that would never have caused harm. These are the numbers the USPSTF reports for each test, next to the benefit numbers from the same statements. They are not a reason to skip screening: the USPSTF weighed them and still recommends each test in its age range. They are what to expect, so a callback is less of a shock.
Two terms
- False positive: the test looks abnormal but there is no cancer. It can cause anxiety and usually leads to more tests and procedures, which have their own risks. (NCI, 2023)
- Overdiagnosis: screening finds a cancer that would never have caused symptoms or become life-threatening. It may then be treated, and there is no way to know whether treatment helped that person. (NCI, 2023)
At a glance
| Test | False positives | Overdiagnosis | Benefit, same statement |
|---|---|---|---|
| Mammogram every 2 years, 40 to 74 | 1,376 per 1,000 women over a lifetime (some women have more than one) | 14 per 1,000 women | About 8.2 breast cancer deaths averted per 1,000 women (model median) (USPSTF, 2024) |
| Low-dose CT, yearly | About 1.9 per person over a lifetime under the 2021 criteria (model) | About 6% of screen-detected lung cancers (model) | NELSON trial: 181 vs 242 lung cancer deaths at 10 years, screened vs not (USPSTF, 2021) |
| PSA, 55 to 69 | About 1 in 6 men screened at least once (ERSPC) | 20% to 50% of screen-detected prostate cancers | About 1.28 fewer prostate cancer deaths per 1,000 men screened (ERSPC) (USPSTF, 2018) |
| Stool tests | FIT: specificity 94% (about 6 in 100 without colorectal cancer test positive). Stool DNA-FIT: 84% (about 16 in 100) | Not quantified in the statement | See the colorectal page (USPSTF, 2021) |
Lifetime figures are estimates from the modeling the USPSTF commissioned; trial figures come from the named trials as the USPSTF reports them. Different sources and time spans, so compare within a row, not across rows.
Mammograms
The USPSTF calls a false-positive result the most common harm of mammography. Its modeling estimates that a mammogram every 2 years from 40 to 74 leads to 1,376 false-positive results and 14 overdiagnosed breast cancers per 1,000 women over a lifetime of screening. Starting at 40 instead of 50 adds about 60% more false positives and 2 more overdiagnosed cancers per 1,000, and 1.3 more breast cancer deaths averted. (USPSTF, 2024)
- Over 10 years: in data the USPSTF cites, the chance of at least one false positive was 38.1% with digital mammography and 35.7% with 3D mammography (tomosynthesis) when screening every 2 years, and 56.3% and 49.6% with yearly screening. (USPSTF, 2024)
- Every 2 years vs every year: yearly screening from 40 to 74 would mean about 50% more false positives and 50% more overdiagnosed cancers. (USPSTF, 2024)
- Most people called back are not found to have breast cancer, the National Cancer Institute says. (NCI, 2025)
Lung CT scans
In the US NLST trial, false-positive rates were 26.3% at the first scan, 27.2% at the second and 15.9% at the third; in the European NELSON trial, 19.8% at the first scan and lower after. Looking back at NLST, one study found that the Lung-RADS reporting system would have cut the first-scan false-positive rate to 12.8%. (USPSTF, 2021)
- Procedures: in NLST, false positives led to an invasive procedure, such as a needle biopsy or surgery, in 1.7% of people screened; complications occurred in 0.1%. (USPSTF, 2021)
- Overdiagnosis: modeling estimates about 6.0% of lung cancers found by screening under the 2021 criteria are overdiagnosed. (USPSTF, 2021)
- Radiation: one low-dose CT is 0.65 to 2.36 mSv, against average yearly background radiation in the US of 2.4 mSv. (USPSTF, 2021)
PSA tests
The false-positive rate depends on the PSA cutoff. In the European ERSPC trial, about 1 in 6 men screened at least once had a false positive, and two-thirds of first-round positive results were false positives. Another major trial found that more than 15% of men screened every 2 to 4 years had at least one false positive over 10 years. About 1% of prostate biopsies lead to complications needing a hospital stay. (USPSTF, 2018)
Trial follow-up suggests 20% to 50% of men diagnosed with prostate cancer through screening may be overdiagnosed, with the rate rising with age, which is part of why the USPSTF recommends against screening from 70. (USPSTF, 2018) More on weighing PSA screening.
Stool tests and colonoscopy
For colorectal cancer, the USPSTF reports pooled specificity of 94% for FIT and 84% for stool DNA-FIT, so roughly 6 and 16 in 100 people without colorectal cancer get a positive result; some of these turn out to have precancerous polyps. A positive result means a colonoscopy, and colonoscopy after a positive stool test led to about 17.5 serious bleeding events and 5.4 perforations per 10,000 procedures, compared with 14.6 and 3.1 per 10,000 for screening colonoscopy. (USPSTF, 2021)
What lowers the chance of a false alarm
- Sticking to the recommended interval: for mammograms, every 2 years gives fewer false positives than every year. (USPSTF, 2024)
- Bringing old results: a first mammogram, or one with no earlier images to compare, is more likely to lead to a callback. See after a mammogram callback.
- Staying in your age range: harms rise outside it, which is why some tests stop. See when to stop screening.
Questions people ask
What is the false-positive rate for mammograms?
Over 10 years of screening every 2 years, the chance of at least one false positive was about 38% with standard digital mammography and about 36% with 3D mammography (tomosynthesis), in data the USPSTF cites. With yearly screening it was about 56% and 50%. (USPSTF, 2024)
How common are false positives on a lung cancer CT scan?
In the large US NLST trial, about 26% of first scans were false positives. Using the Lung-RADS reporting system, one study estimated that rate would have been about 13%. False positives led to an invasive procedure in 1.7% of people screened in NLST. (USPSTF, 2021)
How many men get a false-positive PSA test?
In the European ERSPC trial, about 1 in 6 men screened at least once had one or more false-positive results, and two-thirds of first-round positive results were false positives, the USPSTF reports. (USPSTF, 2018)
What is overdiagnosis?
Finding a cancer that would never have caused symptoms or threatened life. The National Cancer Institute notes that once found it may be treated, and there is no way to know whether treatment helps that person live longer. (NCI, 2023)