NHMRC Public Consultations

Skip Navigation and go to Content
Visit NHMRC website

Review of Chapter 2.3 of the National Statement: Qualifying or waiving conditions for consent submission

Personal Details
First Name: 
Last Name: 
Specific Questions
1. Please comment on the following definition of ‘opt-out’:: 
This definition describes "opt out" as a method of recruitment, rather than a means of waiving the requirement to seek explicit consent. This is misleading. It is, rather, a means of seeking indirect consent or implied consent. Logically, 'opt out' consent is not consent at all, since the researcher has no idea whether a non reply means consent or a failure to respond.
2. Please comment on the rationale provided for an opt-out approach (i.e. Section 3).: 
The rationale needs to be tightened to indicate that it is to be used where there are vast amounts of data and where it is impractical to seek explicit consent. It should also be more explicitly stated that it is to be used mostly only for the certain kinds of low risk medical research. Where it is non-medical research, it should only be available for population studies and large scale studies where there is low risk only. There is a danger that researchers will take the opportunity to use the opt-out means of gaining consent in cases where explicit consent can and should be obtained. For example, knowing that schoolchildren are notoriously bad at returning signed forms, a researcher could argue that an 'opt-out' approach is very likely to allow a better recruitment of participants. This would be very problematic.
3. Please comment on the proposed limited application of an opt-out approach (i.e. Section 4).: 
4.1 needs some further explication. It seems to allow for any kind of research being able to use the 'opt out' approach, when the rationale seems to largely have medical and health research in mind. The notes at the bottom of page 7 confirm this.
4: Please comment on the flow chart (i.e. Section 4).: 
The flow chart is problematic. The right hand branch at 2b begins with data that is collected, no consent is obtained and at 5a it is decided that the value of the data is recognised. This, however, means that HREC is being asked to retrospectively approve data collection. There are a number of cases where researchers might claim that they had collected data, say for quality assurance purposes, but now see the value of it and would like to use it. Unlike research which explicitly seeks ethical approval prior to data collection, in this instance, there is less scrutiny of the ways in which the data has been collected. There are cases where this could be appropriate, such as the use of medical or health information where the research is focussed on establishing relationships between various variables. In these cases, there is already sufficient provision through waiver of consent. What needs to be specified much more clearly the kinds of data where the right hand branch might apply. 6a is of concern, as it suggests that data has been collected without following an HREC process that is of such sensitivity that the proposed research is more than low risk.
5. Please comment on the appropriate mechanism for providing information to participants for the opt-out approach represented at box 6d of the flow chart.: 
HREC approval should also be required where s95 or s95A does not apply. This suggests that the opt out approach can be used with non-HREC approval processes. This could be appropriate only in some cases, but not others.
6. Please comment on the proposed amendments to the National Statement (see Attachment A underlined and in red text).: 
Introduction: The first 5 paragraphs are not connected in any way with the following red additions. It would be better to add a combined 3.2 and 3.3 from the rationale. It might also be worth noting that opt out is not strictly speaking consent (logically it is not), but closer to a waiver of the requirement of explicit consent. 2.3.9 It is unclear why this is here. It suggests that for 'opt out' other levels of ethical review may be used. There does not seem to be a need for this statement. However, as for waiver of consent where health or medical information is being used, it should be the case that only HREC can grant permission to use the opt out approach. This would be consistent. 2.3.10(d)Reasonable need to be defined. Also, it is unclear what happens if all the participants do not received the information. Under the opt out approach, the researchers cannot be sure that all the participants receive the information and so have the opportunity to opt out. In effect, if it cannot be guaranteed that all participants will receive the information, opt-out reduces to waiver of consent. 2.3.10 (f) It is unclear why there would be a follow up phone call if it is opt out. 2.3.10(g)This is a condition under which opt out cannot be used. It should be stated separately, say as a new 2.3.11. This could be stated differently to mean that where identifiable data is used, researchers can only use the opt out means where any information that is published does not identify the participants in any way. 2.3.10(h)surely the more important point here is that the confidentiality of the data will be maintained in accordance with the appropriate level of security, rather than relevant technical standards. 2.3.10(i) Mention might be made of the Australian Code for the Responsible Conduct of Research in relation to data management. 2.3.11 Withdrawal form participation is not feasible once the data has been de-identified. This last clause seems strange. Suggest reword to state that withdrawal is only possible up until the deidentification of the data.
7. Are there situations where an opt-out approach might be appropriate that have not been considered in the proposed amendments?: 
Since there are, in my view, few situations where the opt out approach is appropriate, the answer is "no". It should only be used where the number of participants is very high, perhaps greater than 200, where a high participation rate is required in order to gain meaningful results, where the research is low or negligible risk, where its benefits are significant and where identifiers are removed from any published report or paper.
8. Are there any situations you can think of where the draft amendments would allow an opt-out approach that may be inappropriate?: 
Some types of educational research, and possibly other kinds of research with children. Parents need to be able to provide consent. Any kind of research which does not involve large data sets.
9. Can you provide examples where an opt-out approach may be useful?: 
Use of audit and quality assurance data where the data can be mined to make other kinds of connections.
General Comments

There are significant dangers in the use of the opt out approach, as it could encourage researchers to make use of this approach to avoid seeking consent when it is possible to do so. The flow chart is also problematic, because it seems to suggest that researchers who have collected data for other purposes can now make use of it for research purposes without having had their methods of recruitment of participants scrutinised. It seems to sanction a retrospective ethics approval. Once they have the data and use the opt out approach, it is not known whether participants have received the information or, if they have, they have read it. In effect, there is a reduction to a waiver of consent. It might be more sensible to look at ways in which the cases being considered for opt out approach could be handled under waiver of consent.

Page reviewed: 28 March, 2014