Could harmful research misconduct happen in Australia? What steps are followed if an allegation of research misconduct is made in Australia? Is our system sufficiently robust to deal with allegations impartially and justly?
If an effective vaccine for the Covid-19 virus is found, Australia will undoubtedly see renewed activity by anti-vaccination groups. Australians may be unaware that these groups are still bolstered by egregiously false research into the measles-mumps-rubella (MMR) vaccine by UK paediatrician Andrew Wakefield published in the medical journal, the Lancet, in 1998. His publication of falsified findings was not retracted by the Lancet until 2010. In the meanwhile, MMR vaccination rates fell world-wide leading to many preventable deaths. Australians also may not be aware that, although deregistered in the UK for his proven research misconduct, Wakefield moved to the USA where he remains a hero figure for the anti-vax movement.
Could similar harmful research misconduct happen in Australia or does Australia’s system of research governance and its code of conduct for medical and other researchers make this unlikely? What steps are followed if an allegation of research misconduct is made in Australia? Is our system sufficiently robust to deal with allegations impartially and justly? Do we have adequate deterrents in place at a time when the temptations (publications, promotion, money and fame) to cut corners or worse in biomedical research are so numerous?
The key national document that lays out the expected standards of conduct of research is the Australian Code for the Responsible Conduct of Research (the Australian Code). The Australian Code is auspiced by the National Health and Medical Research Council (NHMRC), the Australian Research Council (ARC) and Universities Australia. The Code’s ‘teeth’ relies on the capacity of the NHMRC and ARC to suspend or withdraw funding for research from an institution, a power based on individual deeds of agreement signed when research grants are awarded. The Australian Code lays down a framework for how allegations of research misconduct are to be investigated by institutions (i.e. universities, research institutes and hospitals). An additional layer of scrutiny was added in 2011 when the ARC and NHMRC established the Australian Research Integrity Committee (ARIC). Put simply, ARIC takes the form of an ombudsman, with powers to review the processes that institutions have used in determining allegations of research misconduct. ARIC is not a merit appeals body and does not re-hear any evidence. ARIC is not empowered to review sanctions or disciplinary actions taken by an institution. Its role is to investigate and review complaints about how allegations of research misconduct have been handled by institutions and then advise the CEO of the ARC or NHMRC as to whether the institution has adhered to the provisions of the Australian Code.
A revised version of the Australian Code was released in 2018. In our view the revised version paradoxically weakened and strengthened different parts of the research governance framework. The Code weakened the standards expected and protection offered in two major ways. First, the auspicing agencies chose to do away with the requirement to determine whether a breach of the Code should be designated ‘research misconduct’. The requirement is now limited to determining if there has been a ‘breach of the Code’, which can be seen as a euphemistic term. This change came despite public consultation responses opposing it. We believe that this change has the potential to send a message to researchers that misconduct in research is now less serious a matter than it used to be, thereby weakening any general deterrence that the Code offers. Individual institutions are still able to designate sufficiently serious breaches of the Code as ‘research misconduct’ if they so choose but it is our view that this should occur in all cases that meet the definition as set out in the Code.
Second, the revised Code advises institutions that when selecting a panel of peers to adjudicate serious misconduct allegations, there was no longer a clear requirement that the majority of the panel were to be independent of the institution. We see this second change as further increasing the risk of the already relatively frequent conflicts of interest (some overt and some probably subconscious) that can undermine community confidence in decision making in our institutions. One aspect of the Code that, on the other hand, was strengthened was the clearer guidance now provided to institutions about the initial investigation of allegations of research misconduct.
In Australia there have been calls for a single national agency (akin to the US Office of Research Integrity) to be established to seek to reduce the inherent conflicts of interest for institutions dealing with allegations of research misconduct. Although a valid case that can be made for such an agency, we are not yet convinced. Australia has a unique, pragmatic and relatively inexpensive system for dealing with allegations, a system which can be improved by just a few changes, as we will come to. The addition of the review powers of ARIC to this system only took place nine years ago. We argue that the success or otherwise of ARIC needs to be subject to an external review before any steps towards the establishment of an Office of Research Integrity are seriously contemplated. Such a review should also examine whether the existence of ARIC has been sufficiently publicised and is sufficiently recognised by researchers. The review could usefully examine whether more detailed public reporting of the findings of ARIC’s investigations is desirable.
As former and original members of ARIC, we cannot reveal any detail of matters handled by the committee as institutions participated in reviews on their understanding with the NHMRC and ARC that the deliberations of ARIC would be confidential. What follows is on the public record or has been revealed at public meetings or is of such a general nature that confidentiality is not threatened.
How ARIC works
ARIC’s work is dependent upon receipt of requests for review. Requests generally come from a person dissatisfied with the processes followed by an institution but can be made by other parties (including the NHMRC and ARC). A review will only commence when it is clear that the institution concerned has finalised its actions in the matter or that there has been manifestly excessive delay. Anonymous requests for review may be accepted. Timelines apply but may be modified at the discretion of ARIC. As a four-member committee at the outset, ARIC met predominantly by telephone with infrequent face-to-face meetings in Canberra.
In its review of any matter, ARIC is obliged to observe due process and afford natural justice to all parties. Parties are invited to comment on draft findings (and relevant parts of the analysis) of any review before a report is finalised. ARIC receives secretarial and administrative support from the ARC and the NHMRC. In practice which agency provides the support for a particular review is based on whether the ARC or the NHMRC funded the research funding identified in any allegation. ARIC’s reports are confidential and are provided to the relevant CEO (of the ARC or NHMRC). Each report contains findings and, if needed, recommendations to the CEO, who is free to accept or reject the recommendations.
The workload of ARIC and its nature
The case load of ARIC has not been enormous. From our own recollections the number of reviews requested between 2011 and 2017 was around thirty, of which ten or so were determined to be out of time. These relatively small numbers may in part reflect a lack of publicity for ARIC but are also likely to reflect the general effectiveness of most of the internal investigations and reviews of research integrity complaints conducted by Australia’s institutions.
Nearly all reviews have been requested by the person who initially brought the allegations of research misconduct to the attention of the institution. In most cases these were less senior researchers within the institution but in a few instances, the request came from an independent researcher who had observed problems with published research, or another agency that had an interest in the allegations. The institutions involved to date have predominantly been universities with a small number of reviews involving medical research institutes and hospitals.
Of the first thirteen completed reviews, in eight instances ARIC advised that the institution’s investigation was not conducted in accordance with the Code and with the institution’s policies and procedures. In four instances ARIC found that the investigation was conducted appropriately and in the remaining case, the institution’s approach was deemed acceptable but there was room for improvement. When deficiencies in institutional processes have been identified, there have been a number of recurring themes including:
• Poorly managed conflicts of interest among those conducting the initial assessment of allegations.
• Lack of procedural fairness, particularly in not allowing all parties to present, or comment on, key evidence.
• Failure to adhere to required processes of the Australian Code and those of the institution.
• Institutional processes that were inconsistent with the Australian Code.
• Poor communication with complainants, including failure to provide an adequate description of any findings and the reasons for those findings.
The impact of ARIC
In the absence of an external review or a formal process for seeking feedback from institutions and from those who lodged complaints with ARIC, it is difficult assess the impact of ARIC or the value of its work. Requests for review by ARIC create unplanned work for institutions and considerable effort is usually apparent in their providing material and responses to ARIC. Whether the existence of ARIC influences how institutions now approach the handling of misconduct investigations is unknown. Levels of awareness of ARIC among researchers are also unknown.
The responses of institutions to the role of ARIC and its findings, based on their communications with ARIC, varied. Some institutional leaders appeared to value ARIC’s reports and expressed a willingness to learn from a review while others were more defensive in response to criticisms, especially those of conflicts of interest that had been overlooked or poorly handled.
The future of ARIC
The revised guide to institutions on investigating allegations of research misconduct contained in the 2018 edition of the Australian Code came in to effect in July 2019. As mentioned, in our view the guide now provides clearer and more precise information about handling the assessment of research misconduct allegations; thus theoretically it may be that institutional processes will be improved, resulting in fewer requests for review to ARIC. On the other hand, the revised guide places less emphasis on the use of external panel members for inquiries into allegations of serious research misconduct; thus it may be that where internal panels are used predominantly, more dissatisfaction with outcomes will result, leading to more requests to ARIC for review. The Code now prefers the term ‘breach of the Code’ to ‘research misconduct’, leaving it in the hands of institutions as to whether research misconduct is found for serious breaches; whether this change will impact on referrals to ARIC is unpredictable. More requests may also flow from the inclusion in the revised Australian Code of an obligation on institutions to inform parties to misconduct inquiries of the existence of ARIC.
These nine years of experience of ARIC can be interpreted in different ways. Those who appreciate that universities in general take their responsibility to deal with allegations of research misconduct seriously may be satisfied with the existence and work of ARIC. Those who argue that this responsibility should be taken out of the hands of the universities and given to an independent central agency may wish to focus on the problems that ARIC has identified in its reviews. It is relevant to note that in the USA, which has a central Office of Research Integrity, the initial responsibility for assessing and adjudicating allegations of research misconduct still rests with the institutions. It is also relevant that a recent parliamentary review in the UK led to a decision that this responsibility should remain in the hands of the institutions.
How can the Australian research integrity system be strengthened?
We believe that leaving the assessment and conduct of panel hearings with Australia’s universities and other institutions remains the preferred option, provided that the following changes are made:
• The requirement to make a finding of research misconduct in relevant cases should be restored in the Australian Code.
• Panels conducting hearings into allegations of serious research misconduct should have a majority membership of external members. At least one panel member, preferably the chairperson, should have knowledge and experience in the process requirements for conducting disciplinary tribunals.
• The NHMRC and ARC must do more to ensure that the research community is more aware and informed about ARIC and its work.
• The annual reports of both ARC and NHMRC should contain more detail of the reviews conducted by ARIC and should also provide data on the instances of allegations of research misconduct notified each year to the two funding agencies.
• The working party appointed to conduct the next revision of the Australian Code should include some members who are not employed by the universities and must include a nominee from ARIC.
Until such changes are made and an external review of ARIC is conducted, we argue that establishing a central national office of research integrity is neither necessary nor justified.