|
| Definition | Studies undertaken |
|
| (1) Developing a complex intervention | |
|
| (1.1) Identifying the evidence base by carrying out a systematic review | (i) Reviewed the epidemiological evidence for early life risk factors for obesity.
(ii) Improved understanding of the target behaviour.
(a) Systematic review of parents’ experiences of bottle-feeding to understand
how parents decide on quantities and frequency of formula-milk feeds.
(b) Systematic review of determinants of early weaning: “Determinants of
early weaning and early use of cow’s milk” identified determinants of
noncompliance with infant feeding recommendations.
(iii) Identified existing systematic reviews and checked the controlled trials register for trials of interventions during infancy. |
|
| (1.2) Identifying/developing appropriate theory by drawing on existing evidence and theory, supplemented if necessary by primary research, for example, interviews/focus groups with “stakeholders”, that is, those targeted by the intervention or involved in its development or delivery | (i) Literature review and team discussions to decide on theory, behaviour change techniques, and intervention strategies.
(ii) Qualitative studies with all stakeholders to refine intervention content. These included interviews and focus groups with mothers (recipients of the intervention) and healthcare providers (who would deliver the intervention). In order to optimise the intervention, an iterative process was used with involvement of mothers, behavioural scientists, doctors, midwives, and health visitors. |
|
| (1.3) Modelling process and outcomes by using a “causal modelling approach” that could include a range of primary and desk based studies to design the intervention, identify suitable measures, and predict long-term outcomes. | (i) Used a causal modelling approach to link “behavioural determinants” to “behavior” and “short-term and long-term outcomes”.
(ii) Developed and validated a questionnaire for use in the trial to assess change in key constructs along the causal pathway targeted by the intervention. |
|
| (2) Assessing feasibility and piloting methods | |
|
| (2.1) Testing procedures for their acceptability, compliance, and intervention delivery | (i) Tested components independently for feasibility and acceptability and final adaptation of the intervention.
(ii) 1 year pilot trial of combined intervention components. |
|
| (2.2) Estimating recruitment and retention and identifying potential barriers to these, using a mixture of qualitative and quantitative methods | (i) Recruitment through post-natal wards, GPs, Health Visitors, midwives, pharmacies, NHS database, charities, and the media to identify most efficient and effective methods.
(ii) Pilot trial over 1 year. |
|
| (2.3) Determining sample size by anticipating the effect sizes in a pilot study | Pilot trial was too small and no previous trials in this area hence used data from observational studies to estimate sample size. |
|
| (3) Evaluating a complex intervention | |
|
| (3.1) Assessing effectiveness by using a randomised controlled trial where possible, choosing the primary and a range of secondary outcomes, and collecting data on predictors or mediators of effect and any possible adverse effects | Set up explanatory RCT (ISRTCN number 2081469). Primary outcome is growth-related and data on a number of secondary outcomes along the causal pathway are also collected. Weight faltering in the babies and reduced quality of life in mothers monitored real time as potential adverse effects reported to independent data monitoring committee. |
|
| (3.2) Understanding change processes provide insights into why an intervention fails unexpectedly or why a successful intervention works and how it can be optimised. Process evaluation nested within a trial can be used to assess fidelity and quality of intervention delivery, clarify causal mechanisms, and identify contextual factors associated with variations in outcomes. Process evaluations should be conducted to the same high standards and reported just as thoroughly as evaluation of outcomes | (i) Intervention fidelity assessment using prespecified checklists.
(ii) Qualitative study nested within the trial-individual interviews with mothers in the intervention and control groups and intervention facilitators to explore how feeding decisions are made, how the intervention might work (or why it may not work) and can be optimised, to identify key ingredients that could be included in future interventions and other contextual factors.
(iii) Mediation analyses to understand how the intervention achieved any effects.
|
|
| (3.3) Cost-effectiveness analyses should be included if at all possible, so that the results are useful to decision makers | Cost-consequence analysis planned and data collection on health service utilisation and maternal quality of life in addition to cost of delivering the intervention. |
|
| (4) Implementation and beyond | |
|
| (4.1) Dissemination by publication in peer-reviewed literature and also communication with policy makers | Peer reviewed publications, conference presentations, public engagement activities, newsletters, and open access web deposition at the end of the trial. |
|
| (4.2) Surveillance, monitoring, and long-term outcomes to measure rare or long-term impacts, using routine data sources and record linkage or by recontacting participants | Consent to recontact participants and access routinely collected health and anthropometry data. If intervention is shown to be effective, process and outcome data could inform a future pragmatic trial. |
|
