Skip to main content
Search form
  • 00:15

    ROBERT SIMPSON: Hello there, I'm Dr. Robert Simpson.I'm specialist physician working in rehabilitation medicine.I'm also an assistant professor in physical medicinerehabilitation at the University of Toronto in Canada.My research interests are in complex interventionsin rehabilitation with a specific focus

  • 00:39

    ROBERT SIMPSON [continued]: on mindfulness and yoga.I'm going to talk today about how to use the Medical ResearchCouncil framework for developing and evaluatingcomplex interventions in health care.What I mean by complex interventions?Well, complex interventions are definedas any type of intervention, where

  • 01:00

    ROBERT SIMPSON [continued]: there are multiple active, potentiallyinteractive components that are contributing to the outcomesthat we observe.And that can present challenges for researchersin terms of understanding, which aspects of an interventionare leading to which effects, which

  • 01:21

    ROBERT SIMPSON [continued]: aspects of the intervention are essential to see those effects.And which aspects of the intervention may be redundant.I'm going to talk through a worked example of howI used the MRC framework for evaluatingthe use of mindfulness-based interventionsfor people with multiple sclerosis.Here's a diagram.

  • 01:43

    ROBERT SIMPSON [continued]: The diagram is designed to show, to illustratethat complex interventions are essentiallycontain so-called black box wherewhat leads to the outcomes we see after the intervention isoften unknown.And part of the reason why we conduct this research,

  • 02:04

    ROBERT SIMPSON [continued]: is to gain a better understandingof what's happening, and that allowsus to refine interventions.So that we're only exposing peopleto those parts of the interventionthat are essential.Why focus on complex interventions

  • 02:25

    ROBERT SIMPSON [continued]: in rehabilitation?Well, complex interventions is for pretty much everythingwe do in rehabilitation.Rehabilitation is about minimizing disabilityand optimizing an individual's abilityto function independently.Rehabilitation is a longitudinal process.It takes place over time, and it involves

  • 02:46

    ROBERT SIMPSON [continued]: multiple different stakeholders, not justthe patient and the doctor in termsof the traditional relationship, but itinvolves multiple other health care providers and patientsfamilies, too.So for example, doctors, nurses, alliedhealth professions, psychologists,to name but a few.And rehabilitation is a characteristically focused

  • 03:09

    ROBERT SIMPSON [continued]: on developing patient and family goals thatare what the patient is hoping to achievefrom being in rehabilitation.So there's tremendous variability,and it's a very complex area.So it never boils down to one individual thingthat makes somebody better after disabling illness or injury.

  • 03:31

    ROBERT SIMPSON [continued]: It's a complex process.Why did I become interested in this area?Well, it really from working with patients with a disablingillness and injury.I was struck by how challenging it

  • 03:56

    ROBERT SIMPSON [continued]: was for people to adjust and adapt to newfound impairments and disability.And I was particularly struck by my MS patientsof whom I had many, how often they described how stressfulhaving the condition was?How frequently they had comorbid anxiety and depression?

  • 04:20

    ROBERT SIMPSON [continued]: And that they told me that there wasn't muchthat they could do about it.They didn't want to take antidepressantsbecause it caused side effects and theycouldn't access any other treatments that were effective.So I became interested in the question of,what other interventions could we consider?

  • 04:41

    ROBERT SIMPSON [continued]: The other thing that influenced me was the United Kingdomgovernment had recently released a government documentthat was focused on the axiom, no healthwithout mental health.And essentially what that means is we can't expect peopleto feel better if we don't.

  • 05:02

    ROBERT SIMPSON [continued]: It also address their mental healthwhen addressing physical health conditions.I knew from my team previously working in general practicethat mindfulness-based interventionswere effective for other people whohad these similar types of problems stress, anxiety,and depression.But I didn't know if they would be effective for peoplewith multiple sclerosis.

  • 05:31

    ROBERT SIMPSON [continued]: And thus, I came across the Medical Research Councilframework for developing and evaluatingcomplex interventions.What is the MRC framework?The MRC framework is a guide for researchers.It outlines a series of steps that researchersshould consider and work through when developing a research

  • 05:53

    ROBERT SIMPSON [continued]: idea through to the eventual process of testingin a full scale trial, and moving on beyond thatto implementing a new practice in the clinical realm.On the MRC framework, it starts by suggestingthat researchers should be really clear about the research

  • 06:15

    ROBERT SIMPSON [continued]: question.And what population are you hopingto study an intervention, and whatare the characteristics of that population that makethis intervention necessary.They also recommend that you try and findout what evidence already exists for this intervention, not just

  • 06:36

    ROBERT SIMPSON [continued]: the evidence of effectiveness, but also evidence for safety.And also try to elicit, whether thereis a coherent theoretical basis for the intervention.And this is important because theycan get the types of decisions that we make.It can also influence, whether wethink an intervention is likely to be safe in a given

  • 06:59

    ROBERT SIMPSON [continued]: population, where mechanisms of actionbecome important where there's potential for overlapsynergism or interacting in a more negative way leadingto harm.The MRC framework recommends that developing and evaluatingcomplex interventions is not a linear process,but instead is an iterative process.

  • 07:20

    ROBERT SIMPSON [continued]: And working through these steps in a sequencethat fits the needs of the researchthat you're dealing with.So for example, you might develop an intervention,test the intervention, and then decidethat that version of the interventionisn't right, so you would go back to development again

  • 07:40

    ROBERT SIMPSON [continued]: before moving on to test in a pilot or feasibility study.And the whole purpose of this is to refine the intervention upuntil the point that you think that it's ready to betested in a full scale study.And the Medical Research Council recommendsthat although implementation is often seen

  • 08:02

    ROBERT SIMPSON [continued]: as the final stage of research, whereresearch becomes translated into mainstream clinical practice.It normalized and embedded in routine care systemsthat it's really important to thinkabout this from an early stage.Because if you don't identify potential barriersand facilitators to implementation,then finding these over at the last minute

  • 08:23

    ROBERT SIMPSON [continued]: can be very problematic, where research resourceshave been used up.And it's difficult to go back to square one again.This diagram just illustrates the non-linear iterativeprocess of the MRC framework movingbetween developing, piloting, evaluating, reporting,and implementing.

  • 08:51

    ROBERT SIMPSON [continued]: So I'm going to talk now about how I use the MRC frameworkby talking to a worked example.This is about assessing the use ofmindfulness-based interventions for peoplewith multiple sclerosis.So thinking about those steps that the MRCcomplex interventions framework recommends.First of all, which population am I proposing to study?

  • 09:13

    ROBERT SIMPSON [continued]: People with multiple sclerosis.And what's the issue?Well, my anecdotal feeling and viewwas that stress was a problem.So the issue of stress, anxiety, and depression.And then thinking about the intervention.Which intervention proposed to study?Well, I knew that mindfulness-based interventionswere effective elsewhere.

  • 09:34

    ROBERT SIMPSON [continued]: So I wanted to test that intervention in people with MS.The MRC then recommends what's knownabout the intervention in the population of interest,and that's the next step and we carried outby conducting a systematic review and meta analysis.The next step in the MRC framework is thinking about,is it feasible to deliver this intervention in the context

  • 09:57

    ROBERT SIMPSON [continued]: that we are proposing to study?Even if something is going to be effective in reviews and metaanalysis, we can assume that it will beeffective in a local context.Next we have to think about issues like accessibility,acceptability is no good bringing an interventioninto mainstream practice if it turns out

  • 10:18

    ROBERT SIMPSON [continued]: that people can access it for a variety of reasons.One of which might be physical disability.One of which might be cognitive impairment.Or that they don't find it acceptableas a practice is just basically our antithesis to whatpeople are looking for, or the practices, for example, usea language that isn't acceptable to people with disability.

  • 10:42

    ROBERT SIMPSON [continued]: And then finally implementation thinking about barriersand facilitators.And I'm going to talk through how I used all of these stepsin my research project using the MRC framework.Defining the population is stress prevalent in people

  • 11:04

    ROBERT SIMPSON [continued]: with MS. To answer this question,we wanted to access data that was going to be reliable.What I mean by that is a sample sizethat's large enough that we can make assumptions from.For example, if I see 10 people in my clinicand they all tell me they've got stress,

  • 11:25

    ROBERT SIMPSON [continued]: that doesn't tell me much about MS on a population level.We were fortunate enough to accessa nationally representative cross-sectional primary caredatabase, which had been designedfor studying multi morbidity or the occurrenceof multiple long term conditions and individuals in Scotland.

  • 11:45

    ROBERT SIMPSON [continued]: One of those conditions was multiple sclerosis.Other conditions included anxiety and depression.So it was very relevant for our study.We used a statistical method called logistic regression,and that allowed us to compare peoplewith MS against controls, and explore the impact,

  • 12:07

    ROBERT SIMPSON [continued]: explore the prevalence of anxiety and depressionin people with MS versus control versus controls.That allowed us to assess for the prevalence of anxietyand depression in people with multiple sclerosisversus controls, while also beingable to control for additional variables, which

  • 12:27

    ROBERT SIMPSON [continued]: might influence outcomes.For example, age, sex, and deprivation.What we found was that any type of comorbiditywas more common in people with MS versus controls.With an odds ratio of two, we foundthat having an additional physical health condition

  • 12:49

    ROBERT SIMPSON [continued]: was also common.Twice as common in people with MS versus controls.What was particularly notable was that mental healthcomorbidity anxiety and depression in particularwas three times as common in people with MS versus controls.We also found that as the number of additional physicalconditions increased so that the prevalence of mental health

  • 13:10

    ROBERT SIMPSON [continued]: impairment.So first of all here, we have a forest plot illustratingwhat I've just been describing.And that shows up at the top.On the right-hand side, we've got the forest plot.On the left-hand side, we've got the tabledescribing the mental health conditions.And then on the right-hand side, we call it the forest plot.

  • 13:32

    ROBERT SIMPSON [continued]: The line down the middle of the forest plotillustrates where there is no effect or no differencebetween the populations.So what we can see for the top to their anxiety and depressionare significantly more prevalent in peoplewith multiple sclerosis versus controls.There are another two below that, drugs misuse

  • 13:56

    ROBERT SIMPSON [continued]: and eating disorder.They also fall within the statistically significantcategory.However, the small sample size makes these likely aspurious finding.This graph here illustrates that as the numberof additional physical conditions increases,so too does the prevalence of mental health impairment.

  • 14:18

    ROBERT SIMPSON [continued]: The top two lines represent peoplewith MS, split into females on the top males below.And then the bottom two lines represent the controlpopulation females on the top, males below.And what the graph serves to illustrateis that this linear relationship that we observeas the number of physical health conditions increases so too

  • 14:41

    ROBERT SIMPSON [continued]: does the prevalence of mental health impairment.You can see that the linear pattern is the same.It affects both people who don't haveMS and people who do have MS. It's only the MS population.This is a far more striking and a highly and a more prevalentfinding.So people with MS are much more likely to be affected

  • 15:02

    ROBERT SIMPSON [continued]: by mental health comorbidity.Moving on to the next step of the MRC framework.This is about defining the intervention.And this is about what are mindfulness-based

  • 15:23

    ROBERT SIMPSON [continued]: interventions?Mindfulness-based interventions originallyderived from ancient oriental Buddhist and yogic meditationtechniques.And a traditional definition mightbe Satipatthana, which means a clear comprehensionthrough awareness.Contemporary definitions of mindfulness

  • 15:43

    ROBERT SIMPSON [continued]: relate more to mindfulness-based interventions.And the original mindfulness-based interventionis mindfulness-based stress reduction.This is an intervention that was developedin the 1970s in North America for people with chronic painand stress.Mindfulness-based interventions have since

  • 16:05

    ROBERT SIMPSON [continued]: been extensively studied with derivativesof mindfulness-based stress reduction,including mindfulness-based cognitive therapy.Both MBSR and mindfulness-based cognitive therapy or MBCThave high quality evidence for effectivenessin people with stress, anxiety, depression.

  • 16:28

    ROBERT SIMPSON [continued]: What leads to the change or what arethe mechanisms of action for mindfulness-basedinterventions?This is not fully understood.What evidence we do have suggeststhat the mediators effecting change are increasesin mindfulness, improvements in cognitive and emotionalreactivity.And how much people practice mindfulness at home

  • 16:50

    ROBERT SIMPSON [continued]: or put another way the dose.Thinking back to definition of complex interventions,are mindfulness-based interventionscomplex interventions.Remembering that a complex interventionis any type of intervention that has

  • 17:11

    ROBERT SIMPSON [continued]: multiple active, potentially interactive components.Well, mindfulness-based interventionsare group programs.Their instructor led.They're delivered over eight weeks with a 2 and 1/2 hourclass each week.They come with a daily commitmentof 45 minutes of home practice.They have core meditation techniques,

  • 17:32

    ROBERT SIMPSON [continued]: which include awareness of the breath, the body and movement.They includes psychological education around stress.They include group discussion so-called inquiry.So there are multiple potential active componentshere to consider, and therefore, yes,mindfulness-based interventions are complex interventions.

  • 18:04

    ROBERT SIMPSON [continued]: What is known about the use of mindfulness-based interventionsfor people with multiple sclerosis?The MRC complex interventions frameworkrecommends that in order to answer this type of questionresearchers should seek out an existing high qualitysystematic review, if such a study exists.If not, then the recommendation isthat you carry out this type of study yourself.

  • 18:26

    ROBERT SIMPSON [continued]: The MRC framework doesn't tell youhow to do the systematic review, but it does point youin the direction of the necessary tools that willguide you through this process.The preferred reporting items for systematic reviews and metaanalysis outlines the necessary stepsfor completing a systematic review.A systematic reviews of research method

  • 18:47

    ROBERT SIMPSON [continued]: that allows you to search a large body of literatureand extract the most relevant types of studiesfor your research question.For example, focused on mindfulness-based interventionsfor people with multiple sclerosis.This means developing search termsthat will capture this type of study.

  • 19:07

    ROBERT SIMPSON [continued]: It means developing a data extractiontool that will allow you to extract the type of datathat you're looking for relating to study type, population,intervention and outcomes.The PRISMA guidance also helps youbring together the findings of your systematic reviewand report that.

  • 19:27

    ROBERT SIMPSON [continued]: And that could either be in a narrative formatif there is not much in the way of quantitative evidenceor if there's more quantitative evidence in the studiesthat you're finding, and are sufficientlyhomogeneous in terms of the populations the interventionsand the outcomes, then it may be possible to carrya meta-analysis.

  • 19:49

    ROBERT SIMPSON [continued]: A meta-analysis is a statistical techniquethat allows you to pool findings from different studies.Therefore, increasing your sample sizeand reducing pooled effects such as a standardizedmean difference.This allows increased confidence in the findingsthat you uncover.

  • 20:09

    ROBERT SIMPSON [continued]: We carried out a series of systematic reviewsand meta-analysis.And this is in keeping with the MRC framework guidance.Research goes out of date very quickly.The adage is that as soon as something becomespublished it's out of date.Initially in 2014, we found that there was very few studiesin this area three in total.

  • 20:31

    ROBERT SIMPSON [continued]: Only one of which was a randomized controlled trial.We updated this in 2019.And by that point, there had been a significant increasewith 12 randomized controlled trials.These had been carried out in various locationsaround the world, including North America, Europe,the Middle East.They had included over 700 people

  • 20:53

    ROBERT SIMPSON [continued]: with MS of all different disease phenotypesacross a range of ages, to mostly includingyounger individuals.And mostly including females.A variety of different interventionshave been used, including MBSR, MBCT.And these have been delivered face to face,

  • 21:13

    ROBERT SIMPSON [continued]: in person, or online.In terms of outcomes there was no sufficient detailto carry a meta-analysis.And our meta-analysis demonstratedthat mindfulness-based interventionshave small to moderate treatment effects, whichshow that they improve stress, anxiety, depressionand fatigue, and people with MS. It wasn't

  • 21:36

    ROBERT SIMPSON [continued]: possible to see from the systematic reviewin meta-analysis, which type of mindfulness-based interventionis optimal for people with multiple sclerosis.We also noted high levels of attritionfrom studies up to 40%.And adherence to treatment was often around about 60%.So less than ideal.

  • 22:02

    ROBERT SIMPSON [continued]: Our systematic review and meta-analysisdemonstrated that although mindfulness-based interventionsare effective for managing stress, anxiety,and depression, and people with MS,there were high levels of attrition up to 40%.And relatively low levels of treatment adherence around 60%.

  • 22:22

    ROBERT SIMPSON [continued]: Bearing in mind that practicing mindfulness,or the increased dose of mindfulnessby practicing at home is associatedwith improved effects.It's really important to try and understandwhy people are not adhering to treatment.And partly, this relates to how feasible is this interventionin the population?

  • 22:43

    ROBERT SIMPSON [continued]: The MRC framework recommends that researchersconsider feasibility at this stage in the research process,and test that ideally in a scale modelbefore moving to a full scale trial.We did this by cutting a feasibilityrandomized controlled trial.Testing mindfulness-based stress reductionagainst scale as usual in a group

  • 23:03

    ROBERT SIMPSON [continued]: of people with MS in Scotland.25 people were randomized to receivemindfulness-based stress reduction,and 25 people were randomized her weight loss control,eventually being able to access the intervention.The feasibility of randomized control trialallowed us to test trial procedures for example,

  • 23:23

    ROBERT SIMPSON [continued]: where are people interested?Where are we able to recruit to target?Where we are able to keep people in the study?Where we are able to collect outcome measuresfrom participants?Would people adhere to the treatmentby attending sessions and then completing the home practices?What we found in our feasibility randomized

  • 23:45

    ROBERT SIMPSON [continued]: controlled trial is that there was a great level of interest.We actually recruited to 100% of our targetwell within our three-month recruitment window.We also found that we were able to collect measuresfrom participants and they completed themto a level of 90% post intervention, and 80%three months later.

  • 24:07

    ROBERT SIMPSON [continued]: In terms of adherence to treatment,we saw 60% attendance at mindfulness sessions,and we saw a home practice level of 66%.These are comparable to the findingsthat we saw in our systematic review and meta-analysis.High levels of aggression and low levels of adherence matter.These matter because if people drop out of a study,

  • 24:28

    ROBERT SIMPSON [continued]: then a study can lose the statistical powerto detect meaningful effects.If people don't complete the practices,then interventions are less likely to be effective.One of the strengths of the MRC complex interventions framework

  • 24:50

    ROBERT SIMPSON [continued]: is that it uses mixed methods.And it allows researchers the opportunity to bring togetherquantitative and qualitative research.Therefore, the qualitative research in this studywas nested, meaning that we carried out the interviewsat a point in the research processwhere it was going to deliver maximum information thatwould be of use to us, in terms of optimizing the intervention.

  • 25:13

    ROBERT SIMPSON [continued]: By using qualitative research methods,we were allowed to explore in more detail, why people signedup to take part in the study.What their expectations were of a mindfulness-basedintervention or aspects of the interventionthey found to be beneficial?What aspects they found were not beneficial?What encouraged them to continue with the practice?

  • 25:36

    ROBERT SIMPSON [continued]: And what factors led to people leaving,dropping out of the course?There are many different approachesto qualitative research.We used interviews, which meant that we interviewed individualswith a semi-structured guide.The questions were focused on our area of interest,for example, expectations, what what,

  • 25:59

    ROBERT SIMPSON [continued]: what didn't, what would you change?These types of things.We interviewed 43 people who took part in the course.We interviewed the people delivering the course,the MBSR instructors.And we interviewed clinicians working with patientswith multiple sclerosis.Bearing in mind that in the longer term, if theseinterventions ever to be implemented,

  • 26:20

    ROBERT SIMPSON [continued]: their input is also going to be crucial.We use the framework approach to thematic analysisin order to organize, analyze and describe our findings.Thematic analysis as a qualitative descriptiveapproach, meaning that no attemptis made to generate new theory, instead you simply

  • 26:41

    ROBERT SIMPSON [continued]: summarizing findings.Here's a table outlining the steps to work throughand cutting a thematic analysis using the framework approach.I'm not going to go into this in great detail.Essentially, there are three overarching steps.These are, how you manage your data?The development of descriptive accounts.And then finally, the development of explanatory

  • 27:03

    ROBERT SIMPSON [continued]: accounts.The final stage of the MRC complex interventionis framework.It refers to considering implementation.Specifically barriers and facilitators to implementation.Implementation itself is a complex process.

  • 27:25

    ROBERT SIMPSON [continued]: It can be used to refer to how a new intervention becomesembedded and normalized within a normal context,or context was the UK National Health Service.We wanted to explore barriers and facilitatorsto implementation.And again, the best way to do thisis using qualitative data where you can ask people specificallyabout the area of interest.

  • 27:47

    ROBERT SIMPSON [continued]: We used an implementation tool known as normalization processto analyze our data.Normalization process theory is a toolthat's been developed specificallyfor studying the implementation of complex interventionsin health care.There are four key domains in normalization process theory.These are coherence, how stakeholders

  • 28:10

    ROBERT SIMPSON [continued]: make sense of the new intervention,cognitive participation, which is about stakeholders abilityand willingness to participate in a new intervention.Collective action, which is about the practical steps thatstakeholders must undertake in order to enact an intervention.And reflexive monitoring, which is about,how stakeholders monitor and appraise an intervention,

  • 28:34

    ROBERT SIMPSON [continued]: as it continues in practice.In terms of challenges associated with cutting outthis research project, I didn't meet any unexpected challenges.But I was new to research when I started this process.

  • 28:55

    ROBERT SIMPSON [continued]: I undertook this process as part of a PhD project,and as such I had PhD supervisors.They recommended that I use the MRC complex interventionsframework.And I have to admit that when I first read it,I didn't find the examples that were provided intuitive at all.I didn't really understand the process when I first read it.Through repeated readings of the framework and discussion

  • 29:18

    ROBERT SIMPSON [continued]: with my supervisors, over time I becameclear about what was happening with the complex interventionsframework.Why we were proceeding in the order that we were,and why we were cutting out all these seemingly diverse?And distinct research projects, and how does theyfit together overall?

  • 29:43

    ROBERT SIMPSON [continued]: First of all, I would say, it's really importantto have access to senior academic guidance.I suggest that you seek out a senior academic whohas similar interests to you.Somebody that you feel you can get along with.Somebody you can feel we can ask questions of,as and when the need arises.

  • 30:04

    ROBERT SIMPSON [continued]: Research like this is not somethingthat you do on your own.Research like this involves lots of other peopleat different stages along the way.For example, you might want to access a statistician if you'recarrying a logistic regression.You might want to access a methods expert if you'replanning a thematic analysis or using a different type

  • 30:25

    ROBERT SIMPSON [continued]: of qualitative research.You might want to access a content expert.For example, somebody who's expertin the use of mindfulness or whatever interventionyou choose to evaluate.I suggest that you don't take on too much.For me, I was carrying a PhD.That's a five-year project.If you're carrying a master's MSC project,

  • 30:50

    ROBERT SIMPSON [continued]: I would suggest that you take on one aspect of the frameworkand focus on that and try and do thatwell, for example, a systematic review.I'd also recommend that you learnabout the concept of reflexivity,and try to bring that into your practice as a researcher.Reflexivity is a bit like mindfulness.It's about observing yourself as a researcher

  • 31:13

    ROBERT SIMPSON [continued]: and exploring your own unconscious biases,and how that impacts on your interpretation of dataand how you report your data.Finally, I'd recommend be patient with yourself.This process didn't make sense to me initially,but it does now.And not simply through repeated readings,practice, and putting into action,

  • 31:36

    ROBERT SIMPSON [continued]: puts theory, trying it as a researcher.It really makes a difference.In conclusion, the MRC complex intervention frameworkis a useful guide for researchers.Interested in cutting edge research in this area.By using mixed methods, researcherscan approach complex research problemsfrom a variety of different and complementary perspectives.

  • 31:58

    ROBERT SIMPSON [continued]: They must see framework as a useful general guide,but it doesn't tell you how to do all of this.There are various numerous useful resourcesthat you can use, and have provided linksat the end of this talk.I hope you find this beneficial, and I wish yougood luck with your research.Thank you very much.


Dr. Robert Simpson, Assistant Professor at the University of Toronto, discusses his research on multiple sclerosis and mindfulness in the development and evaluation of a complex intervention in rehabilitation using the MRC framework, including defining a population, an intervention, and using mixed methods.

Looks like you do not have access to this content.

Developing and Evaluating a Complex Intervention in Rehabilitation Using the MRC Framework

Dr. Robert Simpson, Assistant Professor at the University of Toronto, discusses his research on multiple sclerosis and mindfulness in the development and evaluation of a complex intervention in rehabilitation using the MRC framework, including defining a population, an intervention, and using mixed methods.

Copy and paste the following HTML into your website