Monday, 10 December 2012

Presumed consent may not be ethical in setting up NHS DNA database

A key issue in establishing a national NHS  DNA database, expected to be proposed by the Prime Minister this week, is that of informed consent. It should be understood that such a database of patients' DNA is not immediately clinically relevant. The sample is not being taken for diagnostic or potential treatment of the patient; the proposed database is for genomic research. This research may identify links between specific gene mutations and risk of disease, but it is not intended that this will then be used directly in any proposed treatment of the patient. 

It has rightly become a central principle of medical research that there should be full and informed consent  from any person included in a trial or from whom personal information or samples are obtained. The proposed database, unless it is set up on this basis, would breach that fundamental principle. But one suggestion is that in this instance consent will be presumed unless the patient specifically states that they do not wish their DNA data to be stored as part of the database. This is what is referred to as an 'opt out' approach to consent.  It assumes consent unless otherwise stated. An opt out strategy for the data base would assume informed consent  unless the patient specifically says they do not wish their DNA to be included. 

Ethical research depends on the credibility of consent and how it is obtained.  In this case I am dubious about  the robustness or efficacy of an 'opt out' approach. 'Opt out' is used for consent in a few clinical contexts. It is used for example in HIV testing in pregnancy. Unless the woman specifically indicates that she dosn't want their sample to be tested for HIV, then it will be tested. There are specific justifications for this approach, not least a consideration of duties to the unborn child.  But it is unclear how this could work in a research context; or what the specific justification for it would be. Concern for the time it might take to obtain consent, cost of information literature and processing consent forms are the obvious ones. But I know of no research where such concerns have allowed appropriate informed consent to be discarded. Without diligence a presumption of consent  may be open to abuse or mistakes.  It needs to be clear how patients are to be informed of their right not to participate and in what context and how they would be able to opt out without pressure or prejudice. What information is given  about the possible role of the database and how this is given is also important.  If opt out is simply in the small print then this would be worrying. Failure to indicate 'no' is not the same as 'yes'. 

It is not clear why 'opt out' would be needed other than the administrative ones. If the database is of sufficient value then it should be set up on the basis of consent being given through an  'opt in' approach.   In a clinical context 'opt out' can work where it is clear what the consequences might be for the patient or others involved. This is not clear in the case of research, where outcome is uncertain. 

Developing clear ethical guidelines in medical research has taken decades to establish. The key principles were set out in the declaration of Helsinki in 1964. Central to these principles was that consent  cannot be presumed.  Prior to the development of ethical review in research and the development of these key principles  patients and other subjects could be recruited to trials without proper consent. Proactive consent was central to the Helsinki declaration:

"participation by competent individuals as subjects in medical research must be competent individual may be enrolled in a research study unless he or she freely agrees." 

We should be wary of breaching this fundamental principle. Informed consent is not simply that they say 'yes' to being included, or that their consent is presumed if they say nothing, but that they agree after being informed appropriately of potential outcomes and benefits and of any potential harms to those in the trial or whose samples are being stored or used. 

We may feel that a database offers little or no potential for harm. It is after all just a database. But this depends of a number of factors: how the information will be protected, how it will be used now and in the future, and what will happen to the information as the studies based on it develop. This last point is clearly an unknown.  Without  safeguards it would be relatively easy to decide the data could be used in ways not envisaged at the outset.  If this was done then it would be done without consent of those included in the database, because proactive consent had not been obtained.

Whether such a breach could occur may depend then on the arguments presented. I have no doubt that in the future, persuasive utilitarian arguments will be marshalled in favour of potential breaches if it is believed the benefits were sufficiently great or in the 'public interest'. It will be argued that the benefits would outweigh the potential harms and thus justify a breach of the normal principle.  Furthermore, it is possible that the clamour for disclosure will come on civil liberties grounds, with arguments that individuals have a right to know what information is held about them in relation to the outcome of any research.  This after all would appear reasonable if studies using the database found significant links with particular gene mutations. 

There are potentially great benefits to be gained from establishing a national NHS DNA database. But we need to be vigilant about breaches of the fundamental principle of informed consent. Such a database must have proactive and not presumed consent from patients whose data is stored.  

No comments:

Post a Comment