Investigation into the study 'Early pubertal suppression in a carefully selected group of adolescents with gender identity disorders'

Scope

This report investigates the conduct of the study 'Early pubertal suppression in a carefully selected group of adolescents with gender identity disorders' undertaken at the Tavistock and Portman NHS Foundation Trust and sponsored by the University College London Institute of Child Health. The Health Research Authority (HRA) has no remit to investigate the clinical service provided by the Trust.

This investigation was undertaken in line with the Terms of Reference set out in the appendix.

Background

The study 'Early pubertal suppression in a carefully selected group of adolescents with gender identity disorders' was developed in 2010 in response to broadening use of early pubertal suppression treatments across Europe, in the USA and in Australia. There were debates about the evidence of benefits and safety, and whether young adolescents could consent to treatments due to the potential for unknown risks or harms in the future. Some clinicians argued that the consequences of non-treatment were likely to be greater than those of treatment.

The clinical service of the Gender Identity Development Service (GIDS) at the Tavistock and Portman NHS Foundation Trust had noted that there was considerable dissatisfaction from patients with the situation in the UK, where pubertal suppression was not available, with a number of parents electing to take their children overseas for treatment.

The research team also noted that there were theoretical risks with suppression of puberty but no evidence of harm in published data. The study 'Early pubertal suppression in a carefully selected group of adolescents with gender identity disorders' was established to offer the choice of puberty suppression in the UK, within the context of a research study with safety monitoring.

The application and protocol set out the academic literature on the subject, as well as evidence about the use of the drug to achieve puberty suppression in other populations and conditions. The research team described involvement of young people and families attending the clinical service in designing the study, an activity that is now regarded as good practice but was much rarer in 2010.

The sponsor was University College London (UCL) through the Institute of Child Health.

The study was initially submitted to Central London Research Ethics Committee (REC) 1 (now named London Central). This committee noted that ‘Whilst there was a range of opinions expressed by individual committee members including some who were happy to approve the study, there was a significant number who felt that to proceed on the basis proposed without due consideration of alternative study design was unethical.’

In line with Standard Operating Procedures (SOPs), the applicant had the opportunity to appeal or submit a new application. The study was subsequently reviewed by Central London REC 2 (now named London – Bloomsbury) in December 2010, with justification of the chosen study design.

The minutes of the meeting state:

The Consultant Clinical Psychologist and Consultant Child and Adolescent Psychiatrist attended to discuss the application.

In discussion, the Committee noted the following ethical issues.

1. The Committee commented that this was a well-presented study
2. The Members noted that the study had been rejected by Central London REC 1 committee and this was the second review. The first rejection was based on the opinion that this should be a randomised control trial
3. The Committee noted that early pubertal suppression treatment is currently not available in UK. It is used in Holland and in the USA, though it is provided privately
4. The Committee members enquired whether the researcher could access any literature on studies performed in Holland where the treatment is available
5. The Committee wondered whether the interviews were recorded and quotes were to be used shouldn’t specific consent for this be obtained.

Discussion with the researcher:-

1. The Committee questioned the wording on page 96 of the Young Person Information Sheet, some of the advantages/disadvantages were the same and the committee suggested that clarification should be given with the section re-written on the potential benefits and possible disadvantages of treatment
2. The Committee noted the language in Information sheets needs to be more child-friendly and re-written in a less formal manner
3. The Committee pointed out that the reference to REC 4 needs to be changed to Central London REC 2
4. The research team’s explanation in regards to their feeling that this should not be an RCT was persuasive. These well motivated children would have attending the Gender Disorder Clinic, with their supportive family/carers, after a long period of thinking through the issues. The team felt to then offer a possible placebo intervention was inappropriate and would not be acceptable due to the long-term outcome without early suppression. They felt that some in any placebo cohort would seek alternative therapy once it became clear puberty was proceeding
5. The researchers confirmed that the interviews would be tape recorded and that direct quotes might be used, but that these would be anonymised. The patient information sheets will be amended to reflect this.

The Committee thanked the researchers for attending the interview.

Decision - Provisional opinion with Chair’s action.

The research team responded with the changes requested, and was given a favourable opinion in February 2011, with Chair’s action. The favourable opinion letter referred only to the documents submitted in response to the provisional opinion, i.e. the covering letter with response and the modified Participant Information Sheet (PIS), the remaining documents having been reviewed and approved at the meeting.

The study started recruitment in April 2011. The research team reported that all young people and families were given at least two months between discussion of the study and potential risks and benefits, and signing consent to participate.

Dr Viner provided a recruitment summary in response to a request from the HRA in July 2019:

‘During the recruitment period (April 2011- April 2014), there were 137 young people who were potentially eligible for the study; i.e. aged between 11 and 14.15 years and in the Tavistock service for over six months. Of these, 48 young people (35%) requested to be (sic) attend the UCLH clinics for information about and assessment for the study. Four did not seek further involvement after the first appointment, leaving 44 young people taking part in the study.’

An end of study report was submitted in July 2019 and reviewed by the REC Chair in accordance with Standard Operating Procedures. This was after follow up of the last participant in February 2019, but prior to analysis of the data. It included details of the papers and presentations from the interim analysis at three years. At least one paper was a paper referring to participants in this study along with other patients.

Concerns and complaints

Concerns raised by various parties:

• The study design did not provide a control arm to measure the effect of the treatment
• Recruitment to the study started before ethics approval
• Parents and children taking part in this study were not adequately informed of the risks of taking part
• Annual progress reports were not submitted to the REC for this study
• No interim or final report produced as stated in the original research proposal
• There was evidence that the study’s interim results were unfavourable, but that these had been downplayed
• The study had been cited by the Gender Identity Development Service to justify their policy of extending the unlicensed use of Triptorelin prior to the study completing
• That all of the participants at the interim analysis had progressed to cross-sex hormones and therefore the puberty suppression was not acting as a ‘breathing space’
• The findings from this study have not been formally published, neither as a final report, nor as interim data
• Data for this study disclosed to the Tavistock and Portman NHS Foundation Trust at a board meeting on 23 June 2015 suggested that: natal girls had an increase in behavioural and emotional problems after treatment for one year compared with baseline; a significant increase was found in reported self-harm; no significant changes in gender dysphoria were seen
• Concerns about changes to the clinical service in advance of publication of the results of the study, or at an earlier stage than that used in the study

Complaints against the HRA (relating to the actions of its predecessor National Research Ethics Service) were raised by some parties:

• That Standard Operating Procedures (SOPs) were not followed in allowing the applicant to choose to submit to an alternative REC
• That SOPs were not followed when annual reports were not submitted
• The close institutional affiliation between members of London Central REC 2 and the researchers was not transparent and went unrecorded in the REC minutes and the annual report
• That London Central REC 2’s constitution and attendance did not comply with HRA rules.

Considerations relating to concerns about the research

The study involved the use of a drug, gonadotrophin releasing hormone analogue (GnRHa), to suppress puberty.

In 2010 when the protocol was developed, there was already experience of using this treatment for suppression of puberty in conditions where children begin puberty early.

GnRHa was not licensed for use in addressing dysphoria in gender identity disorders. It should be noted that use of ‘off-licence’ treatments is common in paediatric medicine, as many drugs are not specifically licensed by the manufacturer for children once a licence for adult use has been obtained.

Given that GnRHa was already in use in children under 16 for other conditions and was used in the UK for young people over 16 for gender identity disorder, it was reasonable to seek evidence of the effect and safety of the use of GnRHa in children under 16 presenting with gender identity disorder.

When medical research involves testing new treatments it is self-evident that there is therapeutic uncertainty, i.e. that the effect and safety of the treatment in a particular population and for a particular condition is unknown. It is therefore intrinsic to medical research that consent is given in a context of uncertainty about the outcome. However, evidence from the use of the treatment in other populations or conditions can help both the Research Ethics Committee and participants to make a judgement about the balance of risk against potential benefit of the new treatment in a research study. In this case, the treatment was licensed for the purpose of blocking progression of puberty. However, in this study the effect of the treatment was known and was not under investigation, and it was the longer-term impact of halting the progression of puberty that was being examined. The study was therefore not a Clinical Trial of an Investigational Medicinal Product (CTIMP).

The research team provided a justification for the choice of an uncontrolled longitudinal observational study design when submitting their application for the second REC review, in response to the concerns raised by the first REC. The service could have chosen to implement the treatment ‘off-licence’ in the same way that many paediatric treatments are used. Instead they chose to follow good practice in introducing the treatment in the context of a research study, where evidence could be gathered to inform future clinical service. A randomised controlled trial would only have been ethical if sufficient numbers of patients could be recruited, and retained, to achieve the relevant statistical power to demonstrate an effect. A randomised controlled trial is relevant where participants feel that there is genuine uncertainty between treatments (whether against placebo or an existing alternative). In this study, the treatment was known to suppress puberty and was licensed for this in a specific population. The uncertainty that was to be examined in the study was how puberty suppression affected progression of dysphoria and wider physical well-being in a group. The research team consulted with patients and families and concluded that patients were seeking puberty suppression and would not be willing to accept the uncertainty of not achieving this outcome. Since the puberty suppression treatment was available in other countries, patients would be unlikely to agree to participate in the study or would drop out if randomised to the control arm (as it would be impossible to blind participants to their treatment allocation), and it would therefore be underpowered and hence unethical. Although the research team did not use this argument, it is worth noting that for a randomised clinical trial to have compared the impact on dysphoria of puberty suppression against an alternative of not suppressing puberty or suppressing it at a different stage, the sample would need to have been adolescents who were uncertain about whether or not they wished to suppress puberty.

The inclusion criteria for the study stipulated that participants should have had significant and persistent gender dysphoria and have been seen by the service at the Tavistock clinic for at least six months, with at least four interviews for assessment and therapeutic exploration of their gender identity development. This population was selected specifically because of the likelihood of progressing to cross-sex hormones and potentially to surgical treatment at the age of 16.

The research team described the purpose of pubertal suppression as ‘to induce a sex hormone-neutral environment to provide young people with space to decide whether to progress further with gender reassignment treatment as an adult.’ This phrase appears to have caused confusion as it has been interpreted by some that the puberty suppression was for use in any children presenting to the clinic, that there would be no change in the course of any gender identify dysphoria during this time, and that the child could then choose to progress to cross-sex hormone treatment or to stop treatment with subsequent onset of puberty in the birth gender. It has been noted that the participants in this study and other research involving early puberty suppression have progressed to cross-sex hormones. This has raised concerns that the treatment might be responsible for generating persistence, rather than ‘creating space to decide’.

It would have reduced confusion if the purpose of the treatment had been described as being offered specifically to children demonstrating a strong and persistent gender identity dysphoria at an early stage in puberty, such that the suppression of puberty would allow subsequent cross-sex hormone treatment without the need to surgically reverse or otherwise mask the unwanted physical effects of puberty in the birth gender. The present study was not designed to investigate the implications on persistence or desistence of offering puberty suppression to a wider range of patients, it was limited to a group that had already demonstrated persistence and were actively requesting puberty blockers.

It has been alleged that the protocol was misleading in not providing information from other research about the extent of persistence in patients treated with puberty suppression in early puberty. Although the Dutch study had commenced and was known to the research team, its findings were not published until after the study 'Early pubertal suppression in a carefully selected group of adolescents with gender identity disorders' commenced. The Dutch study was not able or designed to demonstrate causation. Correlation of persistence could be due to the particular patient population. However, as the impact of suppression on persistence was unknown, the participant information mentioned that the treatment could influence the child’s perceived gender identity and how likely they might be to change their mind about their gender.

It is important to note that the treatment was not intended of itself to prevent gender dysphoria. The purpose of puberty suppression was to prevent further development of puberty in the natal gender. If during that period of puberty suppression the dysphoria had abated then the patient could choose to stop treatment, and puberty in the birth gender would then progress. If the dysphoria continued or got worse, then cross-sex hormones could be implemented after the age of 16, in accordance with existing international guidelines. Worsening behavioural and emotional symptoms of dysphoria in the study participants would therefore not in itself be unexpected or warrant a safety report to the REC.

Concerns have been raised about whether all the risks were fully set out to potential participants and their parents. The Participant Information Sheet was reviewed by the REC, who asked for changes to the information on advantages and disadvantages of the study. These changes were made to the satisfaction of the Chair. It is important to recognise that the written information is only part of the consent process, and that verbal discussion is a key part of consent. Neither is consent a one-off process. The HRA is not in a position to comment on the verbal discussions that took place but has been assured by the descriptions provided about the ongoing nature of the discussions with participants and families about both the nature of their condition, the options available and the unknown issues.

The research team did not submit any annual progress reports during the course of the study, although routine requests for annual reports had been made. This is not unusual, as many studies do not provide progress reports, despite routine annual requests. The study team submitted two substantial amendments to the study to modify the eligibility criteria relating to initial bone mineral density. Both were approved. It should be noted that substantial amendments are routine in all studies. The REC did not request any general update on the study when reviewing the amendments.

It is a requirement that researchers report to the REC any unexpected serious adverse events that are deemed to be related to the research procedures. No such events occurred in the study participants. Concerns have been raised about the safety of participants during the study on the basis of interim data. The interim data that has been presented showed that there was an increase in thoughts about self-harm, but that actions relating to harm decreased, although the numbers at this interim stage were small. As the data relate to an interim analysis, the expectation on research teams is that they review whether the data relating to patient safety is sufficiently different to that expected, such that updated information should be provided to participants or additional monitoring arrangements put in place. It is important to note that the inclusion criteria for the study population were that there was a high likelihood of the young person experiencing severe psychological distress consequent on experiencing full pubertal development before the blocker was implemented. The protocol set out the arrangements for structured follow-up to review mental health, as well as full access to the range of interventions available in the GIDS therapeutic programme. The criteria for withdrawing participants from the study due to deterioration of mental health were set out in the protocol. The research team has described its ongoing review of safety information in line with good practice.

The research team has followed an open and transparent approach, and presented interim findings to academics in the field where the nature of interim findings would be understood. It is a normal part of academic debate that questions will be posed, but it will be understood by academics that before the final analysis, any analysis will be liable to change and may not be statistically robust. Where interim review and analysis provides information or identifies concerns then researchers are expected to identify any actions that might be needed, such as providing updated information to participants to allow them the opportunity to review their participation, amending the safety checks, changing the inclusion criteria, or in some cases terminating recruitment. Any such concerns must be reported to the REC and the proposed actions approved. No such concerns appear to have been raised by this study.

The HRA notes that some of the concerns raised relate to the clinical service at the GIDS. The HRA does not have a remit to investigate or comment on clinical services. However, we note that some of the presentations at conferences and the papers presented to the Trust’s Board refer to a mix of both research activity and clinical service. We acknowledge the importance of the interaction between research and clinical service, with each informing the other. It is common practice in areas of rare conditions (particularly in paediatrics) to take an exploratory case-by-case approach using treatments that are not necessarily licensed for that population or condition. Although this activity contributes to clinical experience, it is not conducted with the rigour and robustness of research activity. A key distinction between research and implementing improvements to clinical practice is that research involves specific collection of data about and from the research participants, and this data is analysed in pre-specified ways in order to answer the questions set out in the study protocol. Implementing new treatments is not, in and of itself, research. It is particularly important in such contexts for all parties to be very clear about these distinctions, both in internal discussions and external presentations. It would appear that these distinctions have not always been recognised because there was insufficient clarity in some of the information about descriptions relating to clinical service and those relating to research. For example, references to patients treated as part of the clinical service have been interpreted as referring to the research study, when in fact closer examination reveals this not to have been the case. At least one presentation refers to patients being referred for early puberty suppression, but does not distinguish those referred specifically to the study or those referred clinically who did not meet the study criteria or who started treatment outside the study recruitment period. Similarly, there is a lack of clarity in descriptions of referral for puberty suppression about which part of the pathway this refers to: an initial start of discussions about hormone treatments, or the actual prescribing of treatments. A concern has been raised that children were referred for early puberty suppression without ethics approval. It would be reasonable for a member of the public to assume that references to provision of puberty suppression could only refer to the research study and not to delivery outside a research study as well. It can be difficult to appreciate the distinction between a research project and the introduction of a new treatment beyond research. It should be noted that it is not ethical to continue recruiting patients to a research study when the approved sample size needed for the study has been reached. Similarly, a different population of patients (such as younger children) cannot be added to a research study so that they can access a new treatment, as this would impact on the statistical analysis.

It is legitimate for a clinical service to make changes to its clinical service separately from, and alongside a research study. Changes to clinical service do not require REC approval and should be managed through the clinical governance arrangements of the NHS Trust. These changes can be based on ongoing review of evidence emerging from a wide range of research, as well as international sharing of learning from clinical practice. REC approval is explicitly restricted to review of research and does not extend to scrutiny or oversight of changes to service that are undertaken beyond a research study. The HRA recommends that both clinical and research staff at GIDS are more precise in their descriptions of the different activities they undertake, and exactly what they refer to. Furthermore, they could provide clearer public descriptions of the separate arrangements for governance and oversight of changes to clinical service and how these are separate from but informed by research undertaken in the service and elsewhere.

During the investigation we identified that the version of the young person’s Participant Information Sheet that was used in the study contained the text as approved by the REC but retained the original footer ‘Version 1.0 04-11-2010’. This is not in line with good practice as researchers are recommended to use version control on documents to avoid risks of participants receiving the wrong version. We have no evidence that the wrong version was used in practice. The HRA recommends that the sponsor ensures adequate oversight of version control of documents used in research.

Considerations relating to complaints against the HRA

It has been alleged that the research team was allowed to choose a second REC to submit to after the initial unfavourable opinion from London Central REC 1, and that the research team selected London Central REC 2 because its members were institutionally affiliated with the sponsor of the study, the UCL Institute of Child Health. Furthermore, it has been alleged that the membership of London Central REC 2 was not properly constituted and that the overall attendance of members was not in accordance with HRA rules.

The first REC to review the study, London Central REC 1 raised questions about the consideration of alternative study designs that the Chief Investigator could not answer in the meeting. However, the minutes also note that the meeting considered a controlled trial would be best even if only some patients opted in. This point was not reasonable from a statistical perspective, as insufficient recruitment would mean that the analysis might be flawed. The minutes also show that London Central REC 1 raised some ethical questions in closed discussion that were not noted as being raised with the applicant, but may have affected the members’ views of the proposed treatment in the study. The committee noted evidence that 80% of teenagers with gender identify disorder do not suffer with the disorder on reaching adulthood. The committee does not appear to have noted that the study population was a specific sub-group of teenagers with persistent gender identity disorder who had reached a specific stage of puberty; the 80% figure would not apply to the population that the researchers intended to recruit for the study being considered here. The committee raised concerns about retarding and re-starting natural hormones, and concerns about bullying of teenagers receiving puberty blockers, neither of which were explored with the applicant but may have affected the views of members about the absence of a control arm. It could be argued that the committee extended beyond its remit in questioning the scientific design, given that independent reviews were included in the application. However, it is not clear from the minutes whether the committee had concerns about the scientific reviews submitted as part of the application.

The role of the REC in confirming that adequate scientific review has been undertaken has since been addressed in member training and guidance, given that RECs are not constituted to have scientific expertise in all areas they review.

It was permissible within Standard Operating Procedures at the time for a researcher to choose to submit a study receiving an unfavourable ethics opinion to another REC, rather than appealing to the same REC. An appeal would normally be lodged when an applicant felt that the REC had issued an unfavourable opinion on a matter that the applicant felt had been adequately addressed, i.e. the applicant felt that the REC was making an error. The first REC in this instance did not feel that sufficient justification was provided by the research team for using an observational design instead of a controlled trial. As the applicant had not been able to fully address their points about alternative designs in the meeting, it was reasonable not to appeal. In the second review, the application did not change because the research team’s intention was to focus on providing the justification that had been insufficient at first review. This justification was provided in the covering letter and in discussion at the meeting. This process is entirely legitimate and within due process.

The difference of opinion between the two RECs should not necessarily be regarded as unreasonable. As noted by the HRA’s National Research and Ethics Advisory Panel, ‘individuals, and by extension, individual committees, will value (weigh or judge) different ethical elements differently and can do so in a way that would be considered reasonable. Reasonable disagreement between the members of the committees will translate into reasonable disagreement between committees. The direct consequence of this is that two RECs (that are procedurally consistent) may legitimately reach different opinions about the same research proposal.’

At the time this study was reviewed, it was expected that Chief Investigators would book their studies for review by a local REC within the domain, which in this case would be London. It was also recommended that research involving children was reviewed by a REC flagged for this type of research, which both RECs were. Although applicants can request review at a REC anywhere in the country, it is common for researchers to still request review at the REC based at or nearest to their place of work, as this allows them to easily attend the meeting in person.

It was a requirement that declarations of interest were stated prior to or at the meeting and minuted, along with the decision about how the conflict was handled in the meeting itself. There is no such record in the minutes of this meeting. The Chair at the time has confirmed that reviewing declarations of interest was his routine practice. One member of the committee that reviewed the second application was co-opted from another REC. Co-option is a routine process whereby members of RECs join a meeting of another REC, to ensure that an adequate number and constitution of members are present in line with SOPs. It has been identified subsequently that the co-opted member was a co-author with several other authors on papers with the Chief Investigator, in an unrelated therapy area of adolescent health. It is not clear whether the potential conflict of interest was declared, whether the committee discussed this potential conflict of interests and agreed that it was not a concern, or whether the other members agreed that the individual concerned could contribute but that they would ensure that it did not influence their decision-making. This was not recorded in the minutes, although the co-opted member was not the lead reviewer for the study. Some other members of the REC were affiliated with UCL. However, the minutes do not show any dissenting voices about the acceptance of the justification made by the research team for the longitudinal observational methodology. At the time it was not standard practice for members to declare any interests on an annual basis. This has been in place in annual reports since 2011/12, including listing members’ employers.

A number of members of London REC 2 had connections with University College London or Great Ormond Street Hospital NHS Trust. It should be noted that the research team conducting the study, including the two individuals who attended the London Central REC 2 meeting were employed at the Tavistock and Portman NHS Foundation Trust and not by University College London, which sponsored the study and employed the Chief Investigator.

The HRA’s predecessor organisation provided a quality assurance system for accreditation of RECs. This included a three-yearly cycle of audits, which included aspects such as constitution, training and attendance. Any inadequacies would be addressed through an action plan leading to confirmation of accreditation. If a REC is found not to be able to meet standards, then its accreditation status would be removed. At the time of the review London Central REC 2 was correctly constituted with eight expert and four lay members, of whom three were lay plus (i.e. meeting additional criteria for independence from research). There were some inadequacies in general attendance at the time: four of the 12 members attended fewer than the required seven meetings. Prior to the review of this study, there was a period of time when the committee had too many experts compared with lay members, although two experts then left and the constitution was within the rules. The categorisation of members as expert, lay or lay plus was correct and in line with the guidance. The Vice-Chair was a Chaplain at Great Ormond Street Hospital (the criteria assign Chaplains as lay plus). The other lay members at the time were a retired pharmacologist, a retired solicitor and a special needs teacher who was also a magistrate. In 2010, the oversight of RECs lay with the relevant geographical Strategic Health Authority. It is not known what actions were taken at the time to address attendance. Where such problems with attendance occur now, the HRA takes forward measures alongside the Chair of the REC to ensure that members are reminded of their terms and conditions. Where members are not able to maintain attendance for a period of time due to pressures of work or personal circumstances (noting that members are volunteers), it is possible to request a break in service. London Central REC 2 was audited in 2010 and accredited in April 2011 after completion of an action plan. However, the actual meeting that reviewed this study was quorate and correctly constituted.

It has been suggested that the HRA failed to follow its own procedures in not requiring progress reports, and failing to withdraw its favourable ethical opinion when none were supplied. As noted above, it is common for researchers not to supply annual progress reports. Researchers are advised that RECs have powers to consider suspending or terminating its opinion, but this step is only taken on very rare occasions. It is not an immediate consequence of not submitting an annual report. Suspension or termination of the favourable opinion is used alongside disciplinary measures that may be taken by employers, research sponsors or professional bodies in serious cases of misconduct where the absence of annual reports accompanies other activities where serious concerns have been raised during the study.

In fact, the HRA has for some time been discussing with REC Chairs whether routine annual reports provide sufficient value compared with the burden to researchers of producing them and to RECs of reviewing them, given that other reports and the submission of amendments provide an opportunity for ongoing oversight by the REC within its remit. Whilst a formal decision to cease annual reporting has not yet been made, in practice annual reports are not currently consistently requested. The purpose of progress reports is to provide an opportunity for researchers to receive continued ethical oversight where any ethical considerations have arisen during the course of the study. However, the primary responsibility for oversight of the conduct of a research study sits with the sponsor and it is their responsibility to seek approval for any amendments that might be needed to address any issues arising during the study. The HRA has been consulting about wider transparency measures of which an important one is the publication of end of study reports. There has been no feedback during the consultation to suggest that routine submission of annual progress reports for all studies aids transparency. There was therefore no failing in HRA procedures in not withdrawing a favourable opinion from the study when annual progress reports were not submitted.

Conclusion

The HRA has acted within its Standard Operating Procedures and its normal practice in relation to this study.

The HRA has previously made changes in a number of areas where practice could be more robust than it was in 2010: declarations of interest are set out in the minutes, affiliations of members are set out in annual reports of each REC, policies relating to attendance and training have been revised. RECs have been provided with guidance about their remit in relation to confirming appropriate scientific validity of studies. The HRA is reviewing its arrangements relating to annual progress reports to provide clarity for researchers and the public about expectations.

Overall the research team involved in the design and delivery of the study 'Early pubertal suppression in a carefully selected group of adolescents with gender identity disorders', worked in accordance with recognised practice for health research, and in some areas such as patient involvement and transparency were ahead of normal practice at the time. We find no cause for the study to be re-reviewed by a REC or considered for suspension or termination. We find no cause for concern about the oversight of the study by the sponsor or Chief Investigator, other than a point about document version labeling.

It is important to the HRA that patients and the public have confidence in health research. The HRA wishes to facilitate good research and encourage good scientific debate, informed by open and transparent scientific practice. It does this by ensuring that the ethical issues are properly debated, promoting research transparency, and encouraging good oversight by research sponsors.

The HRA has previously raised questions about the public perception of the implementation of innovation in clinical practice, both in areas where evidence is still developing and in areas where research is not taking place. We recognise that in certain therapy areas, and particularly paediatric medicine, it can be difficult to obtain the funding and resources to undertake research, and that it is therefore common to use unlicensed medicines based on learning from clinical practice. The HRA is keen to ensure that there is transparent oversight and governance of innovative practice that is undertaken outside of research.

It is important to note that the HRA does not have a remit to comment on the interpretation of any of the publications referred to in this study, that is the role of scientific debate. The HRA’s role in relation to the selection of research questions and the design of new studies is to encourage researchers to involve patients and members of the public, and to be assured that researchers have sought independent peer review. The HRA does not fund, design or deliver any health and care research.

The following recommendations are made:

The HRA should complete its review of the use of annual progress reports and provide a clear position on its role in ongoing oversight of research studies.

The HRA should review how effectively declarations of interest by REC members are managed, and seen to be managed.

The Tavistock and Portman NHS Trust should provide greater clarity both internally and externally about the boundaries between research and clinical service.

Researchers and clinical staff working in gender identity development should consider carefully the terms that they use in describing treatments e.g. avoid referring to puberty suppression as providing a ‘breathing space’, to avoid risk of misunderstanding.

UCL should ensure it has mechanisms to oversee version labeling of documents used in research.

NHS England should provide guidance to NHS organisations on appropriate and transparent oversight and governance of innovative practice undertaken outside research.

Appendix 1

Terms of reference for an investigation into concerns about the study IRAS 38588 ‘An evaluation of early pubertal suppression in a carefully selected group of adolescents with Gender Identity Disorder’.

Introduction

These are the terms of reference for an investigation into concerns raised with the HRA about the study IRAS 38588 ‘An evaluation of early pubertal suppression in a carefully selected group of adolescents with Gender Identity Disorder’ (REC references 10/H0718/62 and 10/H0713/79) sponsored by University College London (UCL) and undertaken at the Tavistock and Portman NHS Foundation Trust Gender Identity Development Service (GIDS).

The concerns relate primarily to the clinical service but include reference to the research study. Questions about the conduct of the study and the approvals arrangements by the Health Research Authority (HRA) and its predecessors have been raised.

As these concerns fall within two policies (see Investigation Method below) the HRA has prepared this Terms of Reference to provide clarity on the scope and purpose of this investigation.

Purpose

The purpose of the investigation is to determine whether:

• the application for approval of study IRAS 38588 provided appropriate and accurate information to allow approval decisions to be made;
• the HRA followed appropriate and due process in the review of the application;
• the first Research Ethics Committee (REC) London - Central Research Ethics Committee and the second REC London - Bloomsbury Research Ethics Committee acted appropriately and within remit in the review of the application;
• the sponsor and Chief Investigator provided appropriate oversight of the study;
• the research team at the Tavistock and Portman NHS Foundation Trust followed the protocol and followed responsible research practices in identifying, recruiting and following up participants in the study;

• the failings by any party (if any), and any remedial action required;
• specifically, any failures to protect participants or potential participants, and any remedial action required;
• any opportunities for improvements by any party to better provide assurances to patients and the public, and to encourage public confidence in health and care research.

Scope of the investigation

This investigation will take a broad approach and explore the activities of the sponsor (UCL), the site (Tavistock and Portman NHS Foundation Trust) and the HRA. It will only consider activities relating to research, and not to the clinical service or the implementation of changes to the clinical service.

The investigation will examine the documents available from all the parties that relate to the initial application and the ongoing conduct of the study.

The investigation will be undertaken in the context of the UK Health Departments Standard Operating Procedures for Research Ethics Committees (SOPs) and related processes and expectations that applied at the relevant point in time. It will take account of whether the interests of patients and the public were protected and promoted, but also whether they could be perceived to be protected and promoted.

Investigation method

The investigation will be undertaken in accordance with the HRA policy and procedure for managing third party concerns and issues V1.3 2018.11.29. As it includes elements that are complaints against the HRA’s predecessor it will also be undertaken in accordance with the HRA complaints policy and procedure V4.5 Final 2019.08.02.

The investigation will be undertaken by senior staff in the HRA, supported by a REC Chair and a non-Executive Director of the HRA. None of these individuals were involved in any discussion related to, or handling of, the study at any time prior to questions first being raised with the HRA in 2019.

Appendix 2

Documents used in the preparation of this report

Documents submitted to RECs (selected documents from HRA database and applicant; in line with practice at the time this study was primarily handled in hard copy and not all documents are available on the more recent database).

Application to London Central REC1 ref 10/H0718/62

• REC Application Form signed 14-07-10
• 6/9/10 - Unfavourable opinion

Application to London Central REC2 ref 10/H0713/79

• 3/11/10 - Cover letter
• REC Application Form signed 14-07-10
• Protocol version v1.0 06-10-10
• Young Person Information Sheet v1.0 04-11-10 (footer Version 1.0 04-11-2010)
• Parent Information Sheet v1.0 04-11-10 (footer Version 1.0 04-11-2010)
• Letter from REC1 (6/9/10 unfavourable opinion)

Response to provisional opinion

• 21/2/11 – Letter responding to provisional opinion
• Young Person Information Sheet v1.1 18-01-11 (footer Version 1.0 04-11-2010) (showing changes in bold)
• Young Person Information Sheet v1.1 10-02-11 (footer Version 1.1 10-02-2011)

28/2/11 – Further Information Favourable Opinion from REC

Amendment 29-3-11 revision to parent information sheet

• Parent Information Sheet 22-03-11 (no footer)
• Young Person Information Sheet 22-03-11 (footer Version 1.0 04-11-2010)

Amendment 24-10-11 changing exclusion criteria

Amendment 05-07-12 changing exclusion criteria

• Protocol v1.2 03-07-12

29/4/13 – Progress report request
2/5/14 - Progress report request
19/8/15 – Progress report request 

Minutes

• 25/8/10 – Minutes London Central REC 1
• 1/12/10 - Minutes London Central REC 2

Documents submitted to the HRA

• Summary of recruitment from applicant July 2019
• End of study report from applicant July 2019
• Summary of ethics timeline prepared by applicant August 2019 

Concerns raised by media and members of the public

• 8/4/19 – email from complainant A to previous Chair London Central REC 1
• 11/4/19 - Letter from complainant B to Chair Bloomsbury REC (previously London Central REC 2)
• 8/7/19 - Letter from Newsnight to HRA
• 2/8/19 – email from complainant A to Member Bloomsbury REC (previously London Central REC 2) who chaired meeting where application was reviewed and draft response
• 21/8/19 – Letter from complainant A to HRA
• 2/10/19 – Letter from BMJ to HRA
• BMJ 

Responses to concerns raised

• Response from Tavistock and Portman NHS Foundation Trust to Newsnight July 2019
• Further response from Tavistock and Portman NHS Foundation Trust to Newsnight July 2019
• Responses from Tavistock and Portman NHS Foundation Trust to Mail on Sunday July 2019
• Response from applicant to concerns forwarded by HRA August 2019

Correspondence between RECs and the HRA

• 19/8/19 – response from previous Chair London Central REC 1 to concerns raised
• 11/7/19 – response from Chair Bloomsbury REC (previously London REC 2) following review of end of study report, and responding to concerns raised

Freedom of Information correspondence

• 1819/FOI/050 from B – received 07 January 2019
• 1819/FOI/053 from B – received 30 January 2019
• 1819/FOI/055 from C – received 01 February 2019
• 1819/FOI/070 from A – received 06 March 2019
• 1819/FOI/070 from A – follow up received 06 April 2019
• 1819/FOI/071 from B – received 06 March 2019
• 1819/FOI/071 from B – follow up received 04 April 2019
• 1920/FOI/001 from B – received 30 March 2019
• 1920/FOI/008 from A – received 30 April 2019 

Policy documentation

• Governance Arrangements for Research Ethics Committees (2001) – as in place at relevant time
• Standard Operating Procedures (2011) – as in place at relevant time
• Definition of Expert and Lay Members (2009) – as in place at relevant time
• Confirmation of accreditation of Bloomsbury REC April 2011