[Introduction to Clinical Trials within Cohorts (TwiCs)]
CLAIRE RELTON: My name's Claire Relton.I'm a senior lecturer in clinical trialsat the Pragmatic Clinical Trials Unitat Queen Mary University, London.[Claire Relton, Senior Lecturer in Clinical Trials]My speciality is efficient practical trial design.And I specialize particularly in recruitment and informed
CLAIRE RELTON [continued]: consent--how you recruit patients, and how you provide information,and how you seek consents within the context of trials.Today, I'd like to talk about trials within cohorts,which is a new approach, fairly new-ish approach,to trial design.[How did you become interested in trials within cohorts?]
CLAIRE RELTON [continued]: Well, I first became interested in trialswhen I started a master's in health servicesresearch and then a PhD in trial design.And I was really interested in howto design trials that were efficientand that didn't mask the actual intervention thatwas being examined, and where I was really interested in trying
CLAIRE RELTON [continued]: to design a trial where the experienceof the patients and the clinicians was the same as itwithin routine care.My background is as homeopath.When I started, I got a Department of Health awardto train as a researcher in orderto improve the evidence base for homeopathy and othercomplementary, alternative medicines.At the time at which I started, the front page of the BMJ
CLAIRE RELTON [continued]: said, now is the time for doctors to be boldand tell patients that homeopathy doesn't work.And there I was as a homeopath and a training researcher.And what they were arguing about wasthat there was 160 randomized controlled trials of homeopathyand there was a huge discrepancy in how people interpretedthese trials.One group said, these trials prove that homeopathy works.
CLAIRE RELTON [continued]: And then other groups were saying very loudly,these proves, the trials, that homeopathy doesn't work.So I thought, well, what do these trials mean?So I started unpacking them, like a kid with a Tonka toy,like taking a car apart, and really tryingto understand what these trials were about.So there were a few issues around homeopathy,but then I realized there were lotsof issues around how randomized controlled trials were
CLAIRE RELTON [continued]: designed.Then I learned things like, hey, 90% of trialsactually fail to recruit.I thought, well, that's a lot of waste of government money,isn't it?Why are people keep on doing this design when many of themfail to recruit the populations that they needand they fail to recruit the numbers.And then I had the experience of actually
CLAIRE RELTON [continued]: being inside a trial where everybodywas behaving differently because it was research.And the routine care practices that wewere trying to implement were completelysubverted by the processes that we needed to go throughin order to be in the trial.So I had an experience of being offered to then design a trial.So I designed a standard, pragmatic trial design.
CLAIRE RELTON [continued]: And that was where we were comparing treatment as usualversus treatments usual plus the offerof an additional treatment, treatment of interest.And I had the experience at the first trial managementgroup of the recruiter coming back and saying,bursting into tears, and I said what's going on.And she said, well, I know that half the patients are going
CLAIRE RELTON [continued]: to get usual care, but I know every time I open uptheir little envelope, I always praythat they're going to be getting the experimental treatment,because otherwise why would they come into the trialif they wanted usual care.They needn't come into the trial.They could have got usual care without that.And I thought, well, that's not a very ethical design, is it?So I then worked with my supervisor,
CLAIRE RELTON [continued]: and we had the idea of actually identifying cohorts, i.e.,big groups of people, and then embedding multiple trialswithin the same cohort, and using the routine data that youcollect from the cohort to facilitate the outcomemeasurement for the trials.And then the other important factorwas that actually, we wanted the information
CLAIRE RELTON [continued]: that we gave patients and the consentsthat we sought from patients, to match that whichhappened in routine healthcare.So that's the very swift precis of how I got interestedin this design and how this design was created.So it was the product of my PhD work with my PhD supervisor.[Why should a researcher consider using this approach?]
CLAIRE RELTON [continued]: The researcher would consider usingthis approach if they had a research question whereone of the comparisons they were wanting to look atwas usual care, and if they were wantingto do an efficient trial design, so,if their research was addressing a practical question where
CLAIRE RELTON [continued]: the answers would help inform patientsand clinicians and decision makers,policy makers in routine healthcare.[usual care, efficient trial design][How is this design different from a standard approachand what are its challenges?]There are quite a few challenges in usingthis design in that it's not the standard approach to trialdesign.So in the standard approach to trial design,
CLAIRE RELTON [continued]: you have a research question.You go and recruit your population.You tell them all about what might happen to themor what might not happen to them.You ask for their informed consent for all possibilities.You then randomize.And if it's a trial where usual care is one of the comparisons,50% of your patients then get usual care,which is something that they could
CLAIRE RELTON [continued]: have got outside the trial, so you disappoint them.And I think this approach to providing informed consentis one of the reasons why many standard trial designs failto recruit the populations they need to timely target.So somebody would be interested in using this design
CLAIRE RELTON [continued]: if they are wanting to conduct a trial where treatment as usual,standard care, is one of the comparisons.So the challenges are that the approachis very different in that rather than takingthe single question, recruit your population,tell everybody about everything, and then follow them up,and then at the end of the trial everything'sdisbanded approach, which is the standard approach.
CLAIRE RELTON [continued]: This is a different approach whereyou recruit a cohort of your patients of interestand then you can then facilitate multiple trials,[multiple trials with the same cohort]but you make sure that the information you give,and the consents that you seek, fitwith how you provide information in usual careand how you obtain consents in usual care,
CLAIRE RELTON [continued]: i.e, you don't tell patients about things, treatmentsthey can't then have.And you rarely tell patients, prospectively,that their treatment will be decided by the toss of a coin.You might tell them afterwards, OK, I think you know,this treatment might work.Or you can tell patients, hey, youhave been randomly selected to try this treatment.Do you want to try it?
CLAIRE RELTON [continued]: That's a much more normal conversation.But telling patients who are quite oftenill that their treatment is not goingto be decided by themselves or a clinician,but by the toss of the coin, is quite an unusual situation.But that's the standard approach.So that's one of the tensions of the standard approach.The infrastructures around how you design and implement trialsare set up to fit the standard approach.
CLAIRE RELTON [continued]: So in terms of the ethics applications,in terms of patient information sheets,in terms of the governance, in termsof how you report these trials, they'reall set up for the standard way.So a non-standard way comes into--has frictions and tensions with the standard approachbecause the systems are set up in a way
CLAIRE RELTON [continued]: to suit the standard approach.But that doesn't mean it's not possible to do.So at the moment, we're conductinga review of trials using cohorts,and we've identified 69 studies that, so far, have been usingit for-- actually done primary research,i.e., they've recruited cohorts, and they're
CLAIRE RELTON [continued]: conducting one or more randomized, controlled trialsusing this design.And we're now exploring aspects of those trials.[What are the advantages of using this design?]The advantages of this approach are
CLAIRE RELTON [continued]: that it's a more efficient trial design,particularly if you're going to be conducting multiple trials.The advantage from the clinician's perspective,which I'm often told by clinicians,is that it avoids the necessity of explainingthe experimental intervention to patients who they're notgoing to get it.
CLAIRE RELTON [continued]: So it makes the clinician-patient conversationswith regards to the trial much, much easier,because the conversations happen as theywere doing routine care.This design provides the opportunityto stage the information so that you providethe information in a way that's relevant to the patient
CLAIRE RELTON [continued]: and the clinician at the point at which they'rein with regards to their healthcare.[staged information sharing] So the first consent you askis say, hey, we're doing research in this area,say, chronic knee pain, for example.Do you give permission for us to use your routine datathat we collect here within this clinical settingfor the purposes of research?
CLAIRE RELTON [continued]: That's an easy question.Most patients will consent to that.So 90% or even more will say, yeah, that's fine.It's a nice simple question.You can then ask for those who say yes, they thenjoin your cohort, so you've now got research data over time,about the progression of people with chronic knee pain.
CLAIRE RELTON [continued]: You can see what interventions theyhave using what their quality of lifeis and you can see how their health and their diseasedevelops.So the second consent that you give to them is you say,in the future, can we contact you again,if we've got a treatment that we thinkyou might be eligible for?N