The screening tests worth having after 65, and the ones that are oversold

The case for screening seems self-evident. Find the disease before it announces itself, treat it while it is small, live longer. For some conditions this is exactly what happens. For others, decades of careful trials have shown that it does not, and the reasons are worth understanding because they recur.

The two concepts that explain most screening disappointments are lead time and overdiagnosis. Lead time is the interval between detection by screening and when the disease would have surfaced anyway; if treatment does not change the course, screening only lengthens the time you know about it, which looks like longer survival in the statistics without anyone living a day longer. Overdiagnosis is the detection of disease that would never have caused symptoms in your lifetime — real disease, correctly identified, that was never going to matter.

Overdiagnosis becomes more likely with age, because competing causes of death multiply. This is why several screening programmes have upper age limits that patients often experience as being written off, and which are in fact a considered judgement about the balance of benefit and harm.

Screening with solid evidence

Colorectal cancer. The strongest case in the list. Faecal immunochemical testing and colonoscopy both reduce colorectal cancer mortality in randomised trials, and colonoscopy additionally reduces incidence by removing precancerous polyps. Most guidelines screen to around 75, and to 85 on an individualised basis where health is good.

Breast cancer. Mammography reduces breast cancer mortality, with the size of the benefit debated and the overdiagnosis rate genuinely uncertain — estimates range from under 5 per cent to around 30 per cent of screen-detected cancers. Programmes typically run to 70 or 74, with continuation beyond that reasonable in women with a long life expectancy.

Abdominal aortic aneurysm. A single ultrasound in men around 65 reduces aneurysm-related mortality. One scan, once, with a clear benefit. It remains oddly under-publicised.

Blood pressure. Not usually thought of as screening, but it is, and it is among the highest-value things measured in any clinic.

Osteoporosis. DEXA scanning in women over 65 identifies people whose fracture risk can be meaningfully reduced. Fracture prevention in this group is one of the better-supported interventions in geriatrics.

Genuinely contested

Prostate cancer. PSA testing is the standard example of a screening test with real but modest mortality benefit and substantial harm. The large trials disagree with each other; the European ERSPC study found a reduction in prostate cancer deaths, the American PLCO trial did not, and the discrepancy is partly explained by contamination of the control arm. Overdiagnosis is high — many detected cancers would never have progressed — and treatment carries meaningful rates of incontinence and erectile dysfunction. Most guidelines now recommend a shared decision rather than routine testing, and against testing men with less than ten to fifteen years of life expectancy.

Lung cancer. Low-dose CT screening in heavy smokers does reduce lung cancer mortality — the NLST and NELSON trials both showed it. This is one of the more genuinely positive recent findings. The catch is a high false-positive rate and the need for well-organised follow-up, so it works within a programme and poorly outside one.

Where age changes the answer

Guidelines increasingly frame the upper limit in terms of life expectancy rather than chronological age, on the reasoning that a healthy 78-year-old and a frail 68-year-old are different patients. Most cancer screening takes five to ten years to produce a mortality benefit, while the harms — the biopsy, the anxiety, the incidental finding, the operation — arrive immediately. If ten years is not likely, the arithmetic inverts.

This is an uncomfortable conversation and often a badly conducted one. "You're too old for that now" lands very differently from "the benefit of this takes about a decade to arrive, and I want to be sure it is worth what it costs you in the meantime."

The tests being sold that you probably do not need

Private health screening packages have proliferated, and the marketing is considerably ahead of the evidence.

Whole-body MRI in asymptomatic people generates a large number of incidental findings, most of which are harmless and all of which require investigation. Coronary calcium scoring has legitimate uses in refining risk in intermediate-risk patients, and is often sold to people in whom it will not change management. Carotid ultrasound in people without symptoms is recommended against by most bodies. Tumour marker panels — CA-125, CEA and similar — perform poorly as screening tests in the general population and are designed for monitoring known disease.

The multi-cancer early detection blood tests now entering the market are scientifically interesting and, at the time of writing, still short of the randomised evidence that would show they reduce mortality rather than simply finding more cancer earlier.

A reasonable position

Take the programmes your national system offers; they are the ones that survived scrutiny. Have the prostate conversation properly rather than either demanding or refusing the test. Ask about aneurysm screening if you are a man near 65. And treat any package priced like a spa day with the scepticism you would apply to any other product sold on fear.