Fair play for “researchers”: can editors and regulators develop a common approach to the need (or lack of need) for ethical review?
There are a number of legitimate and valuable tools for gaining information and evidence for scientific advance and improving health care. These include research, evaluation, audit, and others. There is a real danger that UK “researchers” using tools other than “research”, duly following guidance in UK research regulation stating a lack of need for ethical review, find that journal editors will not consider their manuscript for publication because editors dispute the definition of what needs or does not need ethical review.
Many of these issues, of course, also apply to animal ethics approval.
Questions for discussion
- What current guidance do journal editors use to determine if Research Ethics Committee review is mandatory for publication?
- Can we develop a shared process (using current decision tools) that doesn’t unfairly disadvantage “researchers” (and hinder research based care)?
- Recognising that in inevitable grey areas, judgement is needed and will vary, can we develop a process to resolve differences between regulators and editors, that doesn’t burden the researcher to start with?
- Is this a problem beyond the UK and part of what Greg Koski has defined as the creeping “hyper-regulation” of research?
BACKGROUND
The remit of Research Ethics Committees (RECs) is the consideration of research, and the Health Research Authority (HRA) holds firmly to the view that health research should undergo independent ethical review. Audit, service evaluation and public health surveillance, usually of minimal risk, are excluded under official policy (GAfREC 2.3.12 https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/213753/dh_133993.pdf). Their scrutiny lies elsewhere. While this is contended by some, it can be robustly defended and is current UK policy. Hence “researchers” need to categorise their work at an early stage to determine appropriate review. With its partners, the National Research Ethics Service (now part of HRA) collected the literature in 2009 and devised categorisation (http://www.hra.nhs.uk/documents/2013/09/defining-research.pdf) to help and set up a queries line to arbitrate in case of difficulty. In these it was recognised that there could be no clear, unambiguous answer. Judgement is needed and has to be given if the work is to proceed. In 2013 this guidance was further refined and a decision tool developed (http://www.hra-decisiontools.org.uk/research/).
By and large the distinction is clear (if it looks like a duck, then it is a duck) but difficulties in categorisation can arise and there is the risk that, in following guidance, “researchers” can then be disadvantaged if regulators, reviewers and editors either don’t recognise due regulations or don’t share common guidance. It is a concern that has been voiced on several occasions through our queries line. In such a case there is the possibility that REC review is deemed unnecessary within UK regulation at the beginning but later publication is jeopardised if a journal editor disagrees.
Hence the HRA approached COPE and together it was agreed to put this discussion paper on the COPE website for discussion at the COPE Forum, to seek views and, if possible, establish a common approach which would not disadvantage “researchers” (and would expedite health care work).
Comments
As Editor of a biomedical journal, I find this confusing myself. We now have the mandatory header 'Ethical Approval' for all research papers, to be inserted directly before the Method. We accept the option for the authors to write that no ethical approval was needed for the study, plus the arguments. Thus the reader can decide for him/herself.
As Editor guidance I usually check if individual patient or health care provider data have been used; in such cases I think ethical approval is definitely needed. When the researchers use large properly anonymised database then ethical approval, in my view, is not necessary.
Nowadays I notice that our reviewers also increasingly pay attention to the ethical approval issue.
to post comments
Thanks for replying.
You say
"As Editor guidance I usually check if individual patient or health care provider data have been used; in such cases I think ethical approval is definitely needed"
Is this based on definite guidance (legal or ethical) anywhere?
For example would you think ethical review is needed if data are handled or first anonymised by the clinical team?
to post comments
As an Editor in the Biomedical area a simple rule applies; if a journal publishes papers involving human subjects ethical committee approval is required. The Helsinki Declaration requires ethical approval of studies with human participants.
The only exception is reviews which use secondary data.
The aim is to protect patients and human participants.
This an important responsibility for those journals publishing in the biomedical/psychosocial research arena.
to post comments
Thanks Maggie,
You refer to "papers" or "studies". Do you mean research? Some journals may print "audit" or "service evaluations". Do you feel REC review is needed for these?
When you say "secondary data" do you include tissue? For example how might we review use of tissue from a tissue bank?
to post comments
Hugh - the journal I edit doesn't take audit papers. Service evaluation papers that include data collected directly from patients would - for our journal - require REC approval. I use "secondary" data to refer to reviews or epidemiological/national databases.
We also dont publish any papers on use of human tissue - so have no comments to offer.
to post comments
I think that Editors should be aware of the Helsinki declaration and ethical principals and be expected to competantly use their judgement/discretion in applying these principals and not rely solely on a statement of clearance from an external committee. Research that has a committee approval may not be ethical when subsequently submitted for publication. Some guidelines (e.g. in Australia) indicated that de-identified data that is used for clinical audit as part of quality assurance does not require 'clearance' from a duely constituted research ethics committee. Sometimes it is appropriate to publish this - and in the interests of patient care to do so, provided ethical principals are met. Simiarly in case reports provided there is evidence of full and voluntary informed consent. Also, when the work of ethics committees is taken up with a lot of low risk audit and similar there is less time (and expertise) available for assessment of primary research and high risk studies, which itself has risk. So I'm supporting some discretion and flexibility for Editors I think.
to post comments
It is surely time that regulators, editors and "researchers" themselves stopped equating 'ethical review' with 'independent ethical review'. All parties want to see 'good work' and 'good work', whatever it's categorisation, has to be, by any definition, ethical. The onus should therefore be for the "researcher" to demonstrate to editors and readers that the work presented has had ethical consideration and has been conducted to accepted ethical standards. Independent ethical review is (arguably) one way authors can do this but if independent review is not available, for whatever reason, a couple of short sentences in the methodology section will usually demonstrate and give confidence that the work has been conducted taking into account ethical principles (and by this I don't just mean by referencing 'high level' ethics guidelines). If I see a paper containing the sentence 'this work was considered audit and did not require ethical review' then I advise that it be returned to the authors for revision.
to post comments
Although I don't think we make as much use of audit/evaluation data as we might, and I agree that resources can be wasted in unnecessary pursuit of ethical approval, I think the problem here is when data are collected for one purpose e.g. Clinical service improvement within a particular service context, hopefully with appropriate consent, but then submitted for publication, which could be construed as a different purpose for which consent was not given. That different purpose may well be what would make the paper more interesting to a journal, but if the subjects haven't consented to the use of data for that purpose would that not be unethical? I wouldn't include the use of data routinely collected as a consequence of service contact I.e.n"administrative activity data" which would usually (though perhaps not always, in e context of small area geographical analysis) be sufficiently anonymised.
to post comments
Thanks, agree about discretion and RECS bogged down with work of no risk or consequence.
I suppose one of our problems is whether the WMA Declaration really apples to the types of work we are now seeing? What did the original authors have in mind when they first drafted this document? Did they foresee the changing nature of "research"
I agree Editors have a role in deciding if REC review is needed but our reason for starting this was to see how we can prevent researchers being caught in this "gap" not of their own making? Can we develop a procedure where we resolve these issues between ourselves (regulators and editors) rather than disadvantage researchers who have followed advice given at the initiation of their study? (and now find themselves stuck)
to post comments
I have struggled with this issue several times over the past few years as papers have become more varied and creative in approaches to answering questions of interest to clinical practice and health service management. One arena that seems to confuse not only researchers but IRBs (in the US) is audit/evaluation. Some institutions have taken the approach that anything going to publication beyond the institution where the audit was conducted must have an IRB review. Another area of concern is educational research, often really only an audit or evaluation process, where university students are involved as subjects. I expect IRB review of these projects even if data are anonymized because to me, students are a vulnerable population so their participation must be voluntary and they must be guarded from undue pressure to participate in their faculty's project - a project that will progress to publication. As an editor I can't guarantee that - that should be the responsibility of the university's IRB. There is an increasing interest in online research and that is also a big grey area to me. Researchers who mine blog posts for data and quote participant comments need to be really mindful about whether or not participants in the online community see their postings as "confidential" because they used a log-in to access the site. Some of this is common sense, but within large organizations such as associations and publishing houses, the tendency will always be to err on the side of caution. The concerns may be legal rather than ethical, so I'm not sure that a universal guideline or procedure will work.
to post comments
Lack of evidence of ethical approval is a red flag to editors that participants have not been protected or that data have been fabricated or falsified. A few years ago, I asked a non UK institution to investigate a submission where there was no evidence of ethical approval and the authors refused to provide it. The institution found that the study had never been conducted. Lack of ethical approval was one way in which extensive data fabrication by Fuiji and Boldt was discovered. It is unreasonable to expect editors of international journals w to know local law for the various countries (eg UK, US, Australia, Germany, France, India etc) from which they receive manuscripts. Evidence of ethical approval or oversight by a local IRB not the author is reassuring that the study should have been conducted appropriately. Its lack means that otherwise sound papers and research may be rejected as editors err on the side of caution. This text from the BMJ is helpful. http://www.bmj.com/about-bmj/resources-authors/forms-policies-and-checkl...
to post comments
The underlying issue remains that we advise "researchers" before they start that some work falls outside our remit. We do this in an attempt to provide proportionate review and regulation (something researchers and journals have been advocating for some while). If such advice is given and then publication is denied have we unfairly penalised these people who have simply followed rules and guidance?
So can we develop a means of communication (a memorandum of understanding between regulators and journals that will at least allow us to discuss these situations when they arise (probably infrequently)) and avoid penalising the authors?
This discussion might include
(i) a check that the authors DID consult us
(ii) an assesment to see if the article actually reports the study that was presented to us as regulators
(iii) if necessary an assessment as to its classification.
(Falsification and fraud is an issue but wonder if results can be fabricated, so can ethical review (unless this is closely scrutinised)).
to post comments
Thank you for bringing this issue to the forum. I think there are two main considerations for editors when it comes to ethics of the research we publish.
The first is whether we are satisfied the research was conducted ethically. Even a study that has been approved by an ethics committee may raise ethical concerns and may need to be investigated and potentially discussed with the authors' ethics committee or institution.
The second consideration is whether the authors complied with national regulations and law. This can be a minefield as requirements for ethics approval vary from country to country. If authors did not obtain ethics approval because it was not required in the country where the research was carried out (for example in the case of an audit) and there are no ethical concerns, then I would consider a manuscript without ethics approval. In such cases I may ask to see proof that it did not require approval, e.g. national guidelines or a letter from their ethics committee. In rare cases, I have contacted the ethics committee for clarification.
I think that, as editors, we need to balance making sure that the work that we publish was conducted ethically without disadvantaging researchers who have conducted ethical research and followed local/national research regulation stating that their research did not require ethics approval.
to post comments
As a Chief Editor of a European international clinical journal I frequently encounter problems with research having been done on material from one institution/hospital in an ´audit´ fashion or on large databases with just the comment that ethical approval was not needed or that data were anonymized. It may be stated that a data protection authority approved the use of the data, which is good but not the same as ethical consideration. Apart from the UK, Denmark and the Netherlands are examples. Such statements are no guarantee of proper conduct of a study and it is difficult to work with such submissions, partly because there is the risk that such studies may not be cited, even if good. We request reference to a clear official law or regulation granting exemption. The authors word for exemption is not sufficient, - the appropriate committee or IRB has to have said so. Researchers can not decide solely by themselves what is ethical information gathering, neither from patient case notes or large databases. Studies involving biological material require almost everywhere permission, but when it comes to looking at personal details in case records or databases, should that be different? There is a need for international consensus, that without undue restrictions, guarantees correct handing of clinical as well as experimental or biological data gathering, and ensures that databases are used within proper ethical and safety confines.
to post comments
My question is about the type of epidemiological researches OUT of the realm of the CIOM'S International Ethical Guidelines for Biomedical Research Involving Human Subjects. I wonder if a straightforward answer from the authoritative scholars of this forum could help save resources and unnecessary procrastination.
Should at least be warranted that there is no need for any ethical clearance whenever we conduct and so aim to publish an epidemiologic-zero-risk-research {using anonymous, secondary and by no means trackable information to the individual source of data} exclusively from a population-based cancer registry?
Your inputs will be highly appreciated. procrastination could be saved to more than one epidemiologist throughout.
to post comments