On Friday afternoon (in retrospect a foolish decision – never post anything controversial on twitter just before the weekend) I posted a series of tweets sharing this new project page. The project is a multi-site fully-powered clinical trial of a new manualised intervention for parents of autistic children. Let’s break that down a bit.
- Multi-site = we’re recruiting families from three different locations, but all sites are running an identical study.
- Fully-powered = we are recruiting enough participants that we will be able to confidently state whether the intervention has benefits or not. No hedging our bets with “this approach shows promise” or wishing “if only we had a bigger sample sze”
- Clinical trial = we recruit families and then randomly assign them to one of two different interventions. The first is an established but general introduction to autism for groups of parents. The second is a new but specific training exploring the diagnostic domain of “restricted and repetitive behaviours”, again for groups of parents. This second one is the one being tested, by comparing if against current “best practice”.
- Manualised = importantly, this intervention is enshrined in a detailed manual. This means that we have a clear curriculum for each session. It’s important to make sure all families experience the same thing, but also to eliminate any risk of harm which might come from therapists going “off piste”.
- Intervention = I know the term intervention can ring alarm bells, but I am using it here to mean “a new thing that people wouldn’t otherwise do…” and I suppose I could add “…and that we hope will have benefits”. It doesn’t mean something which is intended to fix, let alone “cure” autism.
- Parents of autistic children = the interventions we are using both involve parents attending a series of groups led by trained therapists. At the groups they will learn more about autism, with therapists following the lesson plan for each week. The parents will learn new vocabulary and concepts to help them understand autism and consequently suport their child more effectively (we hope). In the new intervention we’re testing (as opposed to the established approach we’re comparing it with) this will focus specifically on helping parents understand and respond to restricted and repetitive behaviours that they (the parents) find challenging.
Unsurprisingly, there were some concerns raised on twitter. I’d say these fell into two, closely related, categories. 1. is the intervention likely to help, or could it be harmful? and 2. why has there not been more autistic involvement in the study? I say related, because most people think that if there had been more autistic involvement, the intervention would be more likely to help (and less likely to harm). Most worringly for me, some people pointed out that my involvement in this study seemed at odds with my position as a strong advocate for participatory methods – and even that people might see that this project didn’t have a clear participatory element and therefore think “Oh, if Sue isn’t bothering to involve autistic people, I don’t need to bother either”. The issues were too numerous and complex to address in a series of tweets so I’m going to try to give a bit more detail here, explaining why I stand by me decision to be involved in this trial AND stimultaneously promote a progressive socio-political and scientific agenda.
1. The timeline of research
I powerfully believe that autistic people deserve robust evidence to support service delivery – I spoke about this at a recent conference talk which you can watch here (closing session, about 50 minutes in). I don’t want autistic people to suffer because an untested approach is deployed in services without a proper exploration of the risks and benefits. I don’t want autistic people and their families to muddle through, maybe doing well if they chance upon a supportive and knowledgeable clinician or teacher – doing less well if they’re not so lucky. A good example of where more evidence is needed is on the use of mindfulness techniques. Some autistic people I know find it really helpful and recommend it to others. However I’ve also heard autistic people on twitter saying they found mindfulness not just unhelpful but actively distressing or painful. Large clinical trials are essential to pick apart this kind of apparent contradiction, helping us to determine not just whether an approach is beneficial, but who it is most likely to benefit. If we had that evidence, practitioners could make confident recommendations about who should / should not try mindfulness, and also how to handle it if it doesn’t work out. There’s no greater advocate for the need for robust, fully-powered clinical trials in autism research than the magnificent autistic scholar Michelle Dawson and I’d advise people to follow her for more on that perspective.
The problem is that getting to the big-clinical-trial stage takes time. In the case of our trial, the first seeds of the idea were planted a decade ago. The team have spent those years developing their practice into a manual, seeking funding for a pilot trial with a quaitative arm, running that pilot trial, publishing and sharing that work, developing a new grant proposal, and finally securing funding for the multi-site trial we’re launching soon. This trial is costing more than £1.5m and there’s no way a funder will commit that kind of investment to an approach which hasn’t built up a foundation of theoretical justification, experimental evidence and pilot evaluation. The fact is, there’s virtually nothing – I think actually nothing but I’m scared in case I missed something – that’s had autistic input from the start and is ready to come to trial.
2. Families need support now
Post-diagnostic support for families in the UK is woeful. What little established practice there is tends to focus much more on the social and communication domain (and frequently lacks robust evidence from clinical trials). Of course communication is important. Tons of children come to diagnostic services because parents notice language delays, and it is great if these can be addressed. Learning to speak fuently massively increases a child’s ability to advocate for themselves, to learn, and to thrive. But we know as well – especially in young children – that elaborate and strict routines, and all-consuming interests can cause (or reflect) a huge amount of family stress, and there’s next-to-nothing helping parents understand and respond productively to these. There are a lot of different ways in which this plays out. Parents may try to extinguish or prevent behaviours that children find useful, simply because they are different. Schools may move to exclude children who are seen as disruptive and parents may lack the knowledge and language to advocate for their children in these situations. Families may find themselves living extremely restricted lives as they fear the consequences of disrupting their child’s rigid routine – it’s incredibly hard to work out where there is wiggle-room, and how best to support an autistic child to spread their wings a little, without distressing them. Without any standard provision addressing these problems, parents are left relying on their instincts or seeking expensive, sometimes invasive and often un-proven private therapies.
I think this situation needs to change, and evaluating a promising new intervention is a great way to move practice forward.
3. But is it really “promising” if autistic people haven’t been involved?
Well, I don’t think there’s a simple yes / no answer to that question. First, let’s be clear that this intervention – like the vast majority of approaches that are coming out of the research literature today – is not utterly divorced from community priorities (e.g. item 5 on this list) and experiences. It’s built on a theoretical model and broad range of evidence that includes surveys and interviews with autistic people including children and adolescents. This evidence base has evolved since the first pilot trial, and the team’s thinking has evolved with it – hence, for example, a more refined focus on specific kinds of behaviour that deserve closer attention.
On the other hand, it is certainly true that autistic people were not involved in development of this approach from the very beginning. Parents (as far as I’m aware, non-autistic ones) were closely involved – so the work does have important community input, especially when we consider the fact that this is a parent-group intervention. But still, if we subscirbe to the principle of “nothing about us without us”, clearly ideally autistic people would have been key collaborators too. So how do we move forward from here? I can’t turn back time and change the way this intervention was created. And I’m not prepared to leave families struggling while we wait for something autistic-led to reach the stage where it is trial-ready. I’d also like to emphasise again that I think this approach has real potential for benefit (and minimal risk of harm), despite the fact that it wasn’t co-designed with autistic people.
Moving forward from here
In my opinion, there are important ways to include the community in every stage of the research process. In this specific situation, even though the intervention manual is written and the trial methods are now fixed, here’s a few examples of how I’m trying to do that:
- sharing information on social media: I didn’t need to create a website for this study, or tweet about it. All recruitment will be via clinical referrals, so there’s no reason to promote the study publicly, apart from my desire to understand how the autistic community feels about all of my research. I have learnt a lot – as always – from the types of comments I’ve seen on social media. And this blog post (comments open for 2 weeks below) is another way for people to articulate their concerns. We can use this information to help shape the way we talk about the study, the way we interpret the data, and to develop better interventions and trial designs in the future.
- including autistic people on the trial steering group: I tried to recruit an autistic parent of an autistic child to the Trial Steering Group – this is the committee who oversee the trial while it is running. I haven’t managed it yet but I’m still working on this. As well as helping us to integrate an autistic voice into the study leadership (albeit, I realise, rather tokenistically) ths would also give a community member insights into how clinical trials work which I hope would be useful information for them. For me, this is a key benefit of community partnership – the knowledge flows both ways.
- inspiring a new study on “challenging behaviour”: a key point raised on twitter by Damian Milton concerned the term “challenging” and particularly who gets to define what is challenging. I can’t change the way this word is used in the study materials now, but I am now seeking a Masters project student who’d be interested in exploring this question empirically. For example, if prsented with short stories describing an event involving an autistic child, do autistic and non-autistic adults agree on what counts as “challenging”? If not, what are the differences? And what alternative vocabulary could we use in the future to describe such situations, when someone behaves in a way that is hard to respond to, or maybe dangerous? Watch this space…
- building capcity; assuming my autistic colleagues and friends are willing to accommodate my stance on this trial (and other studies too – my work is far from 100% participatory, I am the first to admit) then I hope in the future they will find a willing collaborator who is not only on board with a progressive research agenda but also has the skills to deliver the research evidence needed to make policy and practice change. I’m continuing to develop my skills as a scientist and taking part in a large multi-site trial is a major milestone. This is going to put me personally, and therefore my research plans generally, in a stronger position to win funding, do some excellent science, and make a difference. There are obviously limits to what I’d take on simply to achieve this “professional development” goal, but it is relevant in terms of building a new and diverse community of practice for autism research, in which I hope to play a role as an effective ally from the scientific community.
In conclusion, I can only apologise if my involvement in this trial seems to undermine my dedication to participatory research. For my part, I feel fully committed to working with autistic people, aligning with their priorities and involving them in powerful positions in the research process. But I also think that merging robust science with this agenda won’t happen overnight and I am comfortable with taking a pragmatic position on a study by study basis. Utimately, I’m a scientist, not a revolutionary.
_________________________________________________
Funding acknowledgement: This project was funded by the NIHR Heath technology Assessment, ref. 16/111/95
Department of Health and Social Care disclaimer: The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health and Social Care.
Thanks Dr Fletcher-Watson for this very clear elaboration of the context in which this study has been planned, you’re right that it does shift the perspective of a reader like myself. Also I’ve now looked at the trial registration, and the articles linked as preparatory studies.
My own comments on twitter were partly misdirected, in that I fixated on subject participation, whereas the subjects in this trial will be the parents of autistic children, not the autistic children themselves. Everyone reading this would probably agree on the importance of having the views and values of autistic people informing all research on autism, but related comment will be less formulaic than “representatives of research subject group should participate in research planning”.
So, I’ll charge through some big broad issues, sorry for lack of nuance, happy to expand or qualify comments. This is just one way of viewing these complex challenges, I don’t claim infallibility, but is the product of a lot of real-life experience, meeting real people, reading, and thinking.
I do assume that this intervention as delivered in the trial will deliver substantial benefit, to both the parents and the autistic kids involved. But without seeing the trial protocol, or manual itself, I base that optimism on prior known facts about the research group, and related comments made by Dr Fletcher-Watson, which indicate a fittingly high level of caution in administering an intervention which is primarily directed toward the behaviour of vulnerable individuals. How generalisable any positive trial findings are in real life will depend on how tightly prescriptive the protocol and manual are. If too tight, the ability of parents to learn good skills from broad-minded clinicians and other parents might be limited, in settings beyond the trial setting, since individual differences in the drivers of behaviours are impossible to cover fully in standardised documents (differences such as developmental abilities, perceptions, emotions, motivation etc.). If too lax, the approach and manual could be wielded as ABA Lite by clinicians so inclined, like so many many interventions currently around with “developmental” in their marketing material, but not in the product.
The central issue underlying the consternation expressed on twitter seems to me to be the big old one in all clinical practice and research which involves young kids and their families. This is the interests of the child as distinct from the interests of their parents, and wider society. Sure, those interests often closely align, but they never align perfectly, and will diverge greatly at times.
History already rightly condemns the “behavioural” interventions championed by Lovaas, whose assumption that the interests of autistic children were identical to those of wider society at that time (conform with contemporary culture, don’t be a nuisance), underpinned the serious harm inflicted on the children he saw, and on many more over subsequent years.
That was the *real* cause of the grave harm, the authoritarian projection of assumed values onto individuals who had their own subjective experience (including suffering), but who literally lacked a voice. The behaviourist approach well suited that benighted project of individual annihilation, but behavioural psychological concepts were not the root evil in play. Behaviourist approaches have their place in all clinical scenarios, just as developmental approaches must also. Parents trying to change the behaviour of an autistic child can be an endeavour which is clearly in the child’s interests, or one which is clearly harmful to the child – just as any parental attempt to change any behaviour of any child can be.
In paediatric medical practice at least, no one has ever come up with a simple, reliable, way to promote the interests of all family members, every time, across all circumstances. Not in thorny questions of life and death (chemotherapy for childhood cancer with a poor prognosis?), not in child protection practice, not anywhere. A starting point however, is for the clinician – or clinical researcher – to try wherever possible to separate out the interests of each individual in the equation, in their own head. This is more, not less, important where interests might (superficially at least) appear to overlap or coincide – who wants every night at the dinner table to be WW3? No one!
Commonly missed ingredients when trying to work out the best interests of a young child, are the child’s subjective experience of the troubling event(s) – as hard as it often is to infer this – and the child’s complex and fluid mix of developmental abilities (including perceptual, including social cognitive). Both tall orders, but both crucial.
The report of the 2015 pilot study, “Managing Repetitive Behaviours in Young Children…” could only raise eyebrows in this regard. The same holds true as for the currently planned study – I bet the interventions delivered in real life provided more unqualified benefit to the autistic kids involved, than a sceptical reader glancing through the paper might surmise.
The language used throughout seems regrettable, it’s not surprising that it has turned some autistics on twitter off the follow-up major study. I haven’t often seen the term “child functioning” in print. Whatever the intention, the phrase hardly sounds humanising. You don’t have to be a Kantian moralist fanatic to wonder whether this sails too close to (sounding like) treating a human as a means to an end – what is the function of a child? Less dramatically, “… RRB can have a disruptive impact on both the child and their family. RRB can interfere with the child’s ability to learn new skills, engage in daily living activities and have been shown to take up large amounts of time.” Which new skills? Whose preferred daily living activities? Whose time?
“RRB can also be stigmatizing, and lead to agitation or aggression if interrupted. Higher levels of RRB are also associated with heightened anxiety, for example, insistence on sameness, may be a maladaptive coping response to anxiety. RRB can interfere with family functioning and have been reported as among the most stressful behaviours for parents to manage in addition to provoking negative parenting styles.” You don’t have to be Michelle Dawson to wonder whether the preceived problem in any of these descriptions should be located in the child.
Better: “It is possible that teaching parents how to consider RRB using a functional analytic framework approach enabled parents of young children with ASD to both consider the communicative function of their child’s RRB and implement tailored strategies to manage them, which in turn may have led to the improved outcomes reported.” Not just possible, almost certain, in my opinion. Note the word “tailored”.
Sure, language like this doesn’t directly harm (or help) any real-life kid, or parent. But I think it clouds the picture, and has unfortunately obscured the potential value and importance of a study like this.
Dr Liam Tjia
Paediatrician, Melbourne
Thanks for sharing this blog post.
One query I have with the study relates to the primary outcome, Clinical Global Impression of Improvement (http://www.isrctn.com/ISRCTN15550611). I would be interested to know more about the rationale for choosing this measure as the primary outcome. Has it been used in previous interventional studies of autistic people, for example? A quick glance of the literature suggests it’s mainly used in drug trials of people with schizophrenia, anxiety, bipolar disorder or depression. Is the measure very well suited for evaluating improvements in autistic people? How are improvements understood in this context? For example, if the intervention being tested leads to a reduction in RRBs then would these changes be regarded as improvements by clinicians? This aspect of the study could also have benefited from involving autistic people. As it stands, there’s a concern that you’re not primarily assessing how effective the intervention is at changing outcomes that are most important to autistic people.
I also recommend making the full study protocol publicly available and, even better, getting it peer reviewed (if it hasn’t already).
Best of luck with the study.
Thank you for writing this out for everyone and engaging with the community. It’s definitely appreciated. I guess my main question is (still) why sensory issues aren’t included in this intervention, or at least talked about here? It seems like the “repetitive behaviors” that are talked about could very much be attributed to regulating sensory environments. Is the intervention going to involve educating parents on sensory differences in autistic people (and more specifically their autistic children)? Or executive function?
Is it going to involve parents learning any nonverbal communication that could help their child – like how to use an AAC device?
I was wondering if there are methods in place for a parent, after learning/understanding more about their autistic child, realizing that they could be autistic themselves? As you know there are commonly undiagnosed parents – is there going to be anything in place for them if they are dealing both with realizing/understanding their child’s needs, but also realizing that they have similar problems as well? It would be interesting, out of a random sample, to see how many autistic parents have sensory sensitivities (but may not realize it), for example.
Also, in the future, do you think the intervention for these parents would look different for non-autistic vs. autistic parents? I’m assuming that everything here is going to be for teaching non-autistic parents (i.e. assumed to be non-autistic as they don’t have a diagnosis?).
Getting stuck in the system of needing funding, and getting autistic voices to be understood as credible in the field (to the people who are reviewing research papers, for example) is hard, I’m sure. I hope this is a good step forward.
One other thing – Is there any funding that could be scrounged up to hire an autistic consultant or even a reviewer for after the study is done, for them to look at and discuss the results and interpretation with you before submitting? That would be a good way to include an autistic researcher. You could even have an autistic researcher collaborate and submit an editorial about the study afterwards. Something to think about.