Aims: to focus on understanding and accommodating the perspectives of the people who will use the intervention, in order to improve uptake, adherence and outcomes
Methods:
1. Carry out iterative qualitative research with a wide range of people from the target user populations throughout the intervention development and deployment
2. Identify ‘guiding principles’ that can inform intervention development by highlighting key behavioural issues that the intervention must address
2

• Approach evolved as learning process developing over 20 interventions for public health and illness management
• Based on >1000 qualitative interviews with users, plus evidence from trials of our intervention effectiveness
• Timely to share what we have learned with novice developers, stimulate debate with experienced developers
3

• Person-centred therapy (Carl Rogers) – based on empathy and respect for autonomy of the person
• Patient-centred medicine
• Theory- and evidence-based medicine
• Usability testing, user satisfaction
• User-centred/human-centred design
4

The process of incorporating a person based approach into intervention development
In-depth inductive qualitative research carried out to inform:
• Intervention planning (including formulating guiding principles)
• Intervention development
• Intervention implementation
Cyclical not linear process
5

• Open-ended questions initially (can use theory-based questions later), to elicit personal stories
‘How do you feel about the intervention now?’
• Ask about intervention not website/app
‘Can you tell me about the last time you tried to follow the intervention advice?’
• Focus groups and junior/independent researchers can be useful for eliciting genuine views!
6

Inductive primary qualitative research (AND/OR qualitative synthesis of existing studies) can provide understanding of user perspective and key context-specific behavioural issues - will help developer to:
• Select theory- and evidence-based techniques that are most acceptable, salient, feasible for target population
• Avoid or modify intervention characteristics that are disliked, impractical, intrusive
• Suggest the need for new intervention characteristics, hence not yet evidence-based
8

Formulating guiding principles can inform intervention development by highlighting the distinctive ways that it will address the key context-specific behavioural issues: a) Identify key intervention design objectives
(based on issues, needs identified as crucial to intervention success) b) Identify key features of the intervention that can achieve those objectives
Key features could include behaviour change techniques (from intervention planning - e.g. goal setting), technology characteristics
(e.g. brief modules for mobile phone), implementation setting (e.g. primary care), etc. 9

Qualitative work at the planning stage identified key behavioural issue that obese people had multiple previous experiences of ineffective long-term weight management
10

POWeR design objective 1: to persuade users that this approach to weight management will work
Key features that can achieve this aim
• Distinctive – containing new, surprising and interesting content, e.g. ‘POWeR tools’
(self-regulation techniques)
• Explicitly evidence based, presenting scientific rationale for recommendations and proof of their effectiveness
• Non-commercial, developed by named team of medical and behaviour change experts, linked to NHS, supported by nurse
11

POWeR design objective 2: to promote long-term adherence and maintenance of weight loss
Key features that can achieve this aim
• Emphasis on building autonomous motivation , e.g. non-prescriptive approach, avoid feelings of ‘deprivation’
(no forbidden foods, choice of eating plans and goals)
• Focus on creating lifestyle-compatible long-term habits
(simple eating goals, less reliance on calorie counting, food diary)
12

Further inductive qualitative research essential to gain insight into whether all intervention components a) comprehensible, acceptable, feasible b) easy to use, motivating, enjoyable, informative, convincing
1. Think aloud studies to elicit range of target users’ reactions to every element of the intervention!
2. Allow users to try intervention for a few weeks, keep diaries -- retrospective interviews about experiences
14

Example: Qualitative study of experiences of the prototype POWeR intervention
Think aloud interviews with 16 people (8 women) aged 18 to
>65 using first session of POWeR:
• introduced users to the POWeR philosophy
• allowed them to choose low-calorie or low-carbohydrate eating plan
• helped them to set weekly goals and make ‘if-then’ plans
15

Key findings influencing intervention modification
Perceptions of choice of eating plan
Positive aspects
“The fact that there’s a choice immediately is attractive cause I can go for either or.”
“I like that it’s a long term plan that you can stick to, rather than a really strict regime that as soon as you stop, it comes back on.”
“I find a low calorie too restricting because you end up with having to count or seeming to count everything.”
“So it’s almost saying, you know, not everything is banned, it’s just some are going to be the more occasional ones. So it makes you feel like you’re not going to give anything up.’”
16

Key findings influencing intervention modification
Goal setting
Concerns
“Ok, so ‘eat as much as you like’ [of low carbohydrate foods] doesn’t –
I’m not sure if it will help me lose weight or not”
“It would be useful for me to know what my daily calorie allowance would be.”
“What is easier about those [other] kind of programmes, is that you don’t have any choices to make, you know, you can’t say, oh, ‘should I have a lettuce leaf or should I have a peanut?’ You just eat the packet of soup or whatever. So actually it makes it extremely easy to lose weight.”
17

Key findings influencing intervention modification
Perceptions of choice of goal setting (positive and negative)
“You think, oh, OK, I’ll just have to do that small goal, I don’t have to change my whole eating habit, it’s just small bits.”
“And for me, having a plan with reasonable goals is really important.
Because I don’t know if I am a typical man but I set myself too higher goals too often and fail.”
“I wouldn’t know what to write at this point, you know … I’m not sure the information that was there helped me to find out what mine would be.”
“It’s very easy to say I’ll reduce my portion size here but one of my goals isn’t to stop snacking so I can just snack as much as I like.”
18

Expanded rationale for ‘healthy habit’ vs. ‘diet’ approach
Made calorie counting available for those who really want or need it (e.g. for use as diagnostic aid if not losing weight)
Provided drop-down menu of simple goal-setting options likely to result in weight loss (e.g. reduced portion size, swapping all unhealthy snacks or high calorie drinks)
19

Person-based intervention implementation
Final stage of qualitative research evaluates experiences of implementing the intervention (in full trial or in real-life roll-out)
Findings triangulated with quantitative evaluations to help explain usage patterns and outcomes, inform further implementation
21

Example: Mixed methods study of community roll-out of POWeR
Randomisation
N=786 (69.5%)
Control
N=275 (35%)
Web only
N=264 (33.6%)
Coach
N=247 (31.4%)
Coaching Protocol:
2 short phone calls from a ‘POWeR coach’ in week 1 and week 4

80 sessions
60
50
40
30
20
10
0
1
% of sample still using POWeR at each session
Web only
47 (17.8%)
Web+ Coach
64 (25.9%) ᵪ
Significance test
2
(1,n=511)=4.93, p=.026
2 3 4 5 6 7 8 web coach

2.5
2
0.5
0
1.5
1 p p p
Control Web only Web + Coach

Exploring coaching uptake
(quantitative data)
• Limitations of coaching:
– only 23.5% had one phone call , 18.6% had both
• Benefits of coaching:
– uptake of coaching associated with
– older age, higher BMI at baseline, hypertension, referred to a weight loss scheme by health professional, lower health literacy (trend)
– more sessions completed, more log-ins, more time online
– more weight loss

“I like to have someone say well done when I have done well” [P5]
“I will have to try harder this week, you know, somebody’s is looking after me. I can’t let her down”(P4)
“The second time she called I was struggling, ‘cos I hadn’t lost any weight for a couple of weeks… and I was a bit frustrated . But just having talking to her, sort of helped me sort of to not go and rubbish the diet. And so, actually after getting off the phone to her I actually changed what I was doing that night to actually get myself back on track. “(P1)
“I expected more support. Don't get me wrong, I didn't expect them to ring me every week but like...I got two phone calls in eight weeks”
[P7]
“The calls were good but they were when you [i.e the POWeR team] wanted to do them rather than maybe when I needed them” [P1]

Key intervention design objective: to deliver support as and when needed by user
Key features that can achieve this objective
Support offered:
– to all, but not as compulsory
– in variety of formats (email, phone, face-to-face)
– at various time-points
– with option for patient to initiate support
27

• Not prescriptive - many different combinations of qualitative and mixed methods can be used
• Not always possible to fully implement the person-based approach if time/resources constrained - but vital to devote sufficient resource to development before trialling
• Not the only good way to develop interventions but provides explicit process for identifying:
– the key behavioural issues from the user perspective
– distinctive intervention features that will address them
28

The PBA complements theory-based intervention development by a) suggesting which BCTs may be most important in a particular context b) providing guidance on how to make them acceptable and persuasive
Intervention design can benefit greatly from being not only theory- and evidence-based but also person-based!
29

Leanne Morrison Steph Hughes
Katherine Bradbury Emily Smith
Ingrid Muller
Sarah Williams
Judy Joseph
Sascha Miller
Elaine Douglas
Laura Dennison
Adam Geraghty
Rosie Essery
Rosie Stanford
Jin Zhang
Mary Steele
Ben Ainsworth
Kate Morton
Alison Rowsell
Plus many other coinvestigators and research team members …
30

31