Shaun is Professor of Health Services Research in the Health Services Research Unit at the University of Aberdeen, UK. Shaun joined the Unit in 2013 via a meandering path to health services research. He trained as a physicist and then did a bioengineering PhD involving lasers, blood and bits of skin. Then came health informatics, amputee rehabilitation and a move to Oslo, followed by eight years at the University of Dundee, three of which as Assistant Director of the Tayside Clinical Trials Unit.
Shaun is active in the field of efficient trial design, particularly pragmatic trial design, improved recruitment and retention interventions for trials, improving diversity in trials and the effective presentation of research evidence. He is leading an initiative called Trial Forge (http://www.trialforge.org) that aims to be more systematic about how we generate and use research evidence in making trial design, conduct, analysis and reporting decisions.
He is, or has been, involved in a range of trials, including an EC FP7 trial of treatments for a rare neuromuscular condition, a trial of a lifestyle intervention for women after pregnancy, a lifestyle change trial run through football clubs and the Scottish Premier League, a trial of nerve stimulation to improve continence among residents in care homes, a lifestyle change trial run through the Scottish breast screening program and a lung cancer screening trial. Finally, he coordinated DECIDE, a 5-year, EC-funded project that aimed to improve the way guideline information is communicated to health professionals, patients and the public, policymakers and others.
We had the pleasure of catching up with Shaun to find out about what is happening to make trials better….
Can you briefly introduce your work and the kind of research you do most recently?
I’m a trial methodologist. So I’ve been interested in the methods we use in trials, for around 20 years: How do we make doing trials easier and better?
I’ve never had a particular interest in a particular clinical field or type of trial. It’s always been the methods. A part of that has always been about who is in the trials: who should we have in our trials to make the results as useful as they can be to the people we hope will use the results of our trial.
In the last four or five years, myself and colleagues have become more interested in how to better design trials to involve groups who have traditionally been under-served. Most of what I’ve done personally on this has been on ethnic diversity in trials.
You are currently leading the Trial Forge initiative. Can you tell us a little more about how this?
Trial Forge came about in 2015, shortly after I moved to Aberdeen. It struck me as odd how little we knew about how to make decisions within trials about trials processes. For example, what would be the most effective way to recruit people? Or the most effective way to recruit a group of people with a certain condition and a specific characteristic? We have almost no evidence to support those decisions, or for that matter how to keep people interested in a trial.
The Trial Forge idea really came about because I thought it might be quite cool to have an initiative for this, rather than a project. Projects are always very short term: you hop from one to another. So I thought if I invented something, then it’s an initiative and it lasts much longer.
Myself and colleagues knew we wanted to do something a bit more systematic to fill gaps with regards to process evidence, although the idea was a bit vague at first. Once we started getting into some nitty gritty, we have mostly concentrated on:
- Outcome selection
In the UK, these are the top three methods worries that keep people awake at night. Recruitment comes first, and if you’ve got recruitment nailed, you worry about the next two things.
One of the most important methodologies that we’ve used and promoted within Trial Forge is Studies Within A Trial (SWATs). These are not new; the methodology existed prior to Trial Forge. But SWATs are now a thing amongst the UK trial community. People know what they are, people do them and – very importantly – the National Institute for Health and Care Research funds them. It’s gone from largely an enthusiasts-only interest thing to something which is pretty mainstream.
The next thing we’re trying to see is more coordination nationally. Being completely free – doing absolutely anything – in trial methodology means you’ll find pieces of evidence, but you can’t really draw any firm conclusions from each of those pieces. We need people doing parallel evaluations — testing the same thing in multiple trials. You can then say something with more confidence. That’s the sort of thing we’re trying to do: to chip away at any obstacles in the path of generating evidence.
And the other big thing we’ve worked on is inclusion. More recently we’ve got much more involved in that.
Trial Forge has helped to create different frameworks for improving trial inclusion and diversity…
Trial Forge has led or contributed to three frameworks:
- INCLUDE Ethnicity Framework
- INCLUDE Impaired Capacity to Consent Framework
- INCLUDE Socioeconomic Disadvantage Framework
These are really intended for trial teams when they’re designing their trial, to help them think about who needs to be in their trial.
So the Ethnicity Framework would prompt the whole trial team to think carefully about: for our trial, given the intervention and the disease area, which ethnic groups are important? Who gets the condition? Who would be candidates for this treatment?
The framework would then ask a series of questions about the disease itself, the intervention and the trial. How might the trial design and intervention delivery make it easier or harder for people from those ethnic groups to take part? You can then make changes to your design, while still at the design stage.
The challenge is how to make these changes. Some are straightforward and others are trickier.
What about recruitment and retention with underserved groups?
We currently don’t have a lot of evidence about how to effectively recruit and retain under-served groups. I’m pretty confident that this will become a priority area, because we have a huge gap. With Trial Forge we’re drawing up a priority list of recruitment and retention SWATs for this.
I’ll very quickly also mention something we are setting up called STRIDE: SupporTing Recruitment and retention Improvements for Diverse Ethnicities. It’s to hopefully try and help trial teams tackle problems highlighted after using the INCLUDE Ethnicity Framework. I think from this we’ll have something a bit more concrete about what to do to involve people from different ethnic groups.
Your work has quite a strong Patient and Public Involvement (PPI) focus
In Aberdeen we’ve always been very keen on PPI. And increasingly on really sitting together from the very beginning: in designing the trial.
It’s finding ways to have those discussions, and how to decide what is important to people who have the conditions in question. Listen to what they say, change the design, and then keep talking to people through the trial.
You recently published a paper on PPI in trial outcome selection. Can you summarise this for us?
We wanted to see if there was disagreement in primary outcome measures chosen by researchers compared with clinicians’ and patients’ priorities.
In two clinical areas – treatment and management of breast cancer and kidney-related illnesses – we asked patients and health professionals to rank outcome measures by importance. And we found that 72% of the time they disagreed with what trialists had chosen as the most important outcome.
We then did another study to found out how many of these trials had included PPI, prompted by a question on twitter from a public contributor. We found that across 44 trials, none had involved patient and public contributors in their outcome selection. Zero was a bit of a depressing surprise. But then is it any wonder then that there’s such a mismatch between primary outcome choices made by trial teams and what patients and health professionals consider most important? I think we all need to pause and think about that.
What are the top things you think need to be changed to help make trials more inclusive? You’ve already mentioned engaging more systematically with underserved groups.
1 . I think probably the starting point for everyone would be asking: who should be in the trial?
That sounds trivial and silly, but when we’re drawing up eligibility criteria for trials, characteristics are generally medical things. It would be a big change if eligibility criteria started to include characteristics like socioeconomic status or ethnicity. It would mean systematic recruitment of people from groups who routinely do not appear in trials.
2. Funders also have a big role. But so do regulators and ethics committees. There are multiple points in the research process where people can ask: what are you doing to ensure that the people in your trial represent the full range of people who could potentially benefit?
3. Finally, I think it’s important – for funders, but also for people reviewing grants on behalf of funders – to recognize that we’re not great at this right now. We don’t have strong evidence to support how to do things well.
We haven’t properly engaged with many of these groups for decades. So it’s not easy. Doing this will cost money, and it may well slow trials down. We need to accept that if we really mean what we say about being inclusive, there will be at least a sort of transition period.
Where do hope to see trial diversity in 20 years’ time?
I guess I’d like things to end up where nobody’s really talking about it anymore. It will be obvious that if a particular disease is more severe or is more prevalent in a certain group, we have to do things in our trial to ensure that those people can take part.
And we will have a nice evidence base that suggests which things are effective for involvement, and which organizations can help us do it. We will know how much it costs, and it won’t raise eyebrows.
I think there’s reason to be optimistic, but I don’t think anybody should believe that in two years’ time – or three, five years’ time – everything’s going to be hunky dory. There’s big stuff to be done.
But I think there are positive changes within the UK research trial world. We just need to keep working in the right direction: lots of little changes, which will hopefully add up to something that’s really quite substantive and important. If we do something well within trial methodology, that might improve dozens – possibly lots of hundreds – of trials.
The views expressed are those of the author. Posting of the blog does not signify that the Cancer Prevention Group endorses those views or opinions.
If you would like to join the Cancer Prevention Group’s Patient and Public Involvement group you can find out more here.