Screening without evidence of efficacyBMJ 2004; 328 doi: https://doi.org/10.1136/bmj.328.7438.521-a (Published 26 February 2004) Cite this as: BMJ 2004;328:521
All rapid responses
All interventions will produce evidence of harms and benefits in
individuals who receive them. It is essential that systems for full
quantitative and qualitative data collection should be in place,
particularly for interventions in healthy populations.
The `AGE` Trial was designed to randomise 195,000 symptom-less women
aged 40-41 at entry to study the effect of annual mammography, primarily
on mortality. Randomisation ended in 1996 and screening will be completed
by December 2004 when complete mortality data will be available. 53,000
were invited for annual screening for eight or nine years; a control group
of 106,000 were `not invited`: an overall shortfall deemed acceptable by
the Data Monitoring and Ethics Committee (DMEC).  No mention is
made about morbidities (physical, psychological, emotional, social,
financial),  although many thousands of women screened in their
forties, both within and without this trial, and from other trials of
mammography in this age group, have experienced its harmful consequences.
An independent DMEC should take into account reliable quantitative
and qualitative evidence available from all sources, not just those
concerning effects on mortality within the trial itself. Whilst there are
valid arguments in general for not publicising trial data until
completion, screening in the `AGE` trial is almost complete. There is
reliable evidence from other sources with 14-yr. follow-up that the
likelihood of benefit is extremely small or non-existent;  there are
also certain harms. There is also concern about contamination of findings
by opportunistic screening in the control arm of the AGE trial.  We
should all refer again to the Wilson and Jungner principles of screening
 to ponder how unwise and unethical it has been to ignore them.
Advocates of screening frequently insist that we must wait for
deciades until the `true` effect on mortality is apparent after lengthy
follow-up before publishing. In the meantime, millions of women are being
harmed; million of £/$ expended; millions of people are going without
basic healthcare. I repeat: is this ethical? Is it just?
 Sue Moss. The `AGE` Trial: a study of the effect on breast cancer
mortality of annual mammographic screening of women starting at ages 40-
41. NCRI. Booklet: Current National Trials 7th National Breast Cancer
Trials Meeting, 21st November 2003.
 NHS Cancer Screening Programmes: Breast Cancer.
 Sue Moss, Howard Cuckle. Timing of publication of AGE trial
results. bmj.com rapid response 5th March 2004.
 Cornelia J. Baines. Mammographic Screening: Are Women Really
Giving Informed Consent? Journal of the National Cancer Institute. 2003
Vol. 95, No. 20.
 Wilson JMG, Jungner G. Principles and practice of screening for
disease. Public Health Paper. Number 34. Geneva, WHO. 1968
Competing interests: No competing interests
Editor – Hazel Thornton questions why the results of the AGE trial of
breast screening by mammography starting at age 40 have not yet been
published . She is correct to state that the trial began in 1991 .
However, recruitment of centres and women to the trial was phased over a 5
year period in order to accommodate the additional workload within the NHS
Breast Screening Programme. Women in the intervention arm of the trial are
offered screening annually until the year of their 48th birthday.
Consequently, screening in the trial is continuing until the end of 2004,
and follow up for many years beyond this.
The trial was designed to compare breast cancer mortality between the
intervention and control arms at 10 years from each woman’s date of entry.
It is well recognised that in a breast screening trial it will be many
years from entry before any impact on mortality can be expected to become
evident. This is due partly to the small number of deaths in the early
years among cases diagnosed after entry, and partly to the effect of lead
time . In the AGE trial the timing of any such analysis is agreed with
an independent Data Monitoring and Ethics Committee. Publishing mortality
findings too early would result in wide confidence intervals and
potentially misleading conclusions. Thus a mortality analysis must await
accrual of the requisite amount of follow up.
Sue Moss, Associate Director, Cancer Screening Evaluation Unit
Institute of Cancer Research, Cotswold Road, Sutton, Surrey, SM2 5NG
Howard Cuckle, Professor of Reproductive Epidemiology, University of
Leeds Screening Centre,Gemini Park, Sheepscar Way, Leeds LS7 3JB, UK
Competing interests : None declared
1. Thornton H. Screening without evidence of efficacy (Letter). BMJ
2. Moss S,.for the Trial Steering Group. A trial to study the effect
on breast cancer mortality of annual mammographic screening in women
starting at age 40. J.Med.Screening 1999;6:144-8.
3. Nystrom L, Andersson I, Bjurstam N, Frisell J, Nordenskjold B.
Long-term effects of mammography screening: updated overview of the
Swedish randomised trials. Lancet 2002;359:909-19.
Competing interests: No competing interests