Dealing with overweight and obesity is complex, as the Foresight report  makes clear. The observations here describe family and provider views on treatment implementation. Our findings confirm the levers and barriers identified in our earlier mapping study . Practical problems such as transport, parents’ work schedules and competing demands on family time were common. All the families we spoke to found these difficult, but there were particular pressures for low-income parents, and this may have implications for access. The providers of this programme clearly worked hard to deliver in low income areas, but in order to optimise the ‘reach’ of such programmes, some of the issues described in the findings here need to be addressed. Delivery partners often put considerable effort into recruiting, retaining and motivating families, which increased uptake but also increased cost. Parents and providers valued highly trained delivery staff, again impacting on costs, and providers often felt the need to make small adaptations to local social and cultural needs. Both providers and parents expressed concerns about the long-term success of the programme, and the way that this could be compromised by an obesogenic environment. In many areas barriers to use, together with concerns about funding, created barriers to commissioning. Where these barriers were not experienced, and where the scheme was perceived as effective locally, commissioners were enthusiastic about continuing to fund MEND.
Most families reported having gained something positive from the experience of participation. However, it was often difficult to sustain the changes associated with the intervention in the longer term. The constraints imposed by the wider environment, the fact that family life was changeable, or simply the ‘pull’ of established ways of life made it hard. Families’ concerns about sustainability were shared by providers who felt that MEND was supportive while it was running, but that children and families needed further support to persevere in the longer term. MEND 7–13 is a 10 week programme with short term follow up. Long term improvement is known to be hard to achieve . A plausible argument has been made  that improvements in child weight management are more likely to take place in families who are engaged and ready to make changes. Among the families we spoke to, living in circumstances which facilitated change (for instance having the time, space and assets to get to and from the programme) was also crucial, as were ‘pilot lights’ for change such as a new school, or a dog with his own exercise requirements.
MEND families have been interviewed in three other research studies to our knowledge. These explore views of MEND delivered through primary care [24
], views while attending sessions [25
], and choices between interventions [26
]. Like us, Turner [24
] found parents wanted advice from someone who they felt had both the professional and personal experience to understand the difficulties they faced. Staniford and colleagues [25
] interviewed families and professionals with experience of a range of obesity treatments. In their study, professionals were disappointed about attrition and lack of long term weight change but also frustrated that families did not become ‘independent’ at the close of the programme. The authors noted that:
“By teaching behaviour change/weight control techniques in a contextual vacuum, participants are highly likely to remain vulnerable to the same environmental influences”  p. 240.
Parenting is an onerous job, which many combine with jobs and job-seeking. Whilst activities within MEND (e.g. supermarket visits and exercise) take some account of context, without wider action on the determinants, creating and maintaining healthy weight may simply be too much. As a community participant in another context tells us, focussing largely on individual behaviour in an unsafe environment can be “like teaching children to swim in a pool full of alligators”  p. 730.
None of these studies contacted non-attenders, and a major limitation of our study was our inability to interview those who had been in touch with MEND but never joined a programme (‘refusers’), or who attended once or only a few sessions (≤25%). This is not a unique failing, a similar study of a family-based child obesity intervention also had low response rates from non-participants . We had reason to be positive about our ability to recruit refusers, having successfully done so following an RCT in the past . Time and interest are likely barriers to responding, but we believe these were exacerbated by two impediments in this study. One was the unexpected sampling challenge associated with service records. Secondly (particularly for busy families), changes to research governance structures meant that we had to take a lengthy route to contact ‘refusers’ in contrast to the more direct methods a decade earlier. Despite this gap in our sample, the ‘good’ attenders we talked to were eloquent in telling us what had made attending and engaging with MEND hard as well as what had ‘worked’ for them.
Methods for understanding and evaluating public health interventions such as MEND which take place in complex social and economic settings are still in their infancy. Influential in terms of theory have been Hawe and colleagues [29, 30] on local context and interventions as events in systems. In terms of methods, the Cochrane Public Health Review Group  has given encouragement to methodological plurality, equity, and attention to users. Guidance on complex interventions and natural experiments have propelled the field forward [6, 32] and funding for robust public health research has increased. In this context it is important to ensure that research results are not viewed as commodities providing simple solutions to complex problems. Implementation issues such engagement, local context, staffing, appropriateness of intervention content, funding constraints and commissioning policies identified here will be common to many public health interventions. The most recent NICE guidance points to the importance of addressing these, particularly the impact of short term funding streams, when developing services for managing childhood obesity .
In this study, a good deal of significance was attached by providers to the positive results of an RCT but, as several pointed out, context – geographical, political and cultural - matters. This adds a further layer of complexity for those wanting to implement evidence-informed programmes. In their analysis of the need for a joined up approach to public health planning for childhood obesity, Hendriks and colleagues point to harsh treatment of interventions which admit to problems, a lack of learning by doing, and a lack of interest in implementation as part of the planning and policy process . Reporting problems and difficulties is counter-cultural, and the norm is to disseminate stories of success rather than learning from what goes wrong. The providers we spoke with clearly felt ‘pinched’ by organisational behaviours which required solutions validated in a research context, but where strong applicability to local context was also needed.
Finally, our interviews were with commissioners/providers and families and not with our colleagues who developed MEND and who also, of course, have expert views. They did not always feel that the perceptions of families and commissioners were correct or fair. However, these perceptions and experiences are among the factors that those implementing weight management programmes take into account.