Skip to main content
Search form
  • 00:00


  • 00:14

    PAT HODDINOTT: My name's Pat Hoddinott.I'm chair in primary care at the University of Sterlingin the Nursing, Midwifery, and Allied Health ProfessionsResearch Unit.And I'm also a GP.I have been a GP for nearly 30 yearsin London and in rural Scotland.And my research interest is in preventing ill health.

  • 00:36

    PAT HODDINOTT [continued]: So particularly behavior change interventionsthat improve the health of patientsand prevent them needing health care services.The Game of Stones study is a feasibilityrandomized controlled trial of text messagesand financial incentives for men living with obesity

  • 00:58

    PAT HODDINOTT [continued]: who want to lose weight.And this was a study that was fundedby the National Institute of Health Researchpublic health research board.And it was a commissioned call for more researchinto men and about how they can lose weight.Because on the whole, men seldom attend the existing weight

  • 01:19

    PAT HODDINOTT [continued]: management programs.They're predominantly a female arena.And I first became involved in thiswhen I worked at the Health Services Research Unitin the University of Aberdeen, where I worked with ProfessorAlison Avenell on the ROMEO systematic reviews,where we did four systematic reviews

  • 01:41

    PAT HODDINOTT [continued]: of the quantitative and qualitative evidencefor how men can address obesity.And that revealed big evidence gaps.And so that was how the project originated.I also did some work on financial incentives,but that was with women around childbirth.

  • 02:02

    PAT HODDINOTT [continued]: But as part of that, I undertook systematic reviewson financial incentive interventions,and also we looked at the mechanismsof action and the theory for financial incentives.[What was your role in the Game of Stones study team?]

  • 02:23

    PAT HODDINOTT [continued]: I co-led the study with Dr. Stefan Dombrowski, who'sa health behavior psychologist.And we had a team of researchers from Queens University Belfast,University of Aberdeen, University of Dundee,Glasgow Caledonia University, Universityof Highlands and Islands, and University of Glasgow.

  • 02:46

    PAT HODDINOTT [continued]: So we really pulled together peoplewith a lot of different strengthsand developed a strong team with very importantpublic-patient involvement from the Men's Health Forum Charity.And we'd worked with them since about 2010 on the ROMEO Review.

  • 03:08

    PAT HODDINOTT [continued]: So this was a lovely continuing relationshipwith public-patient involvement, which wasa real strength to our project.[How did you plan for the trial and ensure it would appealto the target population?]The research question was already formulated for usby the commissioning brief, and then we looked at the gaps

  • 03:31

    PAT HODDINOTT [continued]: in the systematic review evidence.And also, we looked at the theory.And we also did a lot of public-patient involvementwith men in the target populationto find out what would appeal to them.And so text messages were quite attractive.

  • 03:51

    PAT HODDINOTT [continued]: And financial incentives show some promise.In particular, we found that commitment contractsshowed quite a lot of promise.So this is where men would put money up front of their own.And then if they didn't meet weight loss targets,they would lose that money.But that introduces real health inequalities

  • 04:13

    PAT HODDINOTT [continued]: for poor people who haven't got any money to put up that.So we wanted to find a way to address that.And that's what we did.I've already mentioned systematic reviews.I've already mentioned theory.The next step was really to find out the views of the targetpopulation.

  • 04:34

    PAT HODDINOTT [continued]: So through public-patient involvementand through qualitative research,we interviewed quite a lot of men who werestruggling with weight loss.And based on that, we designed a discrete choice experimentto address our main uncertainties.

  • 04:56

    PAT HODDINOTT [continued]: And I'm not going to talk about discrete choiceexperiments, other than they are about people's preferences.And that was led by Professor Marjon Van der Polat University of Aberdeen, who is an expert in this.But it was a fantastic way to resolve the main uncertainties

  • 05:17

    PAT HODDINOTT [continued]: for our trial design, because we used a commercial companyand they had a panel of 1,045 men who were overweightand in our target population.And so we resolved the main uncertaintiesthrough that discrete choice experimentabout the structure of the incentives.

  • 05:41

    PAT HODDINOTT [continued]: Because you could go out there and youcould do 100 little experiments of lotsof different financial incentivesto try and find the right combination,because these incentives were over a whole 12-month period.How do you decide how much money you give at each time point?How do you decide how much weight loss the reward

  • 06:02

    PAT HODDINOTT [continued]: or the money should be for?So the discrete choice experimentwas a great way to shortcut doinga lot of small tests of change.We also added some other survey questionsto the discrete choice experiment.For example, we wanted to know wheremen would like to be weighed.

  • 06:23

    PAT HODDINOTT [continued]: Would they want to be weighed in a pharmacy, a GPsurgery, a community center?How often would men want to receive texts?Again, we were able to ask questionsthat were big uncertainties from a large populationof our target population.

  • 06:43

    PAT HODDINOTT [continued]: And qualitative research would nothave been able to give us such good answersto those uncertainties because of the small sample.So we were quite confident that the main uncertaintiesof the intervention design were being answered by the menthat it mattered to.[How did you then design the trial?]

  • 07:07

    PAT HODDINOTT [continued]: We had differences of opinion between the men,the clinicians, the commissioners, the researchers,the theory the systematic review evidence.So you have to do a trade-off.And we had some key core principles which we all--I think we all--agreed on.The first one is that we really wanted our intervention

  • 07:30

    PAT HODDINOTT [continued]: to be affordable, sustainable, to beable to translate into the real worldand not be too blue sky thinking, not too ideal world.And we wanted it to address health inequalities.We wanted the underserved populationto be part of this intervention.

  • 07:51

    PAT HODDINOTT [continued]: We had a strong focus on future implementation.So we held those core principles in mind as we made decisions.And there were various trade-offs.And who knows whether we made the right decisionsat the end of the day, but you haveto weigh up all of the sources of information,triangulate, and make a decision.

  • 08:12

    PAT HODDINOTT [continued]: So the design was for three-arm parallel-group feasibilitytrial testing financial incentives with daily textmessages.Text messages alone, and then a 12-month waiting listfor the text messages.

  • 08:32

    PAT HODDINOTT [continued]: And we designed it that way because all of the researchevidence suggests that incentives alone don't work.That you need some element of support there as well.And the waiting lists control, that was quitea tricky decision to make.

  • 08:53

    PAT HODDINOTT [continued]: And but we wanted to compare these interventionswith a life as usual.And that was the idea of the waiting list control group.And so the feasibility trial ran for 12 months.We weighed men at baseline, three months, six months,

  • 09:17

    PAT HODDINOTT [continued]: and 12 months.It was a self-care intervention.So the face-to-face contact for men with the researcherswas purely for data collection.But that did include qualitative interviews at three months.And the reason for that is that we wanted

  • 09:38

    PAT HODDINOTT [continued]: to make any refinements to the interventions or the trialprocesses early on, given that the interventions weregoing to run for 12 months.But there were pros and cons of doing that.[How did you recruit participants?]

  • 09:58

    PAT HODDINOTT [continued]: We were informed by the work of Professor IainCrombie at the University of Dundee,who had successfully recruited some underserved populationof men into a text message interventionto address alcohol problems.And so this used two recruitment channels.The first recruitment channel was

  • 10:20

    PAT HODDINOTT [continued]: to identify GP practices in very disadvantaged areasand for letters to be sent to patients whothey knew had a weight problem.And then the second channel was to try and pick up the men whodon't go and see their GP.Because it's quite common for men

  • 10:40

    PAT HODDINOTT [continued]: to keep away from GPs until they've actuallygot a health diagnosis.So we wanted to get the younger men, the working men.And the way that we did that was to set upstands in community venues.We chose venues where men frequent,and we did this at weekends and in the evenings

  • 11:03

    PAT HODDINOTT [continued]: to make sure we got working men.And we set up stands with our big banners for Game of Stones.And researchers went out in pairs,and we set them up in supermarkets, DIY stores,in council venues, in hospital foyers.And it's a particular way of just being there

  • 11:28

    PAT HODDINOTT [continued]: and seeing who comes up and engages with your study.And we found that that was actually very variable,but quite successful.So on average the researchers hadto spend an hour or two hours to recruit one person.

  • 11:49

    PAT HODDINOTT [continued]: So not too inefficient.And we sent researchers in pairs,because they were going to fairly disadvantaged areasto do this.But we did get a different sample of menfrom that recruitment method.And so we do think it was worthwhile

  • 12:09

    PAT HODDINOTT [continued]: [Was there a strategy for retaining participantsthroughout the study?]We had differential retention across the three trialarms, which is very usual in weight loss trials.More of our control group men turned upthan men in the text message-only arm,

  • 12:32

    PAT HODDINOTT [continued]: and more men in the text message-only armturned up than in the incentives arm.And we didn't really employ any strategiesat all for retention.But the qualitative data was quite revealing,because we hypothesize that some men who

  • 12:52

    PAT HODDINOTT [continued]: didn't meet their targets for the incentivesactually didn't attend because of the stigma,because of the emotions attached with nothaving reached the targets.[What were the ethical considerations of this study?]

  • 13:15

    PAT HODDINOTT [continued]: Financial incentives are always tricky.And we spent a lot of time gainingpatient and public involvement in our recruitment materials.And I think it would also help that I'dbeen chair of an ethics committeefor quite a long time, so we had no ethical problems at all.

  • 13:35

    PAT HODDINOTT [continued]: We did offer a financial incentive of 20 poundsfor men who attended the 12-month final assessment, justto reimburse them for their time and to say thank you.But we didn't really encounter any ethics committee issues.

  • 13:56

    PAT HODDINOTT [continued]: [How did you analyze the data?]We were working with the Universityof Dundee clinical trials unit, and theyheld all of the data for recruitment trackingand all of our outcome data.

  • 14:17

    PAT HODDINOTT [continued]: And the data were analyzed by an independent statisticianwho was blind to the group allocation.The qualitative data, there were six people involvedin that team.Two of whom had nothing to do with the study,and that was to give that independent outside perspective

  • 14:38

    PAT HODDINOTT [continued]: to make sure that we didn't have any blind spots.And they assisted with helping to codethe data and the interpretation of the data.And that was all managed through NVivo 10 software.And we used a framework approach.The DCE data was all analyzed by Professor Marjon Van der Pol

  • 15:01

    PAT HODDINOTT [continued]: at the University of Aberdeen, whohas expertise in that method.So we also had a PhD student, Rebecca Skinner,who was doing a parallel line of research lookingat how the intervention was embedding itselfin the real world of the participants.

  • 15:21

    PAT HODDINOTT [continued]: And that work is ongoing.And again, she was managing her data quite separately.[What were the outcomes of the research?]I don't think it's about impact at this stage.This was a feasibility trial.

  • 15:41

    PAT HODDINOTT [continued]: And our primary outcome was, is this feasible?And we had a sample of 105 men, and we had a priory criteria,which our peer reviewers thought were a bit too strict,actually.But we met them.We over-recruited.

  • 16:02

    PAT HODDINOTT [continued]: We had 177 men who wanted to take partin the studies within 3 and 1/2 months.So we had to turn men away, which we were delighted about.And I think part of that was about the name, Game of Stones.And we held a competition for the name for our study.

  • 16:23

    PAT HODDINOTT [continued]: And we asked all of our PPI and stakeholders,put names into the pot, and then we had voting.And unanimously, the men wanted Game of Stones,they felt it would draw them in, they felt bonded to that.And I think that was quite important.And I think probably as researchers,we don't pay enough attention to the acronyms

  • 16:46

    PAT HODDINOTT [continued]: and that immediate engagement in research.And so in terms of presenting our findings,our findings show that this is acceptable and feasible,and shows promise.But we can't say anything else because of the small numberof participants in the trial.

  • 17:10

    PAT HODDINOTT [continued]: [What unexpected challenges did you encounter?]I think overall, it went very smoothly.I think the one thing that we learnt is that this wasa self-care intervention, and it was designed specificallyto take burden away from primary care and the health services.

  • 17:32

    PAT HODDINOTT [continued]: So text messages and incentives.So the researchers were not giving any interventioncomponent.They weren't giving any advice, but because wedid qualitative interviews at three monthsto make any final refinements to the design of the study,actually, men saw that as supportive.

  • 17:54

    PAT HODDINOTT [continued]: And what we realized is that some men just reallywelcomed those open questions and being listenedto about their experiences.And I have always argued--and in fact, published a paper about qualitative researchand feasibility trials--saying that you should keep the trial clean,and either do qualitative research at the very start

  • 18:17

    PAT HODDINOTT [continued]: or at the end.And I broke my own rules there, but for a very good reason.And in fact, it's given us a real lessonabout how we need to--if we go to a full trial--really protocol what happens when

  • 18:38

    PAT HODDINOTT [continued]: the researcher weighs the man.And to make that consistent and standardizedthroughout the trial.And certainly in a full trial, therewould not be any qualitative interviews.[What resources do you recommend for others considering thisapproach?]

  • 18:59

    PAT HODDINOTT [continued]: When we started out, we became very awarethat there was this huge black hole about howto design interventions, and particularly,how you make decisions when you're facedwith a lot of uncertainties?And so what we decided to do was to use the TIDieR guidanceon reporting interventions and use it backwards,

  • 19:21

    PAT HODDINOTT [continued]: so that we could populate it.We were also guided by the RE-AIM frameworkby Russell Glasgow and colleagues,which very much talks about reach and sustainabilityand future implementation.But now I'm pleased to report that I'vebeen working with colleagues at the University of Sheffield,and Bristol in South Hampton, and we have produced guidance

  • 19:45

    PAT HODDINOTT [continued]: on intervention development which is published in BMJ Open.And it addresses a lot of the issuesthat we were struggling with at the beginningof our intervention development study.And that's now publicly available.Index is not a checklist.It's not set in stone.Because there's so little evidence

  • 20:06

    PAT HODDINOTT [continued]: about what works in intervention developmentthat leads to successful change in practice and improvementof health outcomes.It has been estimated that only one out of every 100medical innovations actually makes itinto the real world of practice.So there's a huge black hole here.And index sets out five overarching principles,

  • 20:32

    PAT HODDINOTT [continued]: and 10 actions to consider.They're only to consider, but one of the strongest messagesfrom index is that we report the intervention developmentprocess.So that in 10 years, 15 years time, researcherscan look back at how interventionshave been developed and we can understand much more about what

  • 20:55

    PAT HODDINOTT [continued]: works and what doesn't work in intervention development.So in summary, this is my take on the Game of Stones study,but it was very much a team effort.And I would like to acknowledge all of my co-investigatorsand the excellent researchers we had on the team,together with the men and all the stakeholders that

  • 21:15

    PAT HODDINOTT [continued]: were involved in consultations and workshops.I'd also like to acknowledge Professor Iain Crombie'swork that very much informed our overall design,and Dr. Mark Grindle, who wrote the text messages.Finally, I'd like to acknowledge our funders, NIHR, for funding

  • 21:36

    PAT HODDINOTT [continued]: this research, and emphasize againthat this is my take on the Game of Stones intervention.

  • 22:08



Professor Pat Hoddinott, Chair in Primary Care at the University of Sterling in the Nursing, Midwifery, and Allied Health Professions Research Unit, discusses the Game of Stones randomized controlled trial of text messages and financial incentives for men living with obesity who want to lose weight.

Looks like you do not have access to this content.

Game of Stones: A Feasibility Randomized Control Trial

Professor Pat Hoddinott, Chair in Primary Care at the University of Sterling in the Nursing, Midwifery, and Allied Health Professions Research Unit, discusses the Game of Stones randomized controlled trial of text messages and financial incentives for men living with obesity who want to lose weight.

Copy and paste the following HTML into your website