Does mass screening - especially for cancer - cause more harm than good?

We can't be sure but need to confront the question.

Via the below post on X…

Joseph Marine@DrJMarine

Brilliant essay and reflection – one of the most thought-provoking I have ever read on a medical topic. The immediate subject is screening mammography, but it touches on the entire modern preventive medical ecosystem and shifting public sentiments in covid's wake. The quality of

Mo Perry @MoMoPerry

For @tabletmag, I wrote about mammograms, the epistemic ruins of public health, and the implications of unplugging from the consensus-making machine. https://t.co/aKPnvcfrFn

6:58 PM · Feb 6, 2026 · 52.1K Views

---

7 Replies · 39 Reposts · 179 Likes

…I was alerted to this well-written piece about mammography screening in an American magazine called Tablet:

That piece - which is worth reading in full - actually articulated much of my own feelings about mass screening programs in general.

Broadly speaking, their benefits are hugely overblown, and the harms (through triggering chains of complex treatment for findings which turn out to be incidental, or harmless) are massively underestimated.

However, there are 2 additional points which the authors did not make, which are worthy of consideration.

1. The potential that the test itself causes direct harm

Mammography involves exposure to ionizing radiation, which is technically deleterious to health in the sense that any such exposure carries a theoretical risk of actually causing cancer.

Most screening proponents claim that the dose is very small and the risk-to-benefit ratio is favourable.

But such assessments are based on calculations which involve substantial uncertainty. For example:

• the effects of cumulative exposure from yearly or bi-annual mammograms are essentially unknown. It may well be that the risk curve moves sharply upwards after a certain amount of exposure.

• although it is often said that the radiation from a screening mammogram is equivalent to the natural background radiation received from the environment over a few months, in the former the radiation is being delivered in seconds to a concentrated small volume of tissue, and any comparison is (at best) misleading.

• the benefits are - as I mentioned above - hugely overestimated for many reasons which I don’t intend going into here, but in brief, centre around recent improvements in treatment (benefits of early detection are based on older outcome data) and a phenomenon known as “lead-time bias” (ie that earlier detection in itself makes survival look longer without actually delaying death).

2. The interventions themselves are directly harmful.

The article does go into a little detail about the potential harms that can be caused by the worry and anxiety that a finding of possible cancer can cause, though in my opinion there is a lot more that could be said.

Stress is known to be extremely harmful to health, though less is known about the precise biological mechanisms involved. It is likely that inflammatory pathways are activated. Regardless of the way in which stress and anxiety is harmful, the fact is that the only way of truly accounting for it in any assessment of overall - or net - benefit is to look at all-cause mortality and morbidity. However, proponents of mass screening rarely do that - they just look at the target disease. This myopic focus fails to take into consideration the complexity of human biology. This is a theme of much of my critique of the modern biomedical-industrial complex.

The adverse effects of screening programs are not limited to radiation and stress, however.

In my view, the harms caused by the chain of procedures which follows a finding of potential cancer have been hugely underappreciated.

It seems to me that there is evidence to strongly suggest that the biopsy process itself might be causing some lesions which would - if left alone - be harmless to become dangerous.

There are 2 main pieces of evidence for this:

Biopsies are contra-indicated for certain cancers

Because of the chance of cancerous cells traveling down the needle track and spreading the cancer, needle biopsies are generally contraindicated for certain types of cancer - notably those of the testicle, some kidney tumours and certain suspected ovarian cancers.

This is not thought to be an issue other common forms of cancer, such as those in the breast, though I am not sure how robust the reasoning for that distinction really is.

Incidental findings of cancer are common - if looked for

However, what is known is that cancers are frequently found as incidental findings - if looked for - in people who die of unrelated causes.

I had often wondered how frequent such findings were, and the Tablet article above actually linked to this very interesting 2017 paper:

The above is a systematic review and meta-analysis of the data obtained from 13 autopsy studies between 1948 to 2010 involving 2,363 women who had no history of breast disease during their lifetime. The aim was to estimate the prevalence of incidental breast cancer.

The findings are stunning:

Incidental breast cancer was found in nearly one in five autopsies.

The precise breakdown of the stage of the cancer identified was:

• Invasive cancer: 0.85%

• In-situ cancer (DCIS/LCIS): 8.9%

• Atypical hyperplasia: 9.8%

Interestingly, there was only weak correlation between the incidence of the detection of such cancers and increasing age, but a very strong link between more detailed examination resulting in more cancers being found.

The bottom line is that if you look hard enough at any woman’s breast tissue under a high-powered microscope, you have a roughly 20% chance of finding something that looks like cancer (or a precursor).

Now, clearly mammography does not find anywhere near this number of cancers. But what the study shows us is that many of those that it does find ought not necessarily to be regarded as something which is the precursor of an inevitably fatal and progressive disease, but rather something which is part of the aging process and able to be contained by the patient’s immune system so as to render it harmless.

I am not claiming here that we know how to differentiate between the ones which would not progress and the ones which will. But what I am claiming is that this finding upends many of the mathematical models which support the contention that mass mammography is net beneficial, as the upsides of early detection and removal may have been hugely overstated.

Moreover, it seems that the combination of two phenomena discussed here might be particularly harmful: the possibility that ultimately harmless lesions might be being transformed into something more dangerous, either by the radiation from repeated mammography itself, or by the procedures used to investigate once a suspicious lesion has been detected (ie biopsies).

The existence of substantial numbers of incidental cancers has not received the consideration it should have by the proponents of screening. Their belief system seems to be that removing anything suspicious is automatically net beneficial, hence they generally argue for screening more often, with increasingly more sophisticated systems, which can detect ever smaller lesions.

That approach seems due a big rethink, and it seems doubtful that this applies only to mammography.

Finally, it is worth pointing out that many of the themes I have discussed above have been debated by “mainstream” doctors and scientists over many years - falling largely on deaf ears, probably because the industry around screening and investigation to detect early cancer is huge and growing extremely quickly:

Here are some further articles written on the topic, with thanks to Dan for the spot:

1. Here’s a 2025 Daily Mail piece from Prof Susan Bewley (Professor of Women’s Health Kings College London) titled I’m a top women’s doctor and I would NEVER take up an invitation for a breast screening. Women are being lied to by the NHS and this is the REAL truth about mammograms you must know.

2. From the same author in the BMJ (from 2020): Things should never be the same again in the screening world

3. A WHO bulletin from 2008 (an article I doubt they would consider publishing today because of commercial capture): Screening: evidence and practice

4. A 2011 article on Boston University’s website: Overdiagnosis: Bad for You, Good for Business - SPH Bicknell lecturer says too much treatment makes people sick. (Another article which might not get published today by a major university.)