Cancer Screening: Which Tests Are Worth Doing, and at What Age
Screening means testing people who feel completely well. That is a different proposition from investigating a symptom, and it carries a different balance of benefit and harm. A test only earns a place in a screening programme if it has been shown, in large trials, to reduce deaths from the disease.
Why more screening is not automatically better
Three problems recur.
Overdiagnosis. Finding a cancer that would never have caused symptoms or harm in that person’s lifetime. The person is then treated, with all the associated harms, for a disease that was never going to hurt them. This is not a misdiagnosis; the cancer is real. It simply would not have mattered.
False positives. An abnormal result that turns out to be nothing, after a period of fear and a sequence of further tests, some of them invasive.
False reassurance. A normal result shortly before a cancer presents, which can delay response to symptoms.
Screening programmes exist where trials have shown that the mortality benefit outweighs these harms. Tests outside those programmes often have not been shown to.
The screening tests with good evidence
Cervical cancer. HPV testing, with or without cytology, has largely replaced cytology alone because it is more sensitive. Typical approach: from age 25 to 30, every five years with HPV testing, up to around 65. HPV vaccination substantially reduces risk but does not remove the need for screening.
Breast cancer. Mammography, generally from 40 to 50 depending on the country, every one to two years, continuing to around 74. The benefit is real and the overdiagnosis rate is debated. Women with a strong family history or known genetic mutation need earlier and more intensive surveillance, often including MRI.
Colorectal cancer. Several options with good evidence: faecal immunochemical testing annually or biennially, or colonoscopy every ten years, or flexible sigmoidoscopy. Starting age has moved to 45 in several countries because of rising incidence in younger adults. Colonoscopy has the advantage of removing precancerous polyps, so it prevents cancer rather than only detecting it.
Lung cancer. Low-dose CT annually for people at high risk, defined by age and smoking history, typically 50 to 80 with a substantial pack-year history who currently smoke or quit within the past 15 years. This has been shown to reduce lung cancer deaths and is substantially underused.
Hepatitis B and C. Screening at-risk and, in several countries, all adults once, because both are treatable and both cause liver cancer.
Tests requiring a conversation rather than a default
Prostate cancer, PSA. Reduces prostate cancer deaths modestly while causing substantial overdiagnosis. Shared decision making from 50, or earlier with risk factors. The modern MRI-first pathway reduces the harms considerably.
Keep reading
The rest of this guide is free. Watch a short ad to unlock it and support independent health writing.
Preparing your ad…
Ovarian cancer. No screening test has been shown to reduce deaths in average-risk women. CA-125 and ultrasound produce many false positives leading to surgery. Women with BRCA mutations or strong family history need specialist management, which is a different situation.
Skin cancer. Routine whole-population screening is not established, but self-examination and clinician examination make sense for high-risk individuals: many moles, previous skin cancer, fair skin with significant sun exposure, immunosuppression.
Tests marketed widely with weak or no evidence
Whole-body MRI or CT for healthy people. Not shown to reduce mortality. Produces a high rate of incidental findings requiring investigation, with associated anxiety, cost and procedural risk. CT additionally delivers radiation.
Coronary calcium scoring has a legitimate role in cardiovascular risk stratification in selected people, which is different from cancer screening, but is often bundled into “executive health packages”.
Tumour marker panels in healthy people. CA-125, CEA, CA 19-9, AFP and similar are designed for monitoring known disease, not for detection in asymptomatic people. Used as screening they generate false positives frequently and miss cancers often.
Multi-cancer early detection blood tests are a genuinely promising technology currently in trials. They are being marketed in some places ahead of evidence that they reduce mortality. The honest position is that we do not yet know, and a negative result should not reduce participation in established screening.
Risk-based adjustments that do matter
Family history changes the plan. Worth telling your doctor about:
- Breast, ovarian, prostate or pancreatic cancer in close relatives, particularly at young ages, which may indicate BRCA-related risk
- Multiple relatives with colorectal or endometrial cancer, which may indicate Lynch syndrome
- Any cancer diagnosed unusually young in the family
These may justify earlier starts, shorter intervals, different tests, or genetic counselling.
Prevention outperforms screening
For several cancers, prevention is more effective than any screening test:
- Not smoking, which is by far the single largest modifiable factor
- HPV vaccination, which prevents cervical and several other cancers
- Hepatitis B vaccination, which prevents liver cancer
- Limiting alcohol, a group 1 carcinogen associated with cancers of the mouth, throat, oesophagus, liver, colon and breast
- Maintaining a healthy weight, with obesity linked to at least thirteen cancers
- Sun protection
- Physical activity
- Treating H. pylori, which reduces gastric cancer risk
Symptoms override screening schedules
Screening is for people without symptoms. A new symptom is investigated regardless of when the last screening test was and what it showed.
Worth prompt assessment: unexplained weight loss, a lump, unexplained bleeding from any site, a change in bowel habit lasting weeks, persistent cough or hoarseness, difficulty swallowing, a non-healing sore, a changing mole, persistent bloating, or unexplained persistent pain.
The short version
A short list of screening tests has been shown to save lives: cervical, breast, colorectal, lung in high-risk smokers, and hepatitis. Prostate needs a conversation. Whole-body scans and tumour marker panels sold to healthy people have not been shown to help and reliably cause harm through incidental findings. And no screening result should delay acting on a new symptom.