Lessons learnt from the NAP SACC UK feasibility trial

In this chapter, the practical lessons relating to recruitment, data collection and intervention delivery are examined with a view to informing a full-scale trial. These lessons will also be relevant to the design of other trials within nursery settings. The lessons draw from the process evaluation, experiences of the research team, and discussions with co-applicants, the TSC, the founder of NAP SACC (Professor Dianne Ward), and the lay advisory group. Specific quotes and issues raised in the process evaluation are not repeated here but presented in the form of practical recommendations.

Recruitment

Recruitment of a local authority to participate in a trial should include a thorough check of other early years/preschool health initiatives (including non-health initiatives), which could make recruitment difficult or affect the control groups. This learning particularly draws on the phase 1 findings and decision to change the second local authority from Wales to Gloucestershire, but also the experience within Gloucestershire of other initiatives. The impact of this for recruitment to a full trial could be that it will be more difficult to recruit cities in particular, because many have developed healthy preschool programmes or have other city-wide early years initiatives. It may also mean that recruitment of early years settings may be more difficult in local authorities that are part of the ‘Bristol Standard’ family, where many early years settings are undertaking the Bristol Standard quality improvement programme. These considerations will be taken into account in choice of local authorities for a future full trial.

It was helpful attending the local authority’s regular meeting with child-care providers to inform nursery managers of the trial and, if the study progressed to a full-scale trial, this level of local engagement would be important. Nursery recruitment may be aided by creating a short online film, with interviews from managers, NAP SACC UK partners and parents from the feasibility study highlighting what it involved, as well as the nursery staff and NAP SACC UK partners’ enthusiasm for the intervention and the acceptability of the feasibility study. This approach has been used by GoActive108 (www.cedar.iph.cam.ac.uk/research/directory/goactivestudy/). The film should feature nurseries from deprived areas to highlight potential benefits of reducing health inequalities to attract large nurseries in deprived areas. A future trial should ensure that the memorandum of understanding is signed by the manager (not deputy, to avoid the problem encountered at one nursery, which did not fully take part in the intervention) after a briefing meeting with staff and the manager.

To increase child recruitment rates, the acceptability of using parental opt-out consent for some or all data should be discussed with the ethics committee (of note, opt-out consent was used in the nursery food photography substudy and is used routinely by NCMP for national data collection). Recruitment of children could be increased by involving nursery staff in communicating with parents about the study. This would require a member of the research team to discuss the study and child eligibility with staff (not only the manager) and show staff the recruitment packs. This process could also reduce the number of ineligible children included in the denominator. Attendance by research staff when parents are collecting their children on at least three afternoons per nursery may increase recruitment, allowing parents to consider the study and ask questions. Nursery staff should be provided with a list of children for whom no response has been received so that they can remind parents to return the consent form.

Professor Dianne Ward, the founder of NAP SACC in the USA, has identified lessons for conducting environmental interventions in early years settings. These include:

• Ensure that the intervention is based on a behaviour change theory.
• An advisory group should inform the intervention and delivery.
• Realistic recruitment targets should consider the economic nature of child care.
• Data collection needs to be flexible with realistic sample size calculations because of high rates of child (and sometimes provider) turnover.
• Interventions that are relatively easy to implement are more likely to appeal to a wide variety of providers.109

Of particular note for the NAP SACC UK study is the reality that recruitment, particularly dual recruitment of nurseries and children/parents, is challenging and requires resources and time. Such realism is required with consideration for the sample size and resource requirement for a future trial, particularly if meaningful dietary measures are to be included.

Data collection

Although the acceptability of children wearing accelerometers was good, the number of children meeting the wear-time criteria of ≥ 480 minutes per day was not large. A number of studies that used accelerometers with preschool children have a used less strict wear-time criteria for valid days of accelerometer data. For example, Pate et al.’s study in US preschools included children in the analysis if they provided accelerometer data for ≥ 50% of the school day for 3 days, with a full day being > 6 hours (e.g. 180 minutes per day minimum wear time).48 Stanley et al.52 considered > 3 hours per day on ≥ 1 day as valid wear time for a study in a preschool setting.

Given the turnover of children between early years settings and progression to primary school, consideration should be given to a repeat cross-sectional study design within the RCT to maximise assessment of environmental changes on the children within the nursery at each time point.

The use of laptop computers for CADET data collection was not trialled as it would have been cumbersome to use a laptop in the nursery environment. The use of tablets should be explored further for data collection if adequate battery power and speed of data entry is possible with different devices or models; however, paper data collection was acceptable and practical.

In a full-scale trial, parents should not be asked to report costs of food and PA, health-care use or time off work, because there were too many missing data, these data were highly variable and these costs were unlikely to be affected to a great extent by the intervention.

The methods used in the food photography substudy should be adopted for a phase 3 trial to give more detailed information about portion size and food/drink consumption. Among the parents who responded, a high percentage gave consent for data linkage to NCMP height and weight data collected in schools. Therefore, requesting opt-out consent for data linkage to NCMP should be included in a full trial.

The EPAO data were collected on 1 day for each nursery at baseline and follow-up. With the variability of activities, weather and food provided between days we were concerned that one day of observation was not reliable. The EPAO provides useful contextual information about the nurseries, but it is questionable whether or not it is a reliable enough measure to be included as an outcome measure in a full trial. The original development of the instrument concluded the tool is a ‘stable and reasonably accurate instrument . . . However, a more robust, less subjective measure would be more appropriate for researchers seeking an outcome measure to assess intervention impact.’110

A phase 3 trial should measure how the intervention is sustained over time with respect to staff turnover and whether changes are embedded as ‘the way things are done’ or diluted. Furthermore, an objective measure of oral health should be included as a secondary outcome.

Intervention delivery

To ensure timely completion of the Review and Reflect form and maximise the number of nursery staff attending the two workshops (to coincide with inset days, or avoid planned annual leave and nursery closures), the NAP SACC UK partner and nursery manager meetings and two workshops should be booked with all nurseries at an early stage and prior to randomisation (cancelling the dates in control nurseries after randomisation). This would have the additional advantage of nursery managers being fully aware of the time commitment required for the intervention prior to randomisation as a further check, in addition to the signed memorandum of understanding.

The timing of the elements of the intervention should be:

• NAP SACC UK partner and nursery manager meet to jointly complete the Review and Reflect tool.
• Workshop includes the manager highlighting areas for potential improvement identified in the Review and Reflect tool.
• Staff suggest goals at end of workshop and confirm action plan with NAP SACC UK partner in subsequent meeting.

The intervention should be extended to 1 year to include two cycles of the Review and Reflect and goal-setting process, with a supplementary workshop after 6 months. Process data suggest partners and managers focused initially on easy-to-achieve goals. Allowing for a second cycle of goal-setting would encourage them to extend their goals beyond the ‘easy wins.’ As the workshops appeared important in motivating staff, an additional workshop at this point is recommended to encourage commitment from staff. The intervention could be improved by providing template copies of relevant policies (e.g. nutrition, PA, screen time, oral health, dental registration questions for registration pack).

Communication with nurseries at the beginning and throughout the intervention should make clear when the support from the NAP SACC UK partner will end. In addition, for the control nurseries, there should be ongoing communication so there is not a long period between annual data collection periods. Consideration of provision of an alternative health intervention could be considered for control groups (e.g. sun protection or hand hygiene).

During baseline data collection, it was observed that a small number of children in approximately half of the nurseries were bringing in packed lunches from home. An information sheet111 is available from the Children’s Food Trust on healthy packed lunches for early years and the national Change4Life website112 has a page with ideas for packed lunches. This information could be included in the information provided to the NAP SACC UK partners in a phase 3 trial, as an additional resource. Staff from Public Health England and the Children’s Food Trust anticipated government policy to increase free early years education from 15 to 30 hours for 3- to 4-year-olds (for parents working a minimum number of hours) and in September 2017 may increase lunchbox provision (Alison Burton, Public Health England, 2017, personal communication, and Patricia Mucavele, Children’s Food Trust, 2017, personal communication).23 The five nursery managers attending the lay advisory group thought that increases in lunchboxes would be minimal but would increase in some settings. Therefore, the Review and Reflect tool should also incorporate best practice standards for lunchbox provision. The design of this could further be supported by work on lunchbox quality in the UK,113 Australia43 and USA.114

The pressure of health visitors’ capacity to deliver the intervention necessitates flexibility to consider different models of delivery. If NAP SACC UK is delivered by a less experienced group of staff in a future trial, additional training will be necessary to ensure that these individuals are adequately supported. This should be provided at an earlier stage of the intervention, with additional top-up training to allow staff to discuss emerging issues.

Summary

The feasibility study has identified a number of modifications to improve recruitment, data collection and intervention delivery. These are summarised in Table 28.