A study published in JAMA Internal Medicine looks at lifetime gained with cancer screening tests.
Prof Stephen Duffy, Professor of Cancer Screening, Queen Mary University of London (QMUL), said:
“From its title, one would have expected this paper to be based on analysis of individual lifetime data. However, it is not. The paper’s conclusions are based on arithmetic manipulation of relative rates of all-cause mortality in some of the screening trials. It is therefore difficult to give credence to the claim that screening largely does not extend expected lifetime. In addition, the use of all-cause mortality can give seriously misleading results (see https://pubmed.ncbi.nlm.nih.gov/28533316/; https://pubmed.ncbi.nlm.nih.gov/34851217/; https://pubmed.ncbi.nlm.nih.gov/33099453/).
“It is of some concern that the analysis includes the Canadian National Breast Screening Study as there is now public domain evidence of subversion of the randomisation in this trial (https://pubmed.ncbi.nlm.nih.gov/34812692/; https://borealisdata.ca/dataset.xhtml?persistentId=doi:10.5683/SP3/2DEY36).
“Because of the above, the conclusion that cancer screening does not generally increase life expectancy is not borne out by the evidence.”
Dr Leigh Jackson, Lecturer in Genomic Medicine, University of Exeter, said:
Does the press release accurately reflect the science?
“The press release is a reasonable reflection of the science, however, the focus on 2.1 million individuals is slightly misleading. The study considered many different screening tests and 2.1 million was indeed the total number of included patients however no calculation included that many people.”
Is this good quality research? Are the conclusions backed up by solid data?
“The study is methodologically sound with some limitations which the authors clearly state. I would question the inclusion of colonoscopy based on one included study as this is simply a re-reporting of that study and not a meta-analysis so I’m unsure what that adds. The data seem to clearly show there is huge variation in estimates around survival benefits of screening but with the current best evidence suggesting no major increase in all but sigmoidoscopy screening.”
How does this work fit with the existing evidence?
“The study is a good synthesis of existing high-quality studies and aims to provide an overall estimate based on the best current data.”
Have the authors accounted for confounders?
“The authors have accounted for confounders where possible although they acknowledge this was not always possible and caveat their findings accordingly.”
Are there important limitations to be aware of?
“The study had some particular limitations in length of follow-up. This may have limited the amount of data included and also not considering longer follow-up may tend to underestimate the effects of screening.”
What are the implications in the real world?
“Care should be taken when advising patients on screening based solely on the promise of extended life. Whilst for some people this is undoubtedly true, others may in fact experience a shorter life as a result of screening and on average an individual can expect to not experience any extra life expectancy for all tests other than sigmoidoscopy.”
Is there any over-speculation?
“Limited evidence on colonoscopy and the absence of evidence for cervical screening might suggest these tests should not be included In this study to avoid confusion or over-interpretation. The title, abstract and conclusion give the impression these findings cover all screening tests which is not the case. Conclusions should be limited and specific to those tests which had good data to assess.”
Is there any evidence missing from the meta-analysis that might suggest common cancer screening is effective?
“Studies with longer-term follow-up would be useful to see the effects further. Also, whilst RCTs are gold standard studies, it may be informative to include, perhaps separately, real-world observational studies. Trial participants are not always representative of the wider population. The study also only considered life expectancy, there are many other metrics and outcomes that may demonstrate screening efficacy.”
Is this meta-analysis evidence to suggest we should stop screening for common cancers? How does this shift the risk vs benefits of cancer screening?
“The authors themselves do not advocate the abolition of screening programmes and I would agree that this study alone does not present sufficient evidence to support that action. The study does however suggest that claims of increased life expectancy due to screening may be overexaggerated and unsupported by current trial evidence. This may mean that the risk/benefit balance may have shifted somewhat.”
“Overall, this well-designed and large study questions the assumed position that cancer screening will always deliver significant increases in life expectancy. As treatments for various cancers improve, the window of opportunity for screening to have major benefits diminishes, that is, the difference between catching a cancer a bit earlier may not be as extreme as it once was for prognosis. This combined with the possibility of screening detecting small tumours that may never have progressed to a life-threatening malignancy may combine to explain the current findings. The authors do a good job of considering all-cause mortality as the most important metric for measuring success. This allows us to see whether the intervention has an actual impact on the life of the individual rather than just displacing one cause of death with another at a similar time.”
Prof Paul Pharoah, Professor of Cancer Epidemiology, Cedars-Sinai Medical Center, said:
“The authors of this paper have used the results from published meta-analyses of screening for six different cancers to estimate the number of days of life gained from screening. They show that the only screening test for which there was a statistically significant gain in life is screening for bowel cancer by sigmoidoscopy. There were small gains in life for other screening tests, but these were not statistically significant. The common claim that screening saves lives is not justified by the available data.
“There is nothing fundamentally new about their findings: previous publications have provided estimates of the relative risk of all-cause mortality associated with screening. The authors of this paper have simply converted these relative risks into absolute risks.
“It has been described many times before that the reduction in all-cause mortality from different cancer screening tests is very small. However, screening for any one cancer would only be expected to reduce mortality from the cancer being screened for and any all-cause mortality reduction would be small at best. Demonstrating a small reduction in all-cause mortality is very difficult from a statistical standpoint.
“A rather specific criticism of this paper is that the study they used for the all-cause mortality reduction associated with breast screening mammography included several different estimates for different age-groups of the population. They picked the findings for women over 50 (showing no reduction in all-cause mortality – a relative risk of 1), whereas the result for women of any age showed a small 1% reduction in all-cause mortality (a relative risk of 0.99).”
‘Estimated Lifetime Gained With Cancer Screening Tests’ by Michael Bretthauer et al. was published in JAMA Internal Medicine at 16:00 UK time Monday 28 August 2023.
Prof Stephen Duffy: “Although I do not consider it a conflict of interest, for completeness I declare that I am a co-investigator on some of the trials reviewed.”
Prof Paul Pharoah: “I have no conflicts of interest to declare.”
For all other experts, no response to our request for declared interests was received.