Annals of Internal Medicine invites me to read its guidance article about mammography screening but won’t publish my highly relevant criticism of it
In April, Christine Laine, Editor-in-Chief of Annals of Internal Medicine, wrote to me out of the blue that I might be interested in a paper they had just published, given my important contributions in the area.
It is extremely rare to get a personal email from the editor of a major medical journal alerting you to new paper because of your expertise. I read the paper, which is a guidance statement from the American College of Physicians about mammography screening.1
I replied that all guidelines are hopeless because they do not take into consideration that breast cancer mortality in the randomised trials is a biased outcome, which I demonstrated 25 years ago2 and many times since,3,4 including in our well-known Cochrane review.5
I noted that it is heartbreaking that the women have been disinformed for over 40 years and that Laine’s colleague, Fiona Godlee, previous Editor-in-Chief of the BMJ, declared at a meeting in Oxford where I lectured about mammography screening that she did not go to screening because of my research and that my wife, who is also very bright, says the same. This lecture is publicly available.6
I suggested Laine read my recent article, “Why mammography screening doesn’t and cannot work. And how cancer charities deliberately disinform the public,”7 which said it all, and offered to send something like this to Annals.
I noted that my most recent paper in Annals was very telling. We documented in 2017, once again, that screening cannot work because it doesn’t lower the incidence of advanced cancers.8 We did the study in Denmark, which has an ideal control group because women in 80% of the country were not invited to screening for 17 years.
Laine encouraged me to post a comment to the paper on the web and said that selected comments were also published in the journal. I could also submit a longer commentary “for consideration during our usual peer review process.”
As the issue was very important, and as several statements in the Annals article were false, I submitted a commentary (see full text below).
It took only six days for Laine to reject my article after internal review by the editors, without sending it for peer review as she had promised. She encouraged me to upload a comment on their website instead.
Electronic comments on journal websites are not indexed on PubMed, the US National Library of Medicine’s database of abstracts of scientific articles and letters to the editor in print journals. They are therefore virtually unknown to the scientific community.
It is convenient for medical journals to protect their prestige by avoiding publishing letters to the editor in print that challenge their editorial decisions. I have often discussed this with top researchers. As an example, the late biostatistician Douglas Altman and I have both experienced that every time we submitted a letter to the New England Journal of Medicine about a fatal flaw in research they had just published, it was rejected.
I have described how the editors of this journal, Jeffrey Drazen and Eric J Rubin, refused to inform their readers about fraud in drug trials they had published on Vioxx and a Covid-19 vaccine, respectively.9 The journal is so beholden to the drug industry that it has been nicknamed the New England Journal of Medicalization, and psychiatrist David Healy, who described the Covid-19 vaccine scandal, called it New England Journal of Misinformation.10
I was convinced that my e-letter would not make it to the print journal and I was right. By not publishing it, Annals not only protected their own reputation but also the official, but false narrative, that mammography screening saves lives and does more good than harm.
The Lancet also supported the false narrative when they published a flawed review in 2012 and disinformed the media,11 which killed a debate I started in The Lancet in 2000.12 In 2001, we published the most detailed review of the mammography screening trials ever performed, also in The Lancet,2 which led to a flurry of meta-analyses, none of which were of high quality.3,4
Annals have a long history of protecting mammography screening by publishing flawed, substandard reviews. In 2002, the US Preventive Services Task Force published a seriously misleading review in Annals.13 There was nothing about bias in using breast cancer mortality as the outcome, and there were no data on all-cause mortality or on all-cancer mortality. Under harms, only false positive tests, unnecessary anxiety, biopsies and cost were listed. It is misleading to only talk about unnecessary biopsies when screening increases both tumourectomies and mastectomies substantially.2,5,14
A 2007, a practice guideline in Annals from the American College of Physicians cited our 2001 Lancet review, but not for its data on overdiagnosis.15 Instead, a flawed observational study from Firenze by Stephen Duffy and others was taken as evidence that screening decreases the risk of mastectomy. I called this statistical alchemy.4 In the randomised trials, screening increased mastectomies by 20%,2,5 and our long-term data from Denmark demonstrated that the initial increase in mastectomies when screening is started in a region is not compensated by a drop in mastectomies later on.14
In 2016, Annals published an updated review by the US Preventive Services Task Force,16 which mentioned that all-cause mortality was not reduced with screening and that screening increases mastectomies and tumourectomies, quoting the 2013 version of our Cochrane review.5
However, this review was also hugely misleading. The authors included all the trials in the summary estimate for breast cancer mortality even though only three of them had been adequately randomised. We showed already in 200012 that the two groups of trials, those with adequate and those with inadequate randomisation, had confidence intervals for breast cancer mortality that did not overlap, which is huge heterogeneity. We therefore divided our analyses in two groups in all our meta-analyses, including in our Cochrane review.5 The investigators wrote that they assessed statistical heterogenity but they didn’t pay any attention to it.
The review claimed that screening reduces the occurrence of advanced breast cancer, which is false. This result was driven by two trials, those from New York and the Swedish Two-County trial, whose data we have demonstrated are grossly unreliable.2,5,12
In New York, more women with previous breast cancer were excluded from the screened group than from the control group, which the investigators have admitted.5 It therefore looks like scientific misconduct that the authors of the Annals review, who had read our Cochrane review that gives this information, ignored it.
There are many indications that the Swedish Two-County trial may be fraudulent.3-5 Its results are surprisingly positive, and the mean time of randomisation, as well as the number of women included in the study and the number of breast cancer deaths vary between different papers from the trial, even when the age groups and follow-up periods are the same.5,17 Furthermore, my research group showed that breast cancer incidence
and breast cancer mortality are not compatible with contemporaneous official Swedish breast cancer statistics,17 which is very grave because the researchers used the same cancer register as we did.
Cochrane censorship
The worst censorship came from my own organisation, the Cochrane Collaboration, which I co-founded in 1993 and for which I was the Nordic director.18 In 2001, we submitted our Cochrane review of mammography screening review to the Australian-based Cochrane Breast Cancer Group, which had a financial conflict of interest, as it was funded by the centre that offered breast screening in Australia.
The editors refused flatly to let us include data on the most important harms of cancer screening, overdiagnosis and overtreatment of healthy people, even though such outcomes were listed in our protocol the group had accepted and published. It took me five years, with repeated complaints to the Cochrane leadership, before we were allowed to publish these data in our review.3
Cochrane did not learn from its grave mistake. In 2025, Cochrane created another huge scandal when they rejected our update where we had added more deaths and where it was clearer than ever that screening does not save lives.18
I hope you will read also what I published on the website of Annals19 and what I submitted as a commentary (below the references).
Tell your female friends that they are being defrauded by opinion leaders, the media, cancer charities, and boards of health3,4,11 and that they are wise if they decide not to attend mammography screening.
Somewhat ironically, the total war against common sense, which has harmed women and
caused many of them to lose one or both breasts they would not have lost if there had not been screening, has almost exclusively been directed by aggressive males.3,4 I shall therefore end my article with a cartoon:
References
1 Qaseem A, Harrod CS, Balk EM, et al. Screening for breast cancer in asymptomatic, average-risk adult females: a guidance statement from the American College of Physicians (Version 2). Ann Intern Med 2026;Apr 17. doi: 10.7326/ANNALS-25-05116. Epub ahead of print.
2 Olsen O, Gøtzsche PC. Cochrane review on screening for breast cancer with mammography. Lancet 2001;358:1340-2 and Olsen O, Gøtzsche PC. Systematic review of screening for breast cancer with mammography. Lancet 2001;Oct 20.
3 Gøtzsche PC. Mammography screening: truth, lies and controversy. London: Radcliffe Publishing; 2012.
4 Gøtzsche PC. Mammography screening: the great hoax. Copenhagen: Institute for Scientific Freedom; 2024 (freely available).
5 Gøtzsche PC, Jørgensen KJ. Screening for breast cancer with mammography. Cochrane Database Sys Rev 2013;6:CD001877.
6 Gøtzsche PC. Controversies in evidence – the case against screening in breast cancer. Evidence Live lecture. Oxford 2013;March 25.
7 Gøtzsche PC. Why mammography screening doesn’t and cannot work. And how cancer charities deliberately disinform the public. Substack 2026;March 29.
8 Jørgensen KJ, Gøtzsche PC, Kalager M, Zahl PH. Breast cancer screening in Denmark: a cohort study of tumor size and overdiagnosis. Ann Intern Med 2017;166:313-23.
9 Gøtzsche PC. Whistleblower in healthcare (autobiography). Copenhagen: Institute for Scientific Freedom 2025 (freely available).
10 Healy D. New England Journal of Misinformation. Healy’s blog 2021;Nov 22.
11 Gøtzsche PC. How the UK Cancer Empire killed the debate on mammography screening. And how The Lancet starred as the establishment’s useful idiot. Gøtzsche’s Perspective 2026;July 7.
12 Gøtzsche PC, Olsen O. Is screening for breast cancer with mammography justifiable? Lancet 2000;355:129-34.
13 Humphrey LL, Helfand M, Chan BK, et al. Breast cancer screening: a summary of the evidence for the U.S. Preventive Services Task Force. Ann Intern Med 2002;137(5 Pt.1):347–60.
14 Jørgensen KJ, Keen JD, Gøtzsche PC. Is mammographic screening justifiable considering its substantial overdiagnosis rate and minor effect on mortality? Radiology 2011;260:621-7.
15 Qaseem A, Snow V, Sherif K, et al. Screening mammography for women 40 to 49 years of age: a clinical practice guideline from the American College of Physicians. Ann Intern Med 2007;146:511–5.
16 Nelson HD, Fu R, Cantor A, et al. Effectiveness of breast cancer screening: systematic review and meta-analysis to update the 2009 U.S. Preventive Services Task Force Recommendation. Ann Intern Med 2016;164:244-55.
17 Zahl PH, Gøtzsche PC, Andersen JM, Maehlen J. Results of the Two-County trial of mammography screening are not compatible with contemporaneous official Swedish breast cancer statistics. Dan Med Bull 2006;53:438-40.
18 Gøtzsche PC. Cochrane editorial misconduct: our review of mammography screening. Substack 2026;Jan 17.
19 Gøtzsche PC. Guidance about breast screening overlooks the elephant in the room. Ann Intern Med 2026;May 11.
***
Here is the comment I published on the website of Annals.19 Only 400 words and 5 references are allowed.
Guidance about breast screening overlooks the elephant in the room
All guidelines about mammography screening, including the new one in Annals (1), suffer from a fundamental flaw. They assume that the effect of screening on breast cancer mortality can be translated into an effect on total mortality.
Breast cancer mortality is a biased outcome (2). When overdiagnosed healthy women get treated with radiotherapy or chemotherapy, it increases mortality. And assessment of cause of death in the randomized trials was biased in favor of screening (2).
We therefore need to look at total cancer mortality including breast cancer mortality and total mortality instead. The most updated results for the four adequately randomized trials showed a risk ratio of 1.00 (95% confidence interval 0.96 to 1.04) for total cancer mortality and 1.01 for total mortality (0.98 to 1.03) (3). These are very strong results as they are derived from a total of 25,046 deaths.
In addition, the overdiagnosis means that more women lose a breast if they go to screening (2,4). We don’t hear anything about this, only the opposite. For example, the American Cancer Society claims on their website, with no references, that results from many decades of research clearly show that women who have regular mammograms are less likely to need aggressive treatments like surgery to remove the entire breast (mastectomy).
The guideline in Annals say that that the evidence is insufficient or low-quality that screening causes overdiagnosis of invasive or in situ cancers. This is hugely misleading. The randomized trials (2) and systematic reviews of organized screening programmes with a relevant control group (5) show considerable overdiagnosis, ranging from 30% to 52%. This research is not low-certainty evidence.
Mammography is a very poor screening method that should be abandoned because it is harmful and has no benefits (2-4).
References
1 Qaseem A, Harrod CS, Balk EM, et al. Screening for breast cancer in asymptomatic, average-risk adult females: a guidance statement from the American College of Physicians (Version 2). Ann Intern Med 2026; Apr 17. doi: 10.7326/ANNALS-25-05116. Epub ahead of print. PMID: 41996183.
2 Gøtzsche PC, Jørgensen KJ. Screening for breast cancer with mammography. Cochrane Database Sys Rev 2013;6:CD001877.
3 Gøtzsche PC. The three big lies about mammography screening. Brownstone Journal 2026;March 6.
4 Jørgensen KJ, Keen JD, Gøtzsche PC. Is mammographic screening justifiable considering its substantial overdiagnosis rate and minor effect on mortality? Radiology 2011;260:621-7.
5 Jørgensen KJ, Gøtzsche PC. Overdiagnosis in publicly organised mammography screening programmes: systematic review of incidence trends. BMJ 2009;339:b2587.
And here is what I submitted to Annals as a commentary.
Guidance about breast screening overlooks the elephant in the room
The many guidelines about mammography screening that have been published in the last 20 years all suffer from a fundamental flaw. They assume that an effect of screening on breast cancer mortality can be translated into an effect on total mortality.
Accordingly, the official propaganda about screening, e.g. from the American Cancer Society and screening zealots, states that screening saves lives.1,2
However, we have known for 25 years that breast cancer mortality is a biased outcome. There are two reasons for this.3 First, when overdiagnosed healthy women get treated with radiotherapy or chemotherapy, it increases mortality. Second, assessment of cause of death in the randomized trials was biased in favor of screening. One of the issues was that some women had more than one cancer, e.g. 21% in the Malmö trial. Patients with cachexia and no signs of recurrence of breast cancer would likely be assigned to another type of cancer.
I demonstrated that those trials that reported the greatest reduction in advanced cancers with screening also had the biggest reduction in breast cancer mortality, which would be expected if screening worked, but there was a problem and it was huge.4
The linear relation between advanced cancers and breast cancer mortality was in the wrong place. A screening effectiveness of zero, meaning that the rate of node-positive cancers is the same in the screened group as in the control group, corresponded to a highly significant 16% reduction in breast cancer mortality (P < 0.001). My further analyses demonstrated that assessment of cause of death and of the number of cancers in advanced stages were both biased in favor of screening. And since the size of the bias, 16%, was similar to the estimated effect of screening, if all the trials are included,3 even the very poor and biased ones, this result suggests that screening is ineffective.
Since breast cancer mortality is biased, we need to look at total cancer mortality including breast cancer mortality and total mortality instead. The most updated results for the four adequately randomized trials showed a risk ratio of 1.00 (95% confidence interval 0.96 to 1.04) for total cancer mortality and 1.01 for total mortality (0.98 to 1.03).2 This is a very strong result as it is derived from a total of 25,046 deaths.
What is also part of the official propaganda is that by detecting cancers early, less invasive surgery is needed. For example, the American Cancer Society claims, with no references, that results from many decades of research clearly show that women who have regular mammograms are less likely to need aggressive treatments like surgery to remove the entire breast (mastectomy).2
These claims are wrong. If we assume that tumor growth is constant, then the women have harbored the tumor for 21 years on average before they can be detected on a mammogram.5 This is not early detection.
In addition, the overdiagnosis means that more women lose a breast if they go to screening. The trials found 20% more mastectomies in the screened groups.3 And data from Denmark, which has an ideal control group because screening was only practiced in 20% of the country for 17 years, show a large increase in mastectomies when screening started that was not compensated by fewer mastectomies later.6
We don’t hear anything about this in guidelines. They are therefore all misleading.
The guidance just published in Annals by the American College of Physicians is also misleading because it does not take these issues into account.7 The College recommends that for women aged 50 to 74 years, clinicians should use biennial mammography for screening for breast cancer.
The College mentions that the risk of receiving a false-positive result is 50-60% if a woman gets ten mammograms but that a European guideline found low-certainty evidence that false positives increase psychological distress.
High-quality research contradicts this. Generic questionnaires cannot capture fully the psychological consequences of breast screening. A Danish study based on a specially designed questionnaire that was developed using focus groups found that, even after three years, women who had experienced a false-positive diagnosis had an anxiety level and other psychological problems that fell between that for women with breast cancer and women who were told they did not have cancer.8
I find it confusing that the College uses Grade terms when discussing if the science is reliable, i.e. insufficient, low-, moderate- or high-certainty evidence. This grading is subjective and does not tell the readers what the issues are and if they are important.
We are told that there is low-certainly evidence that screening lowers the occurrence of advanced cancers but in our nation-wide study of Danish data, we found no such effect.9
We are also told that the evidence is insufficient or low-quality that screening causes overdiagnosis of invasive or in situ cancers. This is hugely misleading. The randomized trials,3 a systematic review of organized screening programmes,10 and studies of Danish data9 all show considerable overdiagnosis, ranging from 30% to 52%. This research is not low-certainty evidence.
Mammography is a very poor screening method that should be abandoned because it is harmful and has no benefits.
References
1 Gøtzsche PC. Mammography screening: truth, lies and controversy. London: Radcliffe Publishing; 2012.
2 Gøtzsche PC. The Three Big Lies about Mammography Screening. Brownstone Journal 2026;March 6.
3 Gøtzsche PC, Jørgensen KJ. Screening for breast cancer with mammography. Cochrane Database Sys Rev 2013;6:CD001877.
4 Gøtzsche PC. Relation between breast cancer mortality and screening effectiveness: systematic review of the mammography trials. Dan Med Bull 2011;58:A4246.
5 Gøtzsche PC, Jørgensen KJ, Zahl PH, Maehlen J. Why mammography screening has not lived up to expectations from the randomised trials. Cancer Causes Control 2012;23:15-21.
6 Jørgensen KJ, Keen JD, Gøtzsche PC. Is mammographic screening justifiable considering its substantial overdiagnosis rate and minor effect on mortality? Radiology 2011;260:621-7.
7 Qaseem A, Harrod CS, Balk EM, et al. Screening for breast cancer in asymptomatic, average-risk adult females: a guidance statement from the American College of Physicians (Version 2). Ann Intern Med 2026;Apr 17. doi: 10.7326/ANNALS-25-05116. Epub ahead of print.
8 Brodersen J and Siersma VD. Long-term psychosocial consequences of false-positive screening mammography. Ann Fam Med 2013;11:106-15.
9 Jørgensen KJ, Gøtzsche PC, Kalager M, Zahl PH. Breast cancer screening in Denmark: a cohort study of tumor size and overdiagnosis. Ann Intern Med 2017;166:313-23.
10 Jørgensen KJ, Gøtzsche PC. Overdiagnosis in publicly organised mammography screening programmes: systematic review of incidence trends. BMJ 2009;339:b2587.
Gøtzsche PC. Why you should not get screened for prostate cancer. The American Cancer Society misleads the public and enrich themselves. Gøtzsche's Perspective 2026;March 16.
no type of breast screening has been documented to work