During August-October 2014 it emerged that two protocol amendments were submitted to the ethics committee, and rejected. The first rejection letter is dated 14th February 2013. The objections are uncannily similar to the ones raised by HealthWatch members who did not have the benefit of seeing the requested amendments. It reads as follows (our bold):
The members of the Committee taking part in the review decided that they could not give a favourable ethical opinion of the amendment, for the following reasons:
This amendment seeks to shift the emphasis from a pragmatic opportunity (while resources were being expanded) to compare women who were or were not invited for breast screening in the extended age range, almost a service evaluation, to a randomised trial. The name of the study has changed from evaluation to randomised to reflect this intention.
Ethical concerns
1) Participants in a Control Group who do not know that they are in such a group, as randomisation is done before women are invited for screening:
Women should be randomised after agreeing to be included in a trial of screening in the extended ages. A strong case would have to be made to not arrange the trial in this way.
The original justification for having a group that did not receive screening was that this would have to happen to some women anyway, as on a practical level, there was not the capacity to offer such screening to all in the age extension group. However, as now the researchers are offering more than one additional three yearly scan to those in the screening group, this lack of resources argument loses its validity, as the screening and control groups were going to be selected in batches of 100 within the same geographical area.
The researchers had previously said that any woman could request a mammogram, thus no one who wanted screening would in fact be excluded, but the Substantial Amendment submission states that only those patients in the areas that are taking part in the age extension study can ask for a mammogram every 3 years after the age of 70.
However, the standard screening leaflet says that any woman over 70 can request screening every 3 years, and there is no geographical qualification required – this point also needs clarification.
It may be a reasonable approach for those over 70 years as they will have been aware of the screening program and might have received a screening leaflet, but the younger women at 47 will not be aware that they might be able to ask for an early screening visit, and this is not even mentioned as an option in the standard leaflet either.
The Committee’s concern is that the control group do not know that they are in such a group, that is being used to compare their outcomes with others who are having additional screening. It is not sufficient to say that they can request screening – how would they know as they would not have received any breast screening leaflet in recent years let alone the recent more informative leaflet?
2) There is a lack of equipoise in the approach as to the value of screening in the age extension group:
The standard leaflet is for women aged 50yrs to 70yrs and does reflect current evidence and the view of the Marmot report which concluded that screening does offer benefit.
The researchers have stated that the study is to answer the question regarding the value of screening at the edges of the current screening age range, but still gives these women the standard leaflet, albeit with an extra sheet for those in the age extension study. This extra sheet does not demonstrate equipoise or explain the reason for investigating the pros and cons of screening at the extended age range; it just says that ‘we need more evidence for benefit at the extremes of age’ and there is no mention of the disadvantages.
This additional leaflet needs revising.
3) No consent form has been provided:
Verbal consent or attendance is used for the routine breast screening and might be suitable for this study, if the additional information sheet is expanded. However, the Committee do think that first, the use of randomisation before obtaining consent needs to be resolved and written consent to allow randomisation is strongly preferable.
I regret to inform you that the amendment is therefore not approved. The study should continue in accordance with the documentation previously approved by the Committee.
Undaunted, the trial team had another go and submitted Amendment 2. This was also rejected, on 2nd May 2014 — for not not taking on board what information women need when taking part in an RCT of an intervention that is known to cause harms as well as benefits. The committee raise questions about the competence of the PI in failing to address the previous rejection. The text is as follows:
The members of the Committee taking part in the review decided that they could not give a favourable ethical opinion of the amendment, for the following reasons:
This Substantial Amendment is similar to Amendment 1 that was submitted in February 2013 and given an Unfavourable Opinion. This previous amendment hasn’t been acknowledged in this new Substantial Amendment. Since the revised protocol did not have tracked changes, the Committee found it difficult to ascertain what changes had been made (although this was addressed in the summary of changes, the Committee are not sure if they were reflected fully in the revised protocol) and some of the difficulties raised in the previous response do not appear to have been addressed.
The Committee welcome the set-up of the ethical and trial management groups.
However, the Committee still have concerns about the control group not being aware that they have been enrolled into this study, particularly as now the information sheet does not say that this age extension will be eventually be offered to all women; instead it states that only half the eligible women will be invited. The Committee note that the amendment says that any plans to do so have been pushed back to 2016. This changes the context in which the original application received ethical approval, when it was presented as a pragmatic approach, as facilities were not available to offer additional screening to all the women in the age extension group immediately.
The information sheet needs to provide more information:
It is not entirely clear what determines eligibility as it may depend both on age and where they live.
It mentions benefits and harms but gives no more details and doesn’t spell out that the trial could show that extending screening could result in more harm than good to these participants, as probably most women start from the stance that screening is beneficial.
Telling women that the figures in the routine screening leaflet may not apply to those outside the 50 to 70 age range needs to be expanded to include details of ‘in what way’ those figures might not apply to those in the age extension trial, as the words ‘bear in mind’ provide no tangible information.
There is only a little information about the time scale involved in obtaining the results of this trial, so this point needs clarifying.
A little more detail about the type of ‘other NHS information’ that will be collected is required.
The trial is limited geographically this needs to be clarified clearly; how will women know whether they are in an area that will be scanning women from age 47 or that all women regardless of area who are over the age of 70 can request an additional scan?
The information leaflet will only be sent to those women who are being screened; thus, as previously noted in our previous unfavourable opinion, it is unclear how women aged 47 in the ‘control’ group and not invited for breast screening as part of this trial will be informed that, even though they are in the control group, they may be able to request a scan before 50yr. (Those in the control group over the age of 70 will be aware that they can request to continue to have 3 yearly scans, as they will have received the information as part of their regular screening visits.)
There is no consent form; there is a difference between attending a routine screening visit using implied consent, compared to women acknowledging that that they are taking part in a trial. This is also an important point for those individuals who lack capacity.
I regret to inform you that the amendment is therefore not approved. The study should continue in accordance with the documentation previously approved by the Committee.
It seems that this research ethics committee is rather more assiduous than the one which gave a favourable opinion on the original protocol. We have written to the Health Research Authority (which is being quite helpful) as follows:
Although the two rejections of the substantial amendments that were made 14 months apart were made by only two members of the REC by correspondence, nevertheless I think they identified similar concerns to ours:
1. They identified the seismic shift from a pragmatic evaluation of a presumed (but not yet affordable) ‘good’ to a randomised trial of an uncertain intervention.
2. They identified the lack of truth to participants “there is a lack of equipoise in the approach as to the value of screening in the age extension group” (March 2013) – i.e. in the leaflets that women aged 47-50 and 70-73 get they would understandably think that screening is a benefit.
3. In March 2013 they said that “This extra sheet does not demonstrate equipoise or explain the reason for investigating the pros and cons of screening at the extended age range; it just says that the additional leaflet needed revising “we need more evidence for benefit at the extremes of age” and there is no mention of the disadvantages.”
4. The May 2014 letter identifies that initially there was no TSC or DMEC. This does not bode well to be playing ‘catch-up’ in terms of quality science.
5. This second letter also throws more doubt on the investigators’ competence as they appear to have paid scant attention to the reasons for previous rejection. “This Substantial Amendment is similar to Amendment 1 that was submitted in February 2013 and given an Unfavourable Opinion. This previous amendment hasn’t been acknowledged in this new Substantial Amendment. Since the revised protocol did not have tracked changes, the Committee found it difficult to ascertain what changes had been made (although this was addressed in the summary of changes, the Committee are not sure if they were reflected fully in the revised protocol) and some of the difficulties raised in the previous response do not appear to have been addressed.”
6. The REC says that the documentation “doesn’t spell out that the trial could show that extending screening could result in more harm than good to these participants, as probably most women start from the stance that screening is beneficial.” Quite – extending screening could result in more harm than good!
7. The PI does not appear to know what she is doing when trying to change a pragmatic randomised evaluation into a proper RCT via an ‘amendment’. The PI is now applying to do an RCT (as presumably she and the evidence about screening outside 50-69 age range are in equipoise). The REC say that the leaflet should say that extending screening could result in more harm than good, and they refuse the second amendment.
8. But here’s the problem: what should have happened to the continuing ethical approval of the previous ‘pragmatic’ evaluation (both in March 2013 and May 2014)?
9. Since the Marmot review, we now all appear to agree that we don’t know if breast screening ‘works’ in the extended age ranges. In the light of the new ‘equipoise’ (that led the PI to request an amendment), shouldn’t the REC have gone one stage further and reviewed whether the study they previously thought was ethically permissible was still ethically permissible?
10. In the light of the changed rationale (let alone questions of the competence of the chief investigator, surely the original approval (REC ref 10/H0710/9) should have been withdrawn?
11. As there has been no change in the processes over the last 18 months whilst the pragmatic evaluation continued, a very large number of women have effectively been deceived about the reasons they are being offered screening and the risks and benefits.
12. I would be grateful if the REC could review these comments. I believe that the ethical approval for the age extension evaluation/trial should be halted forthwith and a proper new RCT protocol be invited for submission.
This discussion is now becoming very public, eg at the BMJ. On 3rd August 2014 we asked the head of Public Health England, Duncan Selbie, to put the trial on hold while these issues are examined. A reply was received dated 17th December, which did not address our request and reiterated that PHE was committed to the trial.
As an example of the information given to patients see age extension leaflets for women, see below.