Cancer Screening Tests Compared
Every cancer screening test EBL grades, side by side. Each letter answers one question: in randomized trials, did people invited to screening die less often of that cancer? Finding more cancers is a different question, and the harms of each test sit beside its letter.
Each letter below is copied from the test's own record and answers whether people invited to screening died less often of the cancer in randomized trials. A test can find more cancers without preventing deaths: screening also finds cancers that would never have caused harm, and a positive result leads to further tests. Death from any cause, the outcome that counts every harm and every benefit together, has its own section at the end.
Colorectal cancer
Scopes and stool tests, graded on deaths from colorectal cancer.
- AHard clinical endpointDevice or therapy · Population randomized trials of an invitation to screening against usual care, and their pooled analysisGrade for: Colorectal cancer deaths, an invitation to flexible sigmoidoscopy, adults aged 55 to 64About 1 colorectal cancer death prevented for every 770 people invited, over 15 years. Juul et al., Ann Intern Med, 2022 · 36215714
- AHard clinical endpointDevice or therapy · Randomized trials of the screening strategy, by testGrade for: Colorectal cancer mortality, guaiac FOBT (gFOBT)Endpoint The outcome a trial measures to decide whether the treatment worked, such as heart attacks, blood pressure or a symptom score.: Colorectal cancer mortality
- CHard clinical endpointDevice or therapy · Randomized trials of the screening strategy, by testGrade for: Colorectal cancer mortality versus colonoscopyEndpoint: Colorectal cancer mortality at 10 years
- CHard clinical endpointDevice or therapy · Population randomized trials of an invitation to screening against usual care, and their pooled analysisGrade for: Colorectal cancer diagnoses, an invitation to screening colonoscopy, adults aged 55 to 64About 1 colorectal cancer prevented for every 455 people invited, over 10 years; 15 of 11,843 people screened had major bleeding after polyp removal. Bretthauer et al., N Engl J Med, 2022 · 36214590
- DHard clinical endpoint · Tested, no benefitDevice or therapy · Population randomized trials of an invitation to screening against usual care, and their pooled analysisGrade for: Colorectal cancer deaths and death from any cause, an invitation to screening colonoscopy, adults aged 55 to 64
- DHard clinical endpoint · Not yet testedDevice or therapy · Randomized trials of the screening strategy, by testGrade for: Colorectal cancer mortality, fecal immunochemical test (FIT) specificallyEndpoint: Colorectal cancer mortality
Breast cancer
Mammography, graded on deaths from breast cancer.
- BHard clinical endpointDevice or therapy · Meta-analyses A study that pools the numbers from several trials into one combined estimate. of randomized trials of mammography screening against no screeningGrade for: Breast cancer deaths, women aged 50 to 69
- DHard clinical endpoint · Tested, no benefitDevice or therapy · Meta-analyses of randomized trials of mammography screening against no screeningGrade for: Breast cancer deaths, women aged 39 to 49
Lung cancer
Low-dose CT in current and former heavy smokers.
- AHard clinical endpointDevice or therapy · Large randomized trials of low-dose CT screening against chest X-ray or no screeningGrade for: Lung cancer deaths, current and former heavy smokers aged 50 to 74Lung cancer deaths 20% lower than with chest X-ray; 24.2% of scans were positive, and 96.4% of positive results were false alarms. National Lung Screening Trial Research Team, N Engl J Med, 2011 · 21714641
Prostate cancer
The PSA blood test, graded on deaths from prostate cancer.
- BHard clinical endpointDevice or therapy · Large randomized trials of PSA screening against usual careGrade for: Prostate cancer deaths, men aged 55 to 69At 16 years, 570 men invited and 18 men diagnosed for each prostate cancer death prevented; many of the cancers found would never have caused harm. Hugosson et al., Eur Urol, 2019 · 30824296
Cervical cancer
HPV testing, against the Pap test and against no screening.
- BHard clinical endpointDevice or therapy · Randomized trials of HPV-based screening against cytology or no screeningGrade for: Invasive cervical cancer, HPV-based against cytology-based screening, women aged 20 to 64
- CHard clinical endpointDevice or therapy · Randomized trials of HPV-based screening against cytology or no screeningGrade for: Cervical cancer deaths, one round of HPV testing against standard care, women aged 30 to 59 in rural India
Multi-cancer and whole-body tests
Tests sold to look for many cancers at once, and general check-ups. Disclosure for the multi-cancer blood test: EBL's founder holds an unpaid provider and affiliate account with GRAIL, which sells the test and funded the NHS-Galleri trial cited below. GRAIL pays EBL nothing: no commission and no fee. The letter was set from the trials, and the account does not move it.
- DSurrogate biomarker · Tested, no benefitDevice or therapy · Tested in one large randomized trial; primary endpoint The one outcome a trial names in advance as its main test. Results on other outcomes carry less weight. not metEndpoint: Incidence of late-stage (III and IV) cancer at diagnosis, a stage-shift surrogate for cancer mortality
- DHard clinical endpoint · Not yet testedDevice or therapy · No randomized outcome trial; pooled observational Research that watches what people already do rather than assigning a treatment. It can show that two things travel together, not that one causes the other. yield data onlyEndpoint: Cancer mortality or all-cause mortality in asymptomatic screened adults
- DHard clinical endpoint · Tested, no benefitProtocol · Cochrane meta-analysis of randomized trials (found no mortality benefit → D)Endpoint: All-cause, cardiovascular and cancer mortality in adults unselected for disease
Death from any cause
Whether screening changed how many people died of anything. A pooled analysis of the screening trials found a significant gain in lifetime only for sigmoidoscopy.
- CHard clinical endpointDevice or therapy · Large randomized trials of low-dose CT screening against chest X-ray or no screeningGrade for: Death from any cause, current and former heavy smokers
- DHard clinical endpoint · Tested, no benefitDevice or therapy · Meta-analyses of randomized trials of mammography screening against no screeningGrade for: Death from any cause, women invited to mammography screening
- DHard clinical endpoint · Tested, no benefitDevice or therapy · Large randomized trials of PSA screening against usual careGrade for: Death from any cause, men invited to PSA screening