Screening finds cancer early, and it also finds cancer that never mattered

Key takeaways

• Over a lifetime, mammograms every two years from 40 to 74 prevent about 8 breast cancer deaths per 1,000 women, along with about 1,376 false-positive recalls and 14 overdiagnosed cancers, in Task Force modeling⁴.
• In Task Force models that assume everyone follows through, colorectal screening from 45 to 75 prevents 24 to 28 cancer deaths per 1,000 people, partly by removing growths before they turn into cancer⁶.
• The FDA-approved Shield blood test caught about 83% of colorectal cancers but only about 13% of advanced precancerous growths, and a positive result still needs a colonoscopy⁷.
• The American College of Radiology says the evidence doesn’t support whole-body MRI for people without symptoms, and the Galleri multi-cancer blood test was still under FDA review after a split advisory-panel vote in September 2026¹²¹⁴¹⁵.
• Before any screening test, ask five questions: what it finds, how many in 1,000 it helps, how many get false alarms, how many are overdiagnosed and what happens after a positive result.

In 1999, South Korea launched a national cancer screening program. Thyroid cancer wasn’t part of it. But clinics began offering a thyroid ultrasound as an add-on, for about $30 to $50².

Diagnoses soared. From 1993 to 2011, Korea’s thyroid cancer rate rose 15-fold, according to the National Cancer Institute¹. In 2011, doctors diagnosed about 40,000 cases, more than 100 times the number of thyroid cancer deaths¹.

The death rate barely moved¹. Screening had found thousands of cancers that were never going to hurt anyone. NCI cites an estimate that 90% of Korea’s cases were overdiagnosed¹.

Those patients paid a real price. About two-thirds had the whole thyroid removed, which means taking thyroid hormone for life². In an analysis of more than 15,000 surgical patients, 11% developed underactive parathyroid glands and 2% had vocal cord paralysis². In 2015, a Korean guideline panel recommended against routine thyroid ultrasound in healthy people¹.

Some screening tests do save lives. This article shows how to weigh any test by counting benefits and harms in the same units; a later article covers which screenings fit each age.

A screening test is a bet placed on people who feel fine

A diagnostic test looks for the cause of a symptom. A screening test looks for hidden disease in people with no symptoms at all. That difference changes the math.

Two numbers describe a test’s accuracy. Sensitivity is the share of people with the disease whom the test catches. Specificity is the share of people without the disease whom the test correctly clears.

Because most people screened are healthy, even a good specificity produces many false alarms. The Shield colorectal blood test, for example, has a specificity of about 90%⁷. By our arithmetic, that means about 1 in 10 people without the disease it looks for would get a positive result.

So the best yardstick is deaths prevented per 1,000 people screened, set beside false alarms and overdiagnosed cases per 1,000.

Mammograms show how a modest benefit and frequent false alarms can both be true

In April 2024, the U.S. Preventive Services Task Force (USPSTF), an independent panel of prevention experts, lowered the starting age for mammograms. It now recommends screening every two years for women 40 to 74, a “B” grade³. A key reason: breast cancer in women in their 40s rose about 2% a year from 2015 to 2019³.

The Task Force drew on six computer models from the CISNET research network. For every 1,000 women screened with 3D mammography every two years from 40 to 74, over a lifetime, the models found⁴:

• about 8 breast cancer deaths prevented (range 6 to 11);
• about 1,376 false-positive recalls;
• about 201 biopsies that turned out benign;
• about 14 overdiagnosed cancers (range 4 to 37).

Starting at 40 rather than 50 added about 1.3 prevented deaths per 1,000 women. It also added about 503 false-positive recalls, 65 benign biopsies and 2 overdiagnosed cases⁴. For Black women, who are about 40% more likely than White women to die of breast cancer, the earlier start prevented about 1.8 more deaths per 1,000³⁴.

False alarms are common for any one woman, too. A 2011 study in the Annals of Internal Medicine, led by Rebecca Hubbard, used records from the NCI-funded Breast Cancer Surveillance Consortium on more than 169,000 women in their 40s and 50s⁵. Over 10 years of yearly mammograms, a woman’s estimated chance of at least one false-positive recall was 61%. Screening every other year lowered it to 42%⁵.

None of this means mammograms don’t work. It means a callback is usually not cancer. The Task Force found too little evidence to judge extra ultrasound or MRI for women with dense breasts³.

Colon screening can stop cancer before it starts

Most screening finds disease. Colorectal screening can also prevent it, because a colonoscopy removes polyps, the growths that can turn into cancer.

The Task Force recommends screening from 45 to 75, with an “A” grade for ages 50 to 75 and a “B” for 45 to 49, both from 2021⁶. You can choose among several tests⁶:

• a stool test for hidden blood (FIT) every year;
• a stool DNA test with FIT every 1 to 3 years;
• CT colonography or flexible sigmoidoscopy every 5 years;
• colonoscopy every 10 years.

In the Task Force’s models, screening from 45 to 75 prevented 24 to 28 colorectal cancer deaths and 42 to 61 cancers per 1,000 people⁶. The cost was 10 to 16 colonoscopy complications per 1,000 over a lifetime, such as bleeding or a torn bowel⁶.

There’s a catch with the stool tests. A positive result only helps if a colonoscopy follows, and the Task Force says the benefit depends on it⁶. In one study cited by NCI, people who skipped the follow-up colonoscopy after a positive FIT were twice as likely to die of colorectal cancer⁷.

The newest option is a blood draw. In July 2024, the FDA approved Shield, made by Guardant Health, for adults at average risk⁷. In the company-funded ECLIPSE study, it caught about 83% of colorectal cancers but only about 13% of advanced precancerous polyps⁷.

That second number matters. A test that misses most precancerous growths can’t prevent many cancers the way a colonoscopy can. The American Society for Gastrointestinal Endoscopy says blood tests suit only people who won’t do a stool test or colonoscopy⁸.

It also cites modeling: for every 3 people who switch to a blood test from colonoscopy or a stool test, 2 never-screened people must start using it just to break even on deaths⁸.

The 2021 Task Force recommendation came before Shield’s approval and did not include blood tests⁶.

The PSA test is a decision, not a default

The prostate-specific antigen (PSA) test measures a protein in the blood. For men 55 to 69, the Task Force gave it a “C” in 2018, meaning the choice should be made one man at a time. It recommends against screening at 70 and older⁹.

What the Task Force found per 1,000 men screened over about 13 years⁹:

• about 1.3 prostate cancer deaths prevented;
• about 3 fewer cases of cancer that spread;
• about 240 men with at least one positive PSA test;
• about 2 men hospitalized for complications of a biopsy.

Overdiagnosis is the biggest hidden cost. The Task Force estimates that 20% to 50% of men diagnosed through screening may be overdiagnosed⁹.

Treatment has lasting effects. About 1 in 5 men develop long-term urine leakage after prostate removal, and about 2 in 3 have long-term erectile problems⁹. More than half of men treated with radiation have long-term erectile problems⁹.

Black men and men with a strong family history may weigh the choice differently; our companion article on family health history covers that. As of October 2026, the Task Force lists this recommendation as being updated⁹.

Lung scans help long-time smokers, and too few get one

For adults 50 to 80 who smoked at least 20 pack-years (a pack a day for 20 years, or the equivalent) and still smoke or quit within 15 years, the Task Force recommends a yearly low-dose CT scan. That’s a “B” grade from 2021¹⁰.

In the National Lung Screening Trial, a large U.S. study, CT screening cut lung cancer deaths by about 20%¹⁰. Models estimate that about 6% of screen-found lung cancers are overdiagnosed¹⁰.

Other harms are real but mostly short-lived. In that trial, about 26% of people had a false-positive result on their first scan. About 17 in 1,000 had an invasive procedure, and about 1 in 1,000 had a complication¹⁰.

Few eligible people use it. In one 10-state study the Task Force cited, only 14.4% of eligible adults had been screened in the past year¹⁰.

Tests sold straight to healthy people: what the record shows

Some screenings are marketed directly to the public.

Whole-body MRI. In 2023, the American College of Radiology said the evidence doesn’t justify total-body scans for people without symptoms, risk factors or a telling family history. It warned of vague findings that lead to more tests and costs, and said there was no documented evidence the scans prolong life¹².

RSNA News reported in July 2026 on about 160 people scanned at Weill Cornell in its first 18 months. Of them, 90% had some finding, though most needed no follow-up; about 30% had findings that weren’t worrisome but needed a recheck; and about 1% of findings were cancer or suspected cancer. A radiologist there said the data don’t yet show the scans extend lives¹³.

Multi-cancer blood tests. GRAIL’s Galleri test looks for signals of many cancers in one blood draw. In GRAIL’s own report of the NHS-Galleri trial in England, a randomized study of about 140,000 people, the test didn’t significantly reduce combined stage III and IV cancers. GRAIL did report fewer stage IV diagnoses of 12 aggressive cancers in later screening rounds¹⁴.

On September 23, 2026, an FDA advisory panel reviewed GRAIL’s application for adults 50 and older. According to the FDA’s meeting summary, the panel voted 10 to 0 on safety, 6 to 4 on effectiveness, and 7 to 2, with one abstention, on whether benefits outweigh risks. Some no voters cited the lack of data on long-term patient outcomes¹⁵.

As of the meeting, the FDA itself hadn’t ruled on the test, which GRAIL describes as not cleared or approved¹⁴¹⁵.

Ovarian cancer screening. This one was tested in large trials and failed. In the U.S. PLCO trial, about 32 of every 1,000 women without cancer had surgery after a false alarm, and up to 15% of them had major complications. Screened women were no less likely to die of ovarian cancer, and the Task Force recommends against it¹¹.

How strong is the evidence for each test?

Test	Who it’s for	Evidence rating	What it rests on
Colorectal screening	Adults 45 to 75	Strong evidence	24 to 28 deaths prevented per 1,0006
Mammogram every two years	Women 40 to 74	Strong evidence	About 8 deaths prevented per 1,000; frequent false alarms34
Low-dose lung CT	Long-time smokers 50 to 80	Strong evidence	About 20% fewer lung cancer deaths in the national trial10
PSA test	Men 55 to 69	Strong evidence of a small benefit with real harms	About 1.3 deaths prevented per 1,000; 20% to 50% overdiagnosis9
Shield blood test in place of colonoscopy or stool tests	Average-risk adults	Promising but early	Catches most cancers, few precancerous polyps; no death data yet78
Galleri multi-cancer test	Adults 50 and older (proposed)	Not enough evidence	No significant drop in late-stage cancers, per the company; FDA decision pending1415
Whole-body MRI	People without symptoms	Not enough evidence	Many incidental findings; no survival data1213
Ovarian screening (CA-125, ultrasound)	Women without symptoms	Evidence says no, or unsafe	Unneeded surgery, no fewer deaths11
Thyroid ultrasound	Adults without symptoms	Evidence says no, or unsafe	Korea’s 15-fold rise in diagnoses without fewer deaths1

The screening test scorecard

Use these five questions for any screening test, whether your clinician suggests it or an ad does. The scorecard is our own tool, not a validated instrument. Here it is applied to two tests, with numbers from the Task Force.

Question	Mammogram every 2 years, ages 40 to 74 (lifetime, per 1,000 women)	PSA test, ages 55 to 69 (about 13 years, per 1,000 men)
1. What does it find?	Breast cancer, including some that would never cause harm3	Prostate cancer, including many slow-growing ones9
2. How many does it help?	About 8 fewer breast cancer deaths4	About 1.3 fewer prostate cancer deaths9
3. How many get false alarms?	About 1,376 false-positive recalls and 201 benign biopsies4	About 240 men with a positive PSA; 2 hospitalized after a biopsy9
4. How many are overdiagnosed?	About 144	20% to 50% of screen-found cancers9
5. What happens after a positive result?	More imaging, sometimes a biopsy3	Biopsy; if cancer, monitoring, surgery or radiation, with possible leakage or erectile problems9

By our arithmetic, mammography produces about 170 false-positive recalls for each breast cancer death it prevents. The Task Force still judged its benefit to outweigh its harms³. For PSA, the benefit is smaller and the harms of treatment last longer, which is why the decision is personal⁹.

Bring these questions to your appointment:

• If my result is positive, what’s the next step, and how soon?
• Would I need a colonoscopy, biopsy or scan, and what are its risks?
• Does my family history or health change the numbers for me?
• When can I stop this test?

Prevention starts before the scan

Natural-health values and screening science agree more than you might think. Both prize prevention, and both distrust treating healthy bodies without good reason. The Task Force itself recommends against several screenings, including ovarian and thyroid tests¹¹¹.

Habits matter, too. Not smoking is the surest way to lower lung cancer risk, and the Task Force says screening can stop after 15 years without smoking¹⁰. Still, eating well can’t show you a polyp. Screening and healthy habits do different jobs, and you need both.

Questions readers ask

The bottom line

Screening is prevention’s most tempting promise, and some tests keep it. Colorectal screening, mammograms and lung scans for long-time smokers prevent deaths. They also bring false alarms, extra procedures and some overdiagnosed cancers, and you deserve to see both sides in plain numbers.

Tests marketed to healthy people, from whole-body MRIs to multi-cancer blood tests, haven’t yet shown they help people live longer. Ask the five scorecard questions every time, follow through on any positive result, and keep doing the daily habits that lower your risk to begin with.

Keep exploring: when to stop cancer screening and lower blood pressure naturally.

Related reading

• What a medical test can and cannot tell you
• When extra medical tests do more harm than good
• What a wellness panel really costs after you get the results
• How medical decisions are supposed to get made with you
• A 22% risk cut can mean helping 2 or 3 people in 1,000

Sources

1. National Cancer Institute. “Thyroid Cancer Screening (PDQ), Health Professional Version.” NCI, updated May 13, 2025. cancer.gov
2. The ASCO Post. “South Korean Study Sparks Warnings About the Hazards of Overscreening,” reporting Ahn HS, Kim HJ, Welch HG, New England Journal of Medicine, 2014. December 1, 2014. ascopost.com
3. U.S. Preventive Services Task Force. “Breast Cancer: Screening. Final Recommendation Statement.” April 30, 2024. uspreventiveservicestaskforce.org
4. Trentham-Dietz A, Chapman CH, Jayasekera J, et al., CISNET Breast Cancer Working Group. “Collaborative Modeling to Compare Different Breast Cancer Screening Strategies: A Decision Analysis for the US Preventive Services Task Force.” JAMA, 2024. uspreventiveservicestaskforce.org
5. Kaiser Permanente Washington Health Research Institute. “Yearly Mammograms: Expect a False Positive Each Decade,” reporting Hubbard RA, et al., Annals of Internal Medicine, 2011. October 17, 2011. kpwashingtonresearch.org
6. U.S. Preventive Services Task Force. “Colorectal Cancer: Screening. Final Recommendation Statement.” May 18, 2021. uspreventiveservicestaskforce.org
7. National Cancer Institute, Cancer Currents. “Shield Blood Test for Colorectal Cancer Screening.” October 11, 2024. cancer.gov
8. American Society for Gastrointestinal Endoscopy. “ASGE Position on Blood-Based Colorectal Cancer Screening.” ASGE, 2025. asge.org
9. U.S. Preventive Services Task Force. “Prostate Cancer: Screening. Final Recommendation Statement.” May 8, 2018 (update in progress). uspreventiveservicestaskforce.org
10. U.S. Preventive Services Task Force. “Lung Cancer: Screening. Final Recommendation Statement.” March 9, 2021. uspreventiveservicestaskforce.org
11. U.S. Preventive Services Task Force. “Ovarian Cancer: Screening. Final Recommendation Statement.” February 13, 2018. uspreventiveservicestaskforce.org
12. American College of Radiology. “ACR Statement on Screening Total Body MRI.” 2023.
13. Henderson M. “Whole-Body MRI Expands, Leaving Patients Weighing Risks.” RSNA News, July 1, 2026. rsna.org
14. GRAIL (company release). “GRAIL Reports Second Quarter 2026 Financial Results.” August 5, 2026. grail.com
15. U.S. Food and Drug Administration. “24 Hour Summary: Molecular and Clinical Genetics Panel of the Medical Devices Advisory Committee Meeting.” September 23, 2026 (meeting page: fda.gov). fda.gov