In our previous blog, The Ethics Quagmire: Case Studies you might have read the case study by Kathy Tannous concerning the difficulties she has faced getting ethics approval from three ethics committees. But are there problems closer to home for researchers, in their own institutions, even when only one HREC is involved? We think so. But the solution may also lie with these institutions, in the better application of existing guidelines. We explain how below.
Earlier this year, the Productivity Commission handed the Australian Government its report on Data Availability and Use and Research Australia is working with the Taskforce within the Prime Minister’s Department who is developing the Government’s response. A particular area of focus for us, as the national peak body for the medical research pipeline, is the Productivity Commission’s recommendations on Human Research Ethics Committees (HRECs).
HRECs are responsible for providing ethics approval for most publicly funded human research in Australia that involves people; the research can’t proceed without it. This includes research as diverse as a clinical trial, where patients are receiving experimental new medications and treatments, to surveys and research using existing datasets- the types of data based research considered by the Productivity Commission.
The Commission’s recommendations were in response to testimony from researchers about how long delays in securing ethics approval to conduct research with existing datasets were delaying, and in some cases preventing, the research. This is research with the potential to save lives, to make our health system safer and more efficient, and to help prevent disease and injuries. The Productivity Commission’s recommendations in relation to HRECs focused on streamlining the ethics approval process. Specific actions include that all HRECs should be registered with and subject to audit by the National Health and Medical Research Council; and that they should accept approvals provided by other registered HRECs (mutual acceptance).
There is no doubt that ethics approval processes for research using datasets can be improved and the Commission’s recommendations are good but they will take time, perhaps years, to implement, assuming they are accepted by the Government. In the meantime, Research Australia thinks there is more that can be done within the existing rules to improve the process.
Registered HRECs operate under a set of guidelines called the National Statement on Ethical Conduct in Human Research (‘the Statement’). Quite sensibly, this Statement recognises that not all types of research are the same, or pose the same kinds of risks to people. For example, the risk of death or injury associated with undergoing a new surgery technique is of much greater consequence than the inadvertent public release of some of your personal health information from a dataset of Medicare data being investigated by researchers. The latter might be embarrassing but it’s not going to physically injure you. And what about research that uses data where individual identifying information has been removed? What is the risk here and what are the ethical implications?
The Statement provides for these different circumstances by permitting a more streamlined ethics approval process for low risk research, where the worst consequence for a research subject is ‘discomfort’. It also recognises a category of negligible risk, defined as ‘research in which there is no foreseeable risk of harm or discomfort. Where research involves negligible risk and the use of existing collections of data or records that contain only non-identifiable data about human beings, institutions responsible for the research have the option to exempt the research entirely from the ethics process.
So why are researchers using deidentified datasets regularly being asked to subject their research proposals to ethics review? For this type of research, the answer to streamlining research ethics processes appears not to lie in changing HRECs so much as in applying the existing rules properly. Research with datasets where individuals cannot be identified should be routinely exempted from the research ethics process, and those that use identified data and pose a risk of public disclosure should be treated as low risk and subject to the expedited process provided for in the Statement.
Anecdotally at least, it seems that in many cases this is not happening. Research that could be exempted from ethics approval altogether or follow the expedited process is being subjected to the full ethics approval process. While the reasons for this are not entirely clear, we suspect one is misplaced risk aversion on the part of decision makers- why exempt a research application or bring the decision ‘in house’ (streamlining) when it could just be left to the HREC to decide?
Another could be a reluctance to create a separate process for low and negligible risk research proposals when they can just follow the standard process. What this fails to take into account is the extra burden placed on these low risk and negligible risk proposals, and the unnecessary diversion of resources from the consideration of research proposals that pose higher risks and need the full attention of a HREC.
Research Australia plans to undertake more work in this area, to better understand the disincentives to treating low and negligible risk research proposals differently and what can be done to improve existing practices. If you have a view on this please get in contact- we are keen to hear from you.