The feasibility of using a parenting programme for the prevention of unintentional home injuries in the under-fives: a cluster randomised controlled trial

Burden of disease from unintentional injury

Unintentional injury is a major cause of death and disability in children globally,1 and in the UK it is the leading cause of preventable death in children over the age of 1 year. In addition to those who die from injuries, many more suffer morbidity and possible long-term consequences. Half a million children aged 0–4 years attend UK hospitals every year because of a home injury, representing 78% of all injuries occurring to children in this age group. 2 The type and location of child injuries varies with age and developmental stage. Longitudinal cohort studies have shown that the majority of preschool injuries occur within the home, both in this country3,4 and in other high-income countries. 5 Injuries result from falls, hitting, being hit or crushed by objects, poisoning, and burns or scalds. 2

Risk factors for injury

A number of child, family and environmental factors are associated with increased risk of injury. Socioeconomic disadvantage is the risk factor most strongly associated with child mortality and morbidity. 6–12 Family structures such as single-parent families, step-families and large families, and factors that may relate to caregiver supervision such as very young parents, maternal life events, maternal depression and parental behaviours such as excessive use of alcohol, have all been associated with increased risk of child injury. 3,13–18 Male sex and difficult behaviour in childhood, particularly that relating to antisocial, aggressive or hyperactive behaviour, have been associated with increased incidence of unintentional injuries in the UK19,20 and in other high-income countries. 21

Parenting programmes

Parenting programmes are short-term interventions to promote changes in the behaviour of parents that support children. 22 They have been increasingly recognised as an intervention to improve the life chances of children owing to their effectiveness in reducing antisocial behaviour and conduct disorder, increasing educational attainment, and improving mental health and well-being outcomes in children and their parents. Consequently, they have become a core component of child and family policy in the UK. 23,24 Parenting programmes are usually delivered as face-to-face programmes, either individually or in groups. They have been developed on the basis of two main theoretical approaches, behavioural and relational, although some programmes combine elements of both. Behavioural approaches aim to develop parental understanding of the negative impact of attention to problem behaviour and lack of attention to positive behaviour, and to teach positive discipline practices including praise and time-out; relational programmes aim to improve interactions between parent and child, correcting inappropriate parental interpretations of child behaviour, increasing empathy and understanding of developmental phases.

Analyses of longitudinal studies have shown the influence of parents on child outcomes that are related to injury risk. Positive parenting behaviour and parent–child interaction, and a stimulating home environment have been associated with enhanced development by the age of 3 years25 and improved cognitive and behavioural outcomes in children by age 5 years26 or children who are well adjusted and developmentally competent. 27 The use of positive parenting practices, such as increased use of praise to encourage desirable behaviours, is associated with a reduction in injuries. 28 Supportive parent training can improve childcare practices for mothers with learning difficulties29 and enhanced carer supervision can reduce injury risk to children. 30,31 Parenting interventions have the potential to reduce poor maternal mental health and increase maternal self-efficacy,32,33 to improve maternal–child interactions,34 and to change child behaviour, especially behaviour that is challenging or could place the child at risk of injury. 32,35–37 Reductions in injury risk could also be mediated through information to enable parents to make realistic expectations of their child’s development and skills,38 enhanced parental knowledge of safety practices,39 improvement in the quality of the home environment,40 or through the use of home safety practices such as having a fitted and functioning smoke alarm, using stair gates or keeping sharp objects in a safe place. 41,42 Generic parenting support interventions delivered by health visitors, and which may or may not include a focus on injury prevention, have been shown to reduce injury rates in both prospective observational studies43 and RCTs. 44 Meta-analyses of RCTs measuring one-to-one parenting interventions that are delivered primarily through home visiting and primarily conducted with high-risk or disadvantaged families have demonstrated significantly lower risks of injury, as measured by parental self-report of either medically or non-medically attended injuries,45,46 but it is unclear if group-based programmes can achieve similar effects. Parental understanding of the relationship between injury risk and child behaviour and development is variable, and provision of educational anticipatory guidance has been recommended. 47

There is strong evidence that home safety education with the provision of safety equipment is effective in increasing a range of home safety practices. 42 The features of parenting interventions that are most effective are becoming clearer. A review of ‘what works?’ in parenting interventions has shown that interventions are more likely to be effective if they are delivered early in childhood, if intensity is proportional to need, if they include group activities where parents can benefit from the social aspect of working with peers, if they include formal programmes or manuals to maintain the consistency of the delivery of the intervention which should be delivered by trained staff, and if there is a focus on specific parenting skills and practical ‘take-home’ tips. 48 Parents value programmes that enable the acquisition of knowledge, skills and understanding, and facilitate acceptance and support from other parents. 49 The fear of being perceived as a bad parent may inhibit participation in programmes. Positive outcomes from programmes reduce feelings of guilt and social isolation, increase empathy with children, and give confidence to cope with challenging child behaviour. 50

Cost-effectiveness of parenting programmes

As well as being a health and well-being issue, child injury also has economic impacts. Scarce resources with competing uses in all health systems, and the need to decide between new, ‘efficacious’ primary prevention parenting programmes on the grounds of cost-effectiveness, have increased the significance of economic evaluation as a concept and methodology. 51,52 Recent guidance from the UK Medical Research Council for the development and evaluation of complex behavioural interventions suggests that efficacy and cost-effectiveness should be established before programmes are implemented at the population level. 53,54 However, the meaningful determination of these criteria is often problematic for complex interventions. It is therefore important to develop the conceptual and measurement process by which effectiveness and cost-effectiveness can be evaluated during a feasibility trial.

There are some studies to build upon. The cost-effectiveness of parenting programmes had not been widely studied55 at initiation of this study, but some evidence for modelling costs and longer-term savings56 and estimating the long-term return on investment57 has emerged during the time frame of this study. Previously, a systematic review of economic evaluations of child and adolescent mental health interventions demonstrated that most evaluations were small scale, had short time horizons for assessing outcomes and had limited reporting,58 a finding supported by a recent review of UK programmes to prevent child behaviour problems. 59 Some evidence of cost-effectiveness of parenting programmes has been published for group parenting programmes; a formal evaluation of one programme widely used in English Sure Start children’s centres demonstrated improved child behaviour outcomes for modest costs and considered the programme value for money. 60 This study has been used to produce a costing publication61 now widely used for this type of evaluation.

Objectives

The primary objective of the review was to update the evidence on the effectiveness of parenting programmes in preventing unintentional injury in childhood. The secondary objective was to evaluate the effectiveness of parenting programmes at increasing possession and use of home safety equipment and parental safety practices.

Methods

Results

Description of studies

Twenty-two studies70–90 from 30 articles were included in the review (Figure 2 and see Appendix 1). Some authors reported results from the same study at different follow-up time points in separate papers and several authors reported results from the same study in more than one paper. Sixteen70–72,74,76–78,80,82–85,87,89,90 (73%) included studies were RCTs, two73,81 (9%) were non-RCTs, one86 (5%) was a partially randomised study with two randomised intervention arms and one non-randomised control arm, two75,88 (9%) were CBA studies and one79 (5%) was a quasi-RCT. Four studies73,75,81,88 used clustered allocation. Thirteen studies72–74,78–80,82,84,85,88–90 (59%) were from the USA, three from Australia70,83,87 (14%), two each from Canada76,86 (9%) and England71,75 (9%), and one each from Ireland81 (5%) and New Zealand77 (5%). Fifteen of the studies70–72,74,75,77–79,81,83–85,90 recruited socioeconomically disadvantaged participants. Two studies76,87 recruited participants with a learning disability, and three studies81,88 recruited consecutive newborns from a range of paediatric practices.

FIGURE 2.

Quorum flow chart detailing process of study selection for the review.

Seventeen studies70–75,77–81,83–86,89,90 evaluated multifaceted home visiting programmes aimed at improving a range of child, and often maternal, health outcomes. Three82,88 evaluated paediatric practice-based multifaceted interventions, aimed at improving a range of child health outcomes, all of which included some home visits. Two studies76,87 provided solely educational interventions in the home. None of the studies had injury prevention as a primary focus. All studies provided the intervention to individual parents. Four studies83,88,90 provided opportunities for peer support from other parents, one89 provided informal support from family and friends, and five studies78,80,82,85,90 provided parenting education to groups of parents which as a consequence would also provide opportunities for peer support.

Of 16 studies70,72–75,77–81,84–86,88,89 (73%) reporting medically attended or self-reported injury, two73,75 reported insufficient data to be included in the meta-analyses. Seven studies73,75,76,82,88,89 reported a range of safety outcomes such as use of socket covers and stair gates. Two studies87,89 reported home hazards using different tools, and one study82 reported scores from a home safety index. Ten studies70–72,74,83–86,89,90 measured the quality of the home environment using the HOME inventory. 69 One study73 measured the quality of the home environment using the Massachusetts Home Safety Questionnaire.

Effects of interventions

Medically attended or self-reported injury

Sixteen studies70,72–75,77–81,84–86,88,89 reported medically attended or self-reported injury; 10 of these70,72,74,77,78,80,81,84,88,89 reported results from RCTs and were included in the primary analysis which showed that intervention arm families had a statistically significant lower risk of injury than control arm families (RR 0.83, 95% CI 0.73 to 0.94) (Figure 3). There did not appear to be any evidence of publication bias among the 10 RCTs in the primary analysis {Egger’s test regression coefficient = –0.65 [standard error (SE) 0.49], p = 0.22}. Sensitivity analyses were undertaken for the primary analysis including only RCTs at low risk of various sources of bias. The findings were robust to including only those studies at low risk of detection and attrition bias but the effect size became statistically non-significant when analyses were restricted to studies at low risk of selection bias in terms of inadequate allocation concealment.

FIGURE 3.

Forest plot of medically attended or self-reported injury data – RCTs only. M–H, Mantel–Haenszel.

Home safety outcomes

Studies reported home safety practices and hazards using a variety of methods and scales. Data on total HOME scores at 12 months from three RCTs were included in a meta-analysis. The results showed that there was a non-statistically significant difference in pooled average total HOME scores between intervention and control arm families [mean difference 0.57, 95% CI –0.59 to 1.72; χ² = 0.41, 2 degrees of freedom (df), p = 0.82; I² = 0%], with intervention arm families scoring higher. Armstrong et al. 70 reported organisation of the home environment subscale scores and found a statistically significant difference favouring the intervention arm {mean score intervention arm 5.70 [standard deviation (SD) 0.77] vs. mean score control arm 5.11 (SD 1.16), p < 0.05}. Of the six studies72,83,84,86,89,90 not included in the meta-analysis owing to insufficient data but which reported total HOME scores or organisation of the environment subscale scores, two84,86 found statistically significant differences favouring intervention arm families, and four studies72,83,89,90 found no statistically significant difference between treatment arms, with one study83 finding statistically significant differences only among distressed mothers.

Seven studies73,75,76,81,88,89 reported a range of specific safety practices or use of items of safety equipment, such as use of electric sockets covers and stair gates, but each study measured different practices. Of the seven studies, five75,76,81,88,73 found effects favouring intervention arm families. Two studies87,89 reported measures of home hazards. Olds et al. 89 reported statistically significantly fewer observed hazards in the homes of intervention arm families than control arm families. Llewellyn et al. 87 found that intervention parents identified statistically significantly more dangers within the home and implemented a statistically significantly greater number of precautions to reduce the risk of injury than control arm parents. Three studies used composite home safety measures other than the HOME scale: two used the Home Safety Index82,91 and one used the Massachusetts Home Safety Questionnaire. 73 Families in the intervention arms of these three studies scored statistically significantly higher on safe practices and safer homes than the control group families.

Discussion

We found that parenting interventions, most commonly provided on a one-to-one basis in the home as part of multifaceted interventions to improve a range of child (and often maternal health) outcomes during the first 2 years of a child’s life, were effective in reducing self-reported or medically attended injury among young children. The finding was consistent across studies with little evidence of statistical heterogeneity between effect sizes and was robust to most aspects of study quality and study design. There was evidence that parenting interventions can have a positive effect on the use of home safety equipment and practices.

The strengths of the review include a comprehensive search strategy that included searching grey literature and hand-searching conference abstracts. The analysis adjusted for cluster allocated studies where necessary and a range of sensitivity analyses were undertaken to test assumptions regarding the potential for bias, uncertainty as to the extent to which the intervention was based on a protocol, manual or curriculum, follow-up period and injury type. The findings were robust to these assumptions. Limitations relate mainly to the generalisability of the findings, particularly the study populations which mainly comprised families considered to be ‘at risk’ of adverse child health outcomes. All studies provided the intervention to individual parents, and while several also included some parents’ groups, the findings may not be generalisable to group-based parenting interventions. Similarly, most studies provided the intervention mainly within the home, and so the findings may not be generalisable to parenting interventions provided outside the home.

Most studies used parental reports of injuries which may be subject to biased reporting, particularly as blinding participants to treatment arm allocation is not possible with interventions such as these. The quality of studies was variable, with either half or more of the RCTs included in the meta-analysis being susceptible to bias in terms of allocation concealment and/or outcome assessment. However, despite this, sensitivity analyses demonstrated little impact of excluding studies without blinded outcome assessment on the results. Only two studies included in the meta-analysis reported high attrition rates.

While our review suggests that parenting interventions are likely to improve home safety, there are other plausible explanations for why parenting interventions may reduce childhood injuries. All studies included in the primary and secondary meta-analyses were aimed at improving a range of child (and often maternal) health outcomes. Seven of these studies reported statistically significant improvements in child behaviour, four reported less punitive discipline practices among intervention group parents, six reported increased or improved mother–child interaction and two reported improvements in maternal psychological well-being. It is therefore possible that the reduction in childhood injuries may be mediated through improvements in child behaviour, more effective supervision or discipline practices or more positive interactions between mother and child, all of which may be associated with improved maternal psychological well-being.

Implications for this study

This systematic review supports the evidence that parenting interventions reduce parent-reported and medically attended child injuries and increase home safety practices and behaviour. The findings are stronger with regard to injury reduction than to home safety practices, suggesting that the mechanism may be a generic change in parenting. The review did not identify any group-based, community-delivered programmes such as that proposed for this study.

Learning points

Abstract from manuscript accepted for publication

Parenting programmes have the potential to improve the health and well-being of parents and children. A challenge for providers is to recruit and retain parents in programmes. Studies researching engagement with programmes have largely focused on providers’, policy makers’ or researchers’ reflections of their experience of parents’ participation. We conducted a systematic review of qualitative studies where parents had been asked why they did or did not choose to commence, or complete programmes, and compared these perceptions with those of researchers and those delivering programmes. We used data-mining techniques to identify relevant studies and summarised findings using framework synthesis methods. Six facilitator and five barrier themes were identified as important influences on participation, with a total of 33 subthemes. Participants focused on the opportunity to learn new skills, working with trusted people, in a setting that was convenient in time and place. Researchers and deliverers focused on tailoring the programme to individuals and on the training of staff. Participants and researchers/deliverers therefore differ in their opinions of the most important features of programmes that act as facilitators and barriers to engagement and retention. Programme developers need to seek the views of both participants and deliverers when evaluating programmes.

Implications for this study

The review identified key features of programmes that enabled or hindered parental engagement and retention. Participants appeared to prioritise different issues compared with those delivering, or researching, programmes. However, on exploration, some of these issues were not entirely unrelated. For example, parents emphasised the need for trust in the person delivering the programme, while deliverer training appeared to emphasise the ability to deliver the manualised content of the programme rather than the ability to facilitate a group, including issues such as group cohesion. We have summarised the implications for the development of the parenting programme in this study in the learning points below.

Learning points

The key learning points from this review that are relevant to this study are detailed below.

Commissioning of Parenting UK to develop the programme

Parenting UK was commissioned to develop a group-based parenting programme. The commissioning brief specified that the programme should include:

• parenting skills that have the potential to prevent injuries including, but not limited to:

  • relationship building

  • setting and maintaining boundaries

  • behaviour management

  • appropriate levels of supervision for the age and development of the child

• first aid response to common injury scenarios occurring to children under the age of 5 years in the home, including but not limited to:

  • falls

  • burns and scalds

  • ingestions and poisonings

  • foreign bodies and choking

  • unconsciousness/recovery position

  • life-saving skills/cardiopulmonary resuscitation

  • cuts and wounds

  • broken bones

• safety practices and equipment that, when used in an age-appropriate and/or development-appropriate way, can prevent injuries from occurring.

In addition, Parenting UK was requested to provide the materials, equipment and documentation to support the delivery of the programme during the testing of the intervention. This included the development of a programme manual for those delivering the programme to use as a reference aid. Six sets of materials were required to be produced for subsequent testing in a feasibility study.

Governance of the programme development

A programme development subgroup (PDS) was convened to oversee the development of the programme. Membership included three co-applicants: the chief investigator (JM), an academic with expertise in evaluating parenting programmes (S-SB) and a practising health visitor (BP), together with the Director of the Whoops! Child Safety Project, which provides first aid training for parents. This team communicated with the chief executive of Parenting UK and the manager employed by Parenting UK to carry out the development work. Communication between the manager and the PDS was by e-mail and teleconferences held every 3 weeks between February 2011 and July 2011. The emerging findings of the two systematic reviews conducted during the theoretical phase of this study were passed to the staff at Parenting UK to help inform development of the programme.

The Parenting UK manager was asked to keep a log to record the decisions and rationale for choices made during the development of the programme, to provide an interim report after 3 months, and to provide a final report. The manager from Parenting UK made a presentation to the research team at the end of 6 months (July 2011) to describe the course that had been produced, and seek final sign-off prior to production of the materials.

User involvement

The PAG was consulted prior to Parenting UK commencing work on the programme. During the PAG’s second meeting we asked the mothers what they had liked and disliked about any parent support courses that they had attended and what had been good or not so good about any first aid training or courses that they had attended. We then discussed the possible content and format of a course on first aid and home safety including resources and things they could do at home. They made some very helpful suggestions, which were passed onto the staff at Parenting UK.

At the next meeting of the PAG some of the resources developed by Parenting UK for use in the course were discussed. The outline of the course was described and members discussed some of the proposed activities. They were very positive about the content and resources, and this was fed back to the staff at Parenting UK.

Support for those delivering the programme

The manager from Parenting UK made a number of recommendations regarding support for those delivering the programme. It was acknowledged that support to all the deliverers in a group face-to-face environment would enable shared learning and avoided duplication of information exchange. However, it was recognised that the courses were unlikely to be delivered concurrently across the study centres, resulting in deliverers having different support needs at different times. In addition, not all of the course deliverers would be available at the same time. Therefore, the manager at Parenting UK offered to be available to provide weekly telephone and e-mail support as required for those delivering the four courses planned during the feasibility study. The option of video conference contact was considered but rejected owing to the limited electronic access available to course deliverers in their usual work locations.

Elements required to be captured by the tool

The tool needed to capture the following features of each injury event:

• the date of the event (to confirm occurrence was within the follow-up period)

• the location of the injury event (to confirm that it was an injury occurring within the home, garden, drive or yard)

• the type of injury sustained (e.g. cut, fracture, head injury, etc.)

• the action that was taken [e.g. first aid, telephoned general practitioner (GP), took child to hospital, etc., to determine medically attended from non-medically attended injuries, and to enable the collection of data for any subsequent cost-effectiveness analysis]

• whether the injury was sustained by the index child or by any preschool siblings in the household.

Trial design, funding and approval

The feasibility study was a multicentre, cluster randomised, unblinded trial comparing the First-aid Advice and Safety Training (FAST) parent programme against usual care. The two study centres were Bristol and Nottingham.

The trial was funded by the NIHR HTA programme (reference number 09/02/02) and commenced in January 2011. It was approved by the South West Level 3 Research Ethics Committee (reference number 10/H0106/78) and was registered with the International Standard Randomised Controlled Trial Register (reference number 03605270).

Participants

Recruitment of children’s centres and health visitor teams

Recruitment of children’s centres and health visitor teams began in March 2011 and ended in January 2012. While it was feasible to rank children’s centres on the basis of rate of attendances of preschool children with injuries, it quickly became apparent that the factor most likely to influence the ability of a children’s centre to participate in the study was the engagement of the linked health visitor team. Each children’s centre is linked to a named health visitor team. The degree to which the children’s centre and the health visitor team are linked varies significantly from being based in a shared building and working closely together, to being linked in name only and working independently.

During the period between preparation of the original protocol and commencement of the study, organisational change combined with worsening health visitor capacity issues had resulted in:

1. change of the employer of health visitors in both study centres; neither of the new employers had been involved in initial discussions about the study prior to funding

2. new management structures within health visiting services in both cities in response to workforce capacity difficulties. Previously, health visitors were relatively autonomous professionals. The new structures placed managers as gatekeepers to health visitor teams, and required permission from the manager prior to engagement with the teams themselves. The identification and engagement of managers was difficult and slow in both centres.

In addition, another nationally funded injury prevention study based in children’s centres in both Bristol and Nottingham reduced the number of eligible children’s centres for this study.

The intended model of identification of paired children’s centres and health visitor teams was found not to capture the complex and variable ways in which children’s centres and their health visitor teams worked across the two study centres. In both cities, even though each children’s centre had a theoretical catchment area, parents attended the children’s centre that most appropriately met their needs rather than the one that was closest to them. In addition, children’s centres would not refuse a parent attending simply because they did not live locally. Children’s centres were found to have markedly variable facilities, with some not having the capacity to host both a programme and a crèche concurrently.

The identification of four pairs of linked children’s centres and health visitor teams in each centre was therefore achieved through a process of negotiation and an attempt to achieve a reasonable geographical distribution of participating children’s centres across each city.

Recruitment of parent participants

In our original protocol we anticipated identifying eligible parents through health visitor teams over 4 months. As a result of slower identification than anticipated, recruitment took place over 12 months (between May 2011 and April 2012) and during this period we added alternative strategies to identify the optimal methods for a future trial. Three further methods of identifying eligible families were assessed: emergency department identification using telephone contact, emergency department identification using postal contact, and identification via children’s centres (Table 2). At any one site several strategies ran concurrently. For three of the strategies (health visitor identification, emergency department identification via telephone contact, and children’s centre identification), the names and contact details of parents interested in finding out more about the study were passed to the research team who confirmed eligibility, provided further information and answered questions, and took consent. For the remaining strategy (emergency department identification via postal contact), potentially eligible participants contacted the researchers directly, who then confirmed eligibility and took consent.

Randomisation and allocation

All parents recruited to the study consented to randomisation in the trial. The unit of randomisation was the children’s centre. The original proposal was to stratify the eight children’s centres by study centre (two strata) and randomly allocate within each centre to treatment arm using a remote automated system available through the Bristol Randomised Trials Collaboration based at the University of Bristol. Therefore, two children’s centres in each study centre would be randomly allocated to the intervention arm and two to the control arm. Allocation to intervention or control arms was delayed until after recruitment of families in each children’s centre had been completed to avoid the risk of post-randomisation recruitment bias. It was not be possible to blind parents, children’s centres, health visitor teams or researchers to intervention allocation.

Participation in the trial

In the original protocol we proposed that eligible families be identified by health visitor teams, followed by the researcher contacting the family by telephone to confirm eligibility and answer questions. If interested, families were sent the parent information booklet and consent form. A home visit was arranged to take written consent and collect baseline measures. The parent information booklet stated that the course provided in some children’s centres was a parenting programme containing first aid advice and safety information. While some families took up the offer of a home visit, others declined and we revised our procedure to one of negotiating with each participating family the method to complete the consent form and baseline measures that was most convenient for them: home visit, face-to-face meeting at an agreed location (e.g. the children’s centre), or by post with telephone support from the research team if required. On recruitment to the study, parents were asked to complete a questionnaire to record basic demographic details (e.g. age, sex, ethnic group, marital status, level of educational attainment, age and sex of children in their household). After commencement of the study, two additional questions were added to this initial questionnaire relating to whether or not they had ever previously attended a support course for parents or a first aid course. These questions were added because we recognised that previous participation in courses could influence their decision to participate in the study. A flow chart illustrating the proposed participant experience of the study is shown in Figure 5.

FIGURE 5.

Flow chart of parent participation as per original protocol.

After completion of the baseline measures, parents’ allocation to intervention or control children’s centres was disclosed, and those families in the intervention arm were invited to join the parenting programme. After the parenting programme had completed, a period of 6 months’ follow-up commenced for both intervention and control families. We specified in the original study protocol that during this period parents in both arms of the trial would be asked to complete two 3-month injury calendars in which they could record home injuries occurring to both index children and their preschool siblings. We proposed that a home visit would be offered to all study participants after 3 months to collect the first injury record, to complete a short questionnaire and to give out the second 3-month injury calendar. A further home visit would be offered after 6 months to collect the second injury record and to complete the final study questionnaires. It was believed that home visits would be required to facilitate completion and return of the calendars. After commencing the trial, some parents declined home visits and, therefore, collection of the injury calendars and the interim and final questionnaires were negotiated individually with each participating parent: via home visit, face-to-face meeting at an agreed location, or return by post.

At all stages, parents were offered support to complete the injury records and questionnaires if required. Gift vouchers (£5 value) were given to parents upon completion and return of baseline, 3- and 6-month questionnaires and the injury calendars to thank participants for their time.

A final telephone contact with the families occurred after completion of the final study questionnaires. This telephone contact was from a researcher who the parents had not previously met, to conduct a short interview in order to understand their experiences of participating in the study as a whole.

Training of health visitors and cofacilitators to deliver the intervention

The original protocol stated that the parenting programme would be delivered by a health visitor and cofacilitator from the health visiting team linked to the participating children’s centre. A 2-day training event to enable the facilitators to deliver the programme was designed and provided by the manager from Parenting UK (see Appendix 4). Owing to slow recruitment and the first course starting later than planned, the training was delayed from September 2011 until November 2011 to ensure that the event was relatively close to the first occasion when the facilitators would deliver the programme. The training event took place in Nottingham. Attendees included a trainer from Parenting UK, two members of the research team from Bristol, two members of the research team from Nottingham, the two delivery teams (health visitor plus co–facilitator pair) from Nottingham, and the single delivery team from Bristol. Participants were provided with a training manual, a copy of the parents’ handbook, and had the opportunity to experience the resources available during the programme.

Delivery of intervention

Intervention arm children’s centres hosted the parenting programme and provided crèche facilities. Parents participating in the study who were within the catchment area of these children’s centres were invited to attend the programme and sent details of location, directions and timing of the programme sessions. In addition to the parenting programme, parents in the intervention arm received usual care from their health visitor team following the team’s receipt of the notification of attendance at the emergency centre, minor injuries unit or walk-in centre. Parent participants were given a certificate at the completion of the parenting programme. Details of the parenting programme are given in Chapter 3 and Appendix 2, where the 8-week programme is described.

Children’s centres in the control arm did not host the parenting programme. Parents recruited to the study and living within the catchment area of control arm children’s centres were sent a letter advising them that their local children’s centre was not running a first aid advice and safety course, but that we would like them to help us by completing an injury calendar and some further questionnaires. These families received usual care from their health visitor team. ‘Usual care’ by a health visitor team after receiving a notification of attendance for medical advice following an injury can be considered as one of four different actions: telephone contact, face-to-face contact, referral to services, or no action. The action taken by the health visitor team is determined by the circumstances of the event and the health visitor’s understanding of the needs of that family, and by the workforce capacity of the team.

Outcome measures

Qualitative evaluation of the study

We used qualitative methods to explore the experience of receiving/delivering the intervention, and to explore a range of issues regarding the research methodology for scaling up to a future trial. Interviews, whether by telephone or face-to-face, used topic guides and were digitally recorded where consent was given to do so. An example topic guide is provided in Appendix 9. The content of topic guides was determined following team discussions and based upon previous experience of evaluating interventions with parents, the literature regarding barriers and facilitators to parental engagement in programmes, feedback on the programme from the PAG and the need to determine the strengths and areas for improvement of the programme. Digital recordings were transcribed and analysed thematically. Focus groups also used topic guides and were digitally recorded to validate the notes recorded on a flip chart. These notes were analysed thematically. Digital recordings from focus groups were not formally transcribed.

1. We interviewed a Nursery Nurse and Health Visitor Clerk in Bristol regarding their experience of recruiting to the study.

2. An interview was conducted with one of the research nurses in the Emergency Department in Bristol to determine the factors relating to the experience of attempting to recruit participants through this service.

3. Interviews were held in both study centres with the children’s centre managers regarding their participation in the study, both in recruitment and in hosting the parenting programme (for intervention arm centres). Interviews were either by telephone (Bristol) or face to face (Nottingham), conducted by the research team.

4. We conducted focus groups with parents in the intervention arm upon completion of the course to evaluate their perceptions of participation in the feasibility study and their views on the parenting programme. Focus groups were facilitated by two researchers, with one taking notes on a flip chart. In addition, at the end of each session of the parenting programme the parents participating were asked by the facilitators to record the components of that session that were the most and the least enjoyable and/or helpful for them. These paper records were available during analysis of the focus groups.

5. We held face-to-face interviews with the health visitors and nurses who had delivered the programme in Nottingham. In addition, a joint focus group across both study centres for all the trainers and cofacilitators after all the courses had been delivered was held in Bristol, facilitated by four of the researchers, with one taking notes on a flip chart. The interviews and focus group were used to explore issues regarding the content and delivery of the parenting programme, and to explore issues regarding recruitment and retention of participants for scaling up to a main trial.

At the end of the follow-up period, parents in the intervention and control arms were contacted, where possible, by telephone for a brief interview to collect information on their experience of participation in the study overall, on recruitment, completion of questionnaires and follow-up (including completion of the injury calendar). Telephone interviews were digitally recorded but not formally transcribed and points made by the participants were recorded in a chart to aid reporting using the principles of framework analysis.

Qualitative study outcomes relating to recruitment to, and participation in, the feasibility study are reported in the results section below. The outcomes relating to the parents and deliverers perceptions of the parenting programme are reported in Chapter 7.

Sample size

In both study centres an average of 60–100 children meeting the eligibility criteria for this study are notified to each health visitor team each year. We aimed to recruit a total of eight health visitor teams and up to 96 parents across the two study centres and both treatment arms. This equates to two health visitor teams and 24 parents per treatment arm in each study centre. We anticipated attrition of up to one-third of parent participants between recruitment and commencement of the intervention. As this was a feasibility study, a formal sample size calculation was not undertaken. We considered that collecting outcome data on approximately 32 families in each arm would be sufficient to address our objectives.

Statistical analyses

Data were analysed using Statistical Product and Service Solutions (SPSS; SPSS Inc., Chicago, IL, USA). Descriptive statistics (frequencies, proportions and means) were obtained to describe the characteristics of the participants, and the primary and secondary outcome measures. A rate of injury occurrence per unit period of follow-up in the index children and in their siblings was calculated. The focus of the analytical strategy was on recruitment and retention rates and any difference in the primary outcome to inform a future study. We evaluated the completion of secondary outcome measures by parent participants at each time point. We did not intend to conduct comparative analyses as the study was not powered to detect a difference between groups. Analyses were blind to treatment arm allocation.

User involvement

The PAG met five times during the course of the study with an additional thank-you meeting just after the study had finished. The introductory meeting was facilitated by the children’s centre manager and the mothers read and commented on our information leaflets for parents that were about to be submitted for approval by the Local Research Ethics Committee (SW3 Bristol). They were happy with the leaflets, did not suggest any changes and felt that they understood what people were being asked to take part in. All the mothers felt that the first aid course was a good idea, particularly for first-time mothers. Subsequent meetings discussed the development of the parenting programme, the design of the outcome measures booklet, the injury calendar, publicity materials for use in the children’s centres and the format and running of a main trial.

Telephone survey of children’s centres

During the course of the feasibility study it became apparent that ‘usual care’ for control arm participants may be an insufficient incentive to encourage parents to join the study, and that randomisation to receiving the parenting programme or accessing an alternative course may be more attractive. Through engagement with children’s centres in both study centres, it became apparent that the provision of first aid courses was very common. In order to determine the extent to which children’s centres routinely provided first aid courses for parents, delivered either by staff from the children’s centre or by an external agency, a telephone survey of children’s centres was undertaken in June 2012 in Nottingham and during July and August 2012 in Bristol.

A short telephone survey was designed. The contact details of children’s centres were found by searching websites of local authorities in Bristol, Nottingham City and Nottinghamshire County. A Microsoft Excel (Microsoft Corporation, Redmond, WA, USA) spreadsheet was designed to record the contact details of each children’s centre, the job title of the responder and the survey responses.

Recruitment

Randomisation

In Bristol, two children’s centres were randomly allocated to intervention (n = 17 parents) and two to control (n = 12 parents). As recruitment to the study was slow, effort was directed at improving recruitment in two children’s centre areas first, allowing randomisation of this pair, with randomisation of the second pair of children’s centres 3 months later. As children’s centres had to book rooms for the course for 8 weeks, it was not possible to ask them repeatedly to delay the possible start date of the course because of the slow recruitment to the study. Therefore, a start date had to be agreed, and randomisation had to take place, before 12 families had been recruited in each children’s centre area.

In Nottingham, despite extensive efforts over 5 months, no parents were recruited from one children’s centre area (B), and only two parents from another children’s centre area (S). As it was not viable to randomise these two sites, they had to be withdrawn from the study. The two parents recruited at children’s centres were sent letters advising them that the study was no longer running in their area, but thanking them for their interest and participation to that point. As the primary interest of the study at this stage was to explore whether or not the intervention could be delivered rather than to determine intervention effect sizes, a decision was taken to non-randomly allocate both remaining children’s centres in Nottingham to the intervention arm. This decision allowed the parenting programme to run twice in Nottingham, maximising the opportunities for learning from delivery of two parenting programmes. The decision was taken with the agreement of the chairperson of the Trial Steering Committee. Neither the children’s centres nor parents were aware of their non-random allocation to the intervention arm.

Delivery of the intervention

In Bristol, because there was only one trainer and cofacilitator pair, the parenting programmes needed to run sequentially, as the health visitor and nurse could not be released from existing commitments twice in any one week. The first parenting programme course was run at children’s centre C on a Thursday afternoon, 13.30–15.00, between January and March 2012. Of the 11 mothers recruited at this site, two were not contactable by the time the course was due to start, four were not free on the afternoons when the course was running because of work or other regular commitments, and five stated that they would attend the course. The number of parents attending each session is shown in Table 4. The course was interrupted by two breaks, one when both the trainer and cofacilitator were unable to attend and one due to the half-term holiday.

Participants for the second parenting programme were recruited through children’s centre O. A total of six mothers were recruited prior to the start date of the course, which was held on Monday afternoons, 13.30–15.00 between April and June 2012. One parent worked on the day the course was running and another was not contactable after recruitment, leaving four potential attendees. One parent dropped out because she was getting married and would be spending time abroad during the course, and a second parent dropped out after week 1 of the course, possibly because her English was not strong enough to allow her to participate. The remaining two mothers attended regularly. This course was interrupted by two breaks, one for a bank holiday, the second for half term (Table 5). Two weeks before the course was due to commence the children’s centre manager disclosed that she had released one of the rooms that had been set aside for the crèche to another service provider. As we believed it important that the crèche and the course needed to be hosted in the same venue, the course was held in a nearby community hall already known to the parents attending the course. The health visitor and cofacilitator delivering course 2 had identified some issues with the order of activities during delivery of course 1. Therefore, and with the agreement of the chief investigator, the order of some activities was changed during course 2, while continuing to deliver the same material overall.

Participants completing either course 1 or course 2 in Bristol were given certificates at the end of the programme. All parents in intervention and control arms (including intervention arm parents who had not been able to attend the programme) were sent injury calendars and asked to participate in the follow-up period.

In Nottingham, there were two delivery teams so that the parenting programmes could be delivered concurrently. A decision was taken to stagger the courses by 1 week to facilitate access to the shared box of programme resources. Course 1 was planned in children’s centre T on Tuesday lunchtimes, commencing 21 February 2012, and course 2 was planned in children’s centre C on Monday mornings, commencing 27 February 2012. Despite attempts by telephone, text message and letters to contact participants, none of the six parents recruited to attend course 1 at children’s centre T attended the first session on 21 February 2012, and none of the four parents recruited to attend course 2 at children’s centre C attended the first two sessions on 27 February or 5 March 2012. Of the six recruited parents in course 1, a reason for non-attendance was obtained for four: one had college commitments on the day of the course, one had a family commitment, one was moving house and one was unwell. Of the four parents due to attend course 2, a reason for non-attendance was obtained for one parent who had work commitments on that day.

The inability to deliver the courses to parents recruited to the study in Nottingham allowed the opportunity to explore alternative models of delivery of the intervention. Following discussions between study teams and children’s centres, and with the agreement of the chairperson of the Trial Steering Committee and the programme manager at the NIHR HTA, we attempted to identify new participants for the two courses by opening up the course to any parent attending either of the two children’s centres that were delivering the programme, irrespective of whether or not their child had sustained a medically attended injury in the previous 12 months. Posters were displayed in children’s centres and the children’s centre staff drew attention to the course. Researchers attended the children’s centres to promote the programme. Parents were told that a new course had been developed to help parents keep their child safe from injury and that we were looking for parents to test out the course and tell us what they thought of it. Parents were not formally recruited to the study at this stage but merely had to sign a list to register their interest to attend. Over the course of 1 week at children’s centre T, eight people registered to attend (seven parents plus the teenage son of a parent who was interested in child care), and over 2 weeks at children’s centre C, five people registered to attend (four parents plus the mother of a parent with learning difficulties, who attended with her daughter throughout the course). Hereafter, these parents are referred to as ‘open invite’ parents to distinguish them from those parents formally recruited to the study.

Course 1 at children’s centre T in Nottingham recommenced on 28 February 2012 and ran for 8 weeks until 1 May 2012, with a 2-week break for the Easter holiday. The course was delivered on Tuesday lunchtimes, 13.00–14.50 (90 minutes + 20-minute break). A sandwich lunch was offered from 12.30 to 13.00 to encourage attendance. The eight ‘open invite’ participants attended all eight sessions. One of the parents originally recruited to the study attended weeks 1 and 2 before dropping out of the course after she broke her arm (Table 6). Despite agreement to run the course at children’s centre T, this arrangement turned out not to be convenient for the children’s centre staff as they lost access to a room they normally used at lunchtime, and therefore after week 3 the venue was moved to a community centre a short walk away. This change of venue did not result in any loss of attendance by course participants. The health visitor delivering the course collected feedback from participants at the end of each week.

Course 2 at children’s centre C in Nottingham recommenced on 12 March 2012. The health visitor and cofacilitator delivering this course could be released for a total of only 8 weeks, and as the first 2 weeks has been lost with no attendance, only 6 weeks remained available. Components of weeks 1 and 2 from the original course, and from weeks 7 and 8, were combined, and the sessions were extended from 90 minutes to 2 hours to try to deliver as many of the components in the original programme as possible. The course was delivered between 10.00 and 12.30 (120 minutes + 30-minute break). Participants were encouraged to attend from 09.30 to allow the programme to start by 10.00. The course was run between 12 March 2012 and 30 April 2012 with a 2-week break for Easter (Table 7). Two of the participants dropped out of the study after the 2-week break at Easter.

At completion of the programmes for ‘open invite’ participants in Nottingham, participants were given certificates of attendance and a free first aid kit (provided at a discounted rate from a major pharmacy company). Ethical approval was obtained (University of the West of England, Bristol) to invite the ‘open invite’ participants who had completed the course to join the study at this point; specifically, to complete a form telling us details about themselves and their family, and to complete an injury calendar for a 3-month period. Of the eight participants completing the course at children’s centre T, seven parents consented to participate, and of the three participants completing the course at children’s centre C, two parents consented to participate.

In summary, our efforts resulted in three groups of participants: (1) those recruited via health visitor teams, via emergency departments or via children’s centres where parents were randomised to intervention or control arms, (2) those recruited via health visitor teams, via emergency departments or via children’s centres where parents were purposefully allocated to the intervention arm, and (3) participants attending the intervention following an open invitation to experience the programme. A diagram illustrating the flow of individual participants through the feasibility study is shown in Figure 6. A modified Consolidated Standards of Reporting Trials (CONSORT) flow chart including clusters and participants is shown in Figure 7.

FIGURE 6.

Flow diagram of individual participants through the feasibility study.

FIGURE 7.

Flow diagram of progress of clusters and individuals (modified CONSORT diagram to illustrate different recruitment strategies).

Retention and completion of measures

As identified in our protocol both recruitment and retention to the feasibility study were anticipated to be challenging. A summary of the retention and completion of measures for 51 participants is presented in Table 8. Data are shown for participants according to the three groups previously described: those randomised to intervention or control arms, those purposefully allocated to the intervention arm and those experiencing the intervention following an ‘open invite’.

Of 39 participants allocated to the intervention arm [randomised (n = 17), purposefully allocated (n = 9) and ‘open invite’ (n = 11)], only 15 completed the intervention, although completion was significantly higher (Fisher’s exact test, p = 0.002) among those recruited using the ‘open invite’ approach (85%) than using the other two approaches (31%).

Of the 17 parents randomised to the intervention arm, one withdrew from the study. Of the nine parents purposefully allocated to receive the intervention, two withdrew from the study after the first baseline visit. The remaining seven parents were sent 6-month follow-up questionnaires even though they had failed to attend the course. One of these returned their questionnaire. Reminder letters and a further questionnaire were sent but no further replies were received. These seven parents were not asked to complete injury calendars.

Of the 13 participants in the ‘open invite’ group, one was the mother of a parent with learning difficulties and one was the teenage son of a parent, who was interested in a career in child care. Of the 11 ‘open invite’ parent participants, nine subsequently consented to provide complete questionnaires and a 3-month injury diary.

The data show that approximately two-thirds of those sent injury calendars returned partial or complete calendar pages, suggesting that this new measure to record parent reported injuries in the home was relatively well received. Our early concern, that parents may be reluctant to report injuries occurring to their children for fear of being judged, does not appear to have been supported by this response rate. It was possible to obtain emergency department, NHS walk-in centre and general practice records to validate the parents’ reports of NHS service use for injuries for all of the participants who commenced follow-up.

Characteristics of participants in the study

Table 9 shows the characteristics of participants receiving the intervention and those in the control arm. The 26 parents either randomised or offered the intervention had 31 preschool children between them. The 12 parents randomised to the control arm had 18 preschool children between them. The data show that despite the problems with recruitment we were able to enlist parents, mostly mothers, across a range of ages, and with variable ethnicity, marital status and highest educational level. The index children of the parents recruited to the study were predominantly boys, reflecting the increased prevalence of home injuries in this group. None of the index children was below 3 months of age at the time the parents were recruited to the study.

After recruitment had commenced, we added two questions to our baseline questionnaire for participants: one asking whether or not the parent had previously attended a first aid course, and a second question asking whether or not the parent had previously attended a support course for parents. The rationale for adding these questions was that previous attendance at such courses could have influenced their decision to attend the FAST parent programme. Five parents recruited to the control arm, seven parents recruited to the intervention arm and all nine of the ‘open invite’ parents consenting to participate in the study were asked these questions. Attendance at first aid courses was common: four out of five parents in the control arm, five out of seven parents in the intervention arm and three out of nine parents in the ‘open invite’ arm. Previous attendance at a parent support programme was less common: none of five parents in the control arm, one of seven parents in intervention arm and three of nine parents in the ‘open invite’ arm.

Primary outcome

The primary outcome was the rate of parent-reported medically attended injuries to the index child or their preschool siblings occurring in a home setting. The injury calendar was used to collect this measure. The number of injury calendars sent to parents is summarised in Table 10. Six-month injury calendars for 16 index children and two preschool siblings were sent to 16 parents randomised to receive the intervention (108 child-months) and 3-month calendars were sent for nine index children and three preschool siblings of the nine ‘open invite’ parents (36 child-months). Injury calendars were not sent to the parents purposefully allocated to the intervention arm. Therefore, calendars covering 144 child-months were sent to parents offered the intervention. Six-month injury calendars were sent for 12 index children and six preschool siblings of 12 parents randomised to the control arm (108 child-months).

The number of injury calendars returned by parents, by study group, is summarised in Table 11. Calendars were returned for 60 child-months from 10 parents (10 index children + three siblings) randomised to intervention, and for 23 child-months from six ‘open invite’ parents (six index children + two siblings). Of these 83 child-months of observation from parents either randomised to intervention or in the ‘open invite’ group, 41 child-months were from parents who had completed the parenting programme. Calendars were returned for 62 child-months from six parents randomised to the control arm (six index children + five siblings). Parents who returned pages from the injury calendars did not always return all of the pages they had been sent.

In total, 151 injuries were reported by the parents using the injury calendars. More injuries were reported by parents in the intervention arm of the study. Of the 25 parents eligible to receive the intervention (either randomised or ‘open invite’), who between them had 30 preschool children, 16 parents (64%) reported a total of 120 injuries in 21 index children or their preschool siblings. Of 12 parents randomised to the control arm, who between them had 18 children, seven parents (58%) reported a total of 31 injuries in 12 index children or their siblings. Participation in the intervention arm may have heightened awareness of injuries, which may have increased the risk of enhanced recall bias in this group.

This study was not powered to detect a difference in effect between groups and therefore we have not conducted an intention-to-treat analysis. An analysis of those receiving the intervention compared with the control group has been undertaken. The rate of any child injury reported by nine parents who completed the parenting programme and returned injury calendars (either randomised or ‘open invite’ parents) was 1.7 injuries per child-month (69 injuries over 41 child-months), compared with 0.5 injuries per child-month (31 injuries over 62 child-months) in the children of six parents in the control arm. We defined injuries as those causing a mark that lasted longer than 1 hour, or an ingestion event. However, visual analysis of the free text provided by parents describing the injuries suggests that parents who attended the parenting programme were more likely to report more minor injuries than those in the control arm. For example, one parent reported 24 injuries in her three preschool children over 3 months, while others reported injuries, such as ‘banged knee’ or ‘bumped head’, that required no action other than a ‘cuddle’. While this may illustrate an appropriate heightened awareness of injury following attendance at the parenting programme, it is possible that some injuries reported by parents on the injury calendars did not meet our definition of an injury. These findings suggest that amendments to the instructions on the calendar and the way the calendar is introduced to parents would be appropriate for a future trial to ensure that parents in both arms know when and when not to record an injury. It also suggests that a more objective outcome measure is needed rather than relying solely on parental report.

Heightened parental concern and major injuries were more likely to involve contact with NHS providers, and so these data were also captured in pursuit of an objective outcome. In the nine parents who completed the parenting programme (either randomised or ‘open invite’) and returned injury calendars, there was one injury reported that required NHS provider use over 41 child-months, that is to say a rate of 0.024 injuries per child-month or 0.024 episodes of NHS provider use per child-month. The single episode of NHS provider use was a telephone call to emergency services (‘phoned 999’) but the injury did not require any further NHS provider care. In comparison, in the seven parents randomised to the control arm who returned injury calendars, there was one injury reported that required four episodes of NHS provider use over 62 child-months, that is to say a rate of 0.016 injuries per child-month or 0.065 episodes of NHS provider use/child-month. These four episodes of NHS provider use were one episode each of ‘phoned NHS Direct’, ‘phoned GP’, ‘visited NHS walk-in centre’, and ‘visited emergency department’. One parent in the intervention arm who was unable to attend the parenting programme owing to work commitments but who did complete an injury diary for the 6-month follow-up period reported that her child broke their arm, requiring a visit to the hospital.

In order to illustrate the range of injuries reported by parents using the injury calendars, the number of injury events and the number of children injured in both intervention and control arms of the feasibility study are reported in Table 12. No comparisons between groups have been undertaken owing to the small samples in this feasibility study.

Parent-reported injury outcomes from the injury calendars were validated against emergency department, NHS walk-in centre and general practice records where available. The injury calendars for three children reported three injuries that resulted in NHS service use. Two of these were validated against emergency department, NHS walk-in centre or general practice records. The third was a report of dialling 999 after a child fell off his bike and banged his head. The parent did not report subsequent emergency department attendance. Our methods did not attempt to validate calls to the emergency services, but there was no emergency department record of an attendance for this child subsequent to the 999 call. A review of the records for index children and preschool siblings obtained from emergency departments (n = 61), NHS walk-in centres (n = 61) and general practices (n = 58) identified 12 episodes of NHS provider use. Two of these were for foreign bodies in the nose. As we had not asked parents to report objects up the nose or in the ears, these two have been excluded from further analysis. Of the remaining 10 events, two were reported by parents on injury calendars, four were not reported on injury calendars, and four were in children for whom no injury calendar pages were returned (one of these injuries was for a burn to the hand of a toddler on hair straighteners. The general practice record indicated that this injury had been treated in a burns unit, suggesting that this could have been an injury with significant consequences for the child). Table 13 summarises these data, indicating a sensitivity of the injury calendars of 20% and a specificity of 99%.

Secondary outcomes

The secondary outcomes reported by the parents randomised to intervention or control arms are reported in Table 14. Data have been reported as the proportion of questionnaires completed, frequency counts and scores at baseline and follow-up. Differences between groups should not be overinterpreted owing to the small numbers of participants for whom data are available, and statistical comparisons between groups have not been reported. These data have been included only to illustrate that the recording of this battery of secondary outcomes is feasible. Parental well-being was assessed using the Warwick–Edinburgh Mental Well-being Scale,101 and only one participant at follow-up invalidated the questionnaire by leaving one element unanswered. Mean scores fell slightly in both intervention and control groups at follow-up compared with baseline. Parenting supervision, assessed using the PSAPQ,102 was well completed by those returning questionnaires. Scores for the protectiveness and belief in fate subscales of the PSAPQ increased in a positive direction at the end of follow-up compared with baseline for parents in the intervention arm. Scores for supervision at follow-up were broadly static and those for risk tolerance increased compared with baseline for parents in the intervention arm.

The Strengths and Difficulties Questionnaire95,96 was used to assess child behaviour for children over 2 years; however, the use of different measures below the age of 2 years was problematic as parents reported developmental change between baseline and the end of follow-up, which meant that individual questions asked during infancy at baseline were no longer appropriate by the end of follow-up when the child was a toddler. The Strengths and Difficulties Questionnaire95,96 has not been validated for use in children below 2 years. For a future trial, it would be necessary to identify an alternative single measure that can assess child behaviour appropriate to the child’s developmental age between birth and 5 years, yet still provide a pre-test–post-test change score.

All of the participants returning questionnaires at baseline and follow-up were able to complete the questions relating to first aid knowledge and home safety practices and behaviours. Minor changes to wording and question structure for these questionnaires would be required for a future trial to avoid potential misinterpretation (though no evidence of misinterpretation was identified in questionnaires returned during the feasibility study).

Qualitative outcomes

This section describes the results of the qualitative interviews with parents regarding their recruitment to, and participation in, the feasibility study. Parents’ comments relating to their perception of the parenting programme, together with the perspectives of those delivering the course, are detailed in Chapter 7.

Telephone survey of first aid courses

Twenty-seven children’s centres were identified in Bristol and 84 in Nottingham City and County. In both Bristol and Nottingham some children’s centres worked in partnership with neighbouring centres, so that programmes and courses running at one centre were available to parents attending any of the centres in that cluster. During the telephone survey, if a respondent was responsible for more than one children’s centre, we ensured that their responses applied to all children’s centres under their responsibility. In Bristol, 34 children’s centres worked as 25 individual or clustered centres. We obtained information on 25/25 (100%) centres or clusters in Bristol and on 64/81 (73%) of centres across Nottingham City and Nottingham County areas.

Where information was available, 19/25 (76%) of children’s centres in Bristol and 39/64 (61%) in Nottingham were able to offer first aid courses to parents. Two of the six centres in Bristol not running courses had done so in the past and 21 of the 25 children’s centres in Nottingham (84%) that did not run courses said they would like to host them. The main barriers to hosting courses were lack of capacity in terms of staff or space, or lack of funding.

Across both study centres courses were most commonly run once or twice per year (range 1–6), typically comprising six sessions (range 1–7), with sessions most commonly lasting 2 hours (range 1.5 hours to ‘all day’). Providers varied but were most commonly from the local college or a local authority approved trainer. Voluntary sector providers such as St John’s Ambulance and the Red Cross provided the majority of the remaining courses. Courses were most commonly held during the daytime, with only three centres in Bristol and eight in Nottingham reporting evening or weekend sessions. Courses were most commonly able to host up to 12 parents but often ran with fewer participants and targeted groups of disadvantaged families, adults with few qualifications or parents of children under 1. Full details are provided in Table 15.

Summary of findings

We conducted a feasibility trial of a new parenting programme designed to help reduce the occurrence of home injuries in preschool children. We sought to recruit 96 parent participants with a view to 64 participating in a two-arm trial of the FAST parent programme versus usual care. We tested five methods of parent participant identification and discovered the limitations of each. We were able to recruit 40 parents and, of these, 15 were retained through to the end of a 6-month follow-up period. We also engaged 11 ‘open invite’ parents, of whom nine were retained through to the end of a 3-month follow-up period. We delivered the parenting programme in four settings across the two study centres. We identified issues relating to study design, participant identification, recruitment, retention, intervention delivery and evaluation to inform a future trial. No serious or adverse events were reported.

Strengths and limitations of the feasibility study

Despite the challenges experienced during recruitment of both health visitor teams and parent participants to the feasibility study, we have been able to complete this project on time and within budget. Recruitment to the study took longer than anticipated and the sample originally proposed was not achieved. We carefully monitored recruitment and acted promptly when strategies appeared to be unsuccessful. Multiple strategies were used to determine the most appropriate method of identifying potentially eligible participants and recruiting them. This learning can inform the design of any future trial. The different strategies were implemented with appropriate ethical approval for each amendment to the study protocol, ensuring that the study was conducted with robust research governance. These changes were supported by an experienced trial steering committee that provided constructive advice and guidance. Despite the challenges with recruitment to the feasibility study we were able to run four courses of the parenting programme as proposed, and the evaluation of the delivery provided valuable lessons for the development of both the parenting programme and any future trial.

We used a mixed methods approach to evaluate the feasibility trial, including the development and testing of a new measure for recording parent reported injuries in the home. We also conducted a preliminary validation of the tool against emergency department, NHS walk-in centre and primary care records. Our feasibility trial showed that the proposed quantitative outcomes could be collected in a future trial; for example, 32 out of 48 (67%) partial or complete injury calendars for preschool children were returned by parents. We also used qualitative techniques to explore the experience of participation in the feasibility study, and capture the experience of the parenting programme from multiple perspectives. These outcomes of the feasibility study are able to support the rationale for the conduct of a future trial.

One notable strength of our feasibility study has been the involvement of parent advisors, who have significantly informed our decision-making. We established a PAG that provided guidance on new strategies for identification and recruitment of parents, on materials used during recruitment (e.g. posters), on documentation used in the trial (e.g. parent information sheets) and on the interpretation of outputs from the study. Parent advisors have attended the Trial Steering Committee meetings supported by the PAG facilitators. Our experience of user participation in this feasibility trial has become one of the case studies used in a NIHR study on Public Involvement in Research funded by the Health Services and Delivery Research Programme (reference: 10/2001/41).

The primary limitation of this feasibility study is the difficulty experienced in recruiting the planned number of parents to the study, despite the different identification strategies used. The main reason was the eligibility requirement to recruit parents with preschool children who had already sustained an injury where medical attention had been sought from an emergency department, NHS walk-in centre or minor injuries unit. We were concerned that this criterion would act as a barrier to recruitment and this appears to have been the case. Parents are naturally reluctant to admit that their child has sustained an injury for fear of being labelled as an inadequate parent. This fact compounds the issue that not all parents wish to attend parenting programmes. The health visitor teams and the children’s centre staff that we were working with advised that the need to admit that their child had sustained an injury was the primary reason that parents were reluctant to engage. The programme facilitators running the open access programme reported that they were pleased that the open access method had recruited the families they were seeking; several parents made comments such as ‘oh, that happened to [my child]’ during the course, suggesting that the programme had in fact recruited families where children had sustained injuries. The PAG recommended that the programme should be made available to all parents, ideally before a child has sustained any injuries.

A second limitation to the study has been the challenge of working with health visiting teams, both in asking them to help identify potentially eligible participants in the trial and also as a source of programme leaders to act as deliverers of the intervention. We had consulted extensively with health visitors before submitting our proposal, and one of our co-applicants is a practising health visitor. However, we could not have anticipated the extent to which NHS reorganisation, changes in health visitor managers and reduced workforce capacity would reduce the ability to identify teams that had the interest and capacity to work with us on the feasibility study. The challenge of recruiting parents to the study and the limited capacity of health visitor teams in Nottingham were key contributory factors to the necessity to withdraw two children’s centres where the research team had been unable to recruit sufficient parents to enable randomisation to intervention and control arms. For this reason, in Nottingham, both remaining children’s centres were purposefully allocated to the intervention arm to enable two courses of the parenting programme to run in that study centre.

Where parents had been recruited using the original eligibility criteria, the numbers of parents commencing the parenting programmes was small (two at one setting and three at a second setting). One of the main reasons for this was the long delay between recruitment and randomisation to intervention or control arms (up to 5 months) and to commencement of the course, due to our extended efforts to recruit up to 12 parents to each programme. During this delay some parents returned to work or took up other commitments at the same time as when the programme was due to run. A shorter duration between opening recruitment and commencing the programme would be required in any future trial. Of the five parents who were recruited to the original eligibility criteria and who commenced the programme, four (80%) completed the course. In Nottingham, where we tested the feasibility of ‘open invite’ access to the program, there was a very short delay between invitation and commencement of the programme (maximum 2 weeks) and much higher numbers of parents commenced the programmes (nine in one setting and five in a second setting). Of these 14 parents, 11 (79%) completed the course.

The implications of these strengths and limitations for future research have been described in Chapter 8.

Delivery of the programme

The format of the course as a regular commitment over a number of weeks was positively regarded by the majority of participants as it allowed time for information to be absorbed and gave them

. . . time to try things out.

Parent, Bristol course 1

The length of the sessions, at around 1.5 hours, was considered appropriate. A regular weekly format appeared preferable over a shorter more intense course:

You can’t take it all in in one day.

Parent, Nottingham course 2

A shorter, day-long course had previously been proposed by two Bristol parents who later withdrew after concerns over the time commitment. The status of the programme deliverer was important to the parents. The fact that the trainers had a health background provided the parents with confidence that they were not merely being delivered a pre-set content but that the trainers could add context and background. Participants appreciated a friendly, chatty style of delivery from the trainers. Four Bristol participants specifically mentioned locating the course in a local venue that was on a bus route, that was within walking distance for some and that had parking facilities, and reflected that this was all part of the thoughtfulness that had been a part of the course design. Parents who were not used to attending a children’s centre raised some concerns about using this as a venue; they did not know the staff there and were not confident in the crèche provision. Parents in Nottingham specifically raised the concern regarding children’s centres being perceived as locations where social services support families.

Course content and materials

Parents appeared comfortable with the combination of support for parenting, child development, first aid advice and safety. The majority of parents reflected that more first aid advice would have been helpful:

I thought it was all going to be first aid with a bit of parenting but it’s much more parenting with a bit of first aid.

Parent, Nottingham course 1

Several parents had previously attended courses for their jobs or for life skills, including courses on health and safety and a St John’s Ambulance Course. None had been offered this type of combined course previously. The parents who completed the course held it in positive regard:

. . . it was really interesting and very helpful.

Parent, Bristol course 2

[I] really enjoyed the course.

Parent, Nottingham course 2

The parenting guidance appeared to make sense with the safety advice. A small group of parents praised the quality and professionalism of the course materials:

. . . [it was] really well organised . . . felt really professional . . . very luxurious.

Parent, Bristol course 2

The parent handouts to take home each week were well received with a ‘nice folder’ (parent, Bristol course 2) which was ‘informative’ and ‘helpful (parent, Nottingham course 2).

I think it’s something I’ll refer to at other points, if something happens at least it’s something you can look back to.

Parent, Nottingham course 2

All groups commented that the activity cards caused confusion, especially if they were muddled up:

Every single time we had something with the cards, there was an awful lot of them.

Parent, Bristol course 1

The sessions with cards – there are some humungous amounts of cards.

Parent, Nottingham course 1

Three of the four groups identified that they would like a follow-on course which might support them as their children grew older; other requests for further information included advice about water hazards in the home, bandaging, what items they should have in a first aid box and checking expiry dates on their medicines and tablets, which suggested that they had considered home safety beyond the specific content of the parenting programme.

Group experience

A positive and shared group experience supported engagement with the programme. Development of group cohesion was facilitated by the social elements (e.g. provision of refreshments) and these were regarded as follows:

. . . a bit of a time out.

Parent, Bristol course 2

The chat made it what it is.

Parent, Nottingham course 2

Provision of a free crèche was regarded as

. . . incredible.

Parent, Nottingham course 1

Repeated attendance with the same people provided time for confidence to develop between group participants and the feeling that this was a group where it was safe and supportive. Parents responded to this commitment and felt able to

. . . talk freely [about their concerns].

Parent, Bristol course 1

Another felt that

. . . you can feel more comfortable with what you are saying.

Parent, Nottingham course 2

The programme became a social event with calls being made between participants in both Bristol and Nottingham to establish if other participants were attending. However, there was concern that too many people attending the course might have a negative impact:

. . . you would struggle to have the discussions.

Parent, Bristol course 1

Some parents felt that a group of up to eight parents would be appropriate. In Nottingham, one ‘open invite’ group had provided the option for participants to attend with a friend, an opportunity which was a boon for those who were less confident about attending alone.

Learning new knowledge and skills

Parents were enthusiastic about the opportunity to learn new skills and knowledge. They specified as particularly useful the parenting elements such as understanding the emotional arousal of the brain, how to communicate positively and appropriately with children (Bristol parents, both courses), and looking at a situation from a child’s point of view. Several parents talked of benefiting from the discussions on calm parenting, taking a more measured response to an event and positive praise. Parents also highlighted as important the first aid and safety knowledge, particularly watching the fire safety film, developing a fire escape route, wrapping a burn in cling film, and learning not to panic in an emergency. This last point was particularly important for one mother in Bristol; prior to the course she would have called her mother for help but now reported feeling more confident to manage situations herself. Generally, parents appeared to appreciate take-home messages about increased parental awareness and assessing potential dangers to their children:

I thought it useful in every sort of way . . . I thought it was really helpful.

Parent, Nottingham course 1

Parents reported taking specific actions to improve safety in the home as a result of participating in the course, including adding socket covers, putting locks on cupboards, putting hair straighteners away and installing baby-gates.

Focus

Health visitors and nurses were concerned that the course was too heavily focused on parenting/behaviour change and not enough on safety awareness, anticipatory guidance and first aid advice. This echoed the perspective of some parents who had expressed concerns that the course was predominantly focused on parenting. The health visitors and nurses perceived the materials to be geared towards the assumption that the parents would be difficult to engage, and would have children who were exhibiting challenging behaviour (as has historically been the remit for parenting programmes). In reality, the parents engaging with the course had existing positive relationships with their children, were willing to learn and had healthy children without behavioural difficulties who had sustained injuries during the course of common play and daily life. One health visitor said:

. . . the starting point seems to be that parenting style is the main driver for accidents rather than developmental stage and environment.

Bristol voice 2

Health visitors were concerned that parents might feel ‘duped’ into having signed up to a first aid advice and safety training course, only to experience what was very obviously a parenting programme. They reported a greater need for ‘everyday parenting’ advice, such as how decisions should be made about what activities are safe for a child to try. Health visitors and nurses offered constructive suggestions for how the focus of the course could be better balanced by changing the order and volume of the content (see Content of each session, below).

Materials and resources

Once on the course, the families were engaged positively by the quality of the materials provided; two items specifically mentioned included the information video about how swiftly fire can take hold in a bedroom and the parents’ handbooks with weekly handouts. Some parents were reported to bring their handbooks regularly: some to write notes, some to show how they had undertaken tasks at home following the sessions. However, the materials that programme deliverers needed to take to and from the venues each week were heavy and bulky, particularly the parents’ handbooks and handouts. A wheeled case was suggested as an aid to transporting the materials and resources. Trainers reported that some safety/first aid advice materials appeared inconsistently provided; for example, there was no parent handout on the management of cuts and wounds but there were handouts for other injury types. The two courses in Bristol ran sequentially while the two in Nottingham ran concurrently, resulting in the single resources kit needing to be transported across the city each week for use by both delivery teams. The colour-change heat mug was broken twice in Nottingham. In future, each delivery team would need its own set of resources. The fire safety DVD was highly rated, though in each location it was a challenge to find a machine on which it could be played. Activities often involved picture cards – these all had the same colour border (lime green) on every pack. It was suggested that each pack could have a different coloured border to enable easy identification when cards become mixed together during activities.

Length of the course

The 8-week course was generally perceived as long, and health visitors and nurses were concerned that this may have been a barrier to recruitment. Each course had to have breaks in the middle owing to school holidays or bank holidays, and there was concern this could affect retention of parents (during one course in Nottingham two parents did not return after a 2-week break for Easter). It was suggested that a 6-week course would fit into one half of a school term. One delivery team in Nottingham had negotiated with their manager to have time out from normal activities in order to deliver the course over 8 weeks. When none of the original families turned up for the first week of the course, and they were forced to spend week two finding ‘new’ parents to attend, and there were only 6 weeks remaining. When they were able to restart the course, they were forced to condense the 8-week programme into 6 weeks. They were able to deliver all of the content by merging weeks 1 and 2 and weeks 7 and 8, and extending each session in between by 15–30 minutes. They managed to cover the content of the course in this manner and did not perceive the shorter duration to be detrimental to parent experience. There was a strong recommendation to deliver the course during the morning and not during afternoons, as parents are fresher then and not worried about collecting children from school.

Content of each session

The health visitors and nurses reported that the volume of material that needed to be covered in each session was very large. While they broadly tried to stick to the times recommended, they found there needed to be flexibility to respond to issues raised in discussion and some flexibility to adapt the materials to the educational level and needs of the participants. It was reported that some weeks felt very ‘top heavy’ with parenting and behaviour change instruction, with the injury prevention and first aid content seen as a disconnected ‘add-on’. Weeks 5 and 6 were highlighted as being particularly problematic in this way. Health visitors were concerned that the focus on parenting inadvertently risked stigmatising the attendees. They therefore suggested that the course could be made less stigmatising by reordering the content such that it was led by an injury topic. For example:

1. Injuries at different ages and stages – e.g. ‘today we are going to talk about burns and scalds’. Activity/discussion then explores different burns and scald risks at different ages/stages of development.

2. How to respond if your child has this injury – advice on appropriate first aid response to burns and scalds.

3. How to prevent this injury – discuss how to keep your child safe from burns and scalds at different ages.

4. Let the safety discussion lead into the parenting issues for that week, e.g. communication with your child, boundary setting, etc.

Each week could be titled with a different injury type, so that the parenting and behaviour change content is presented as supportive and complementary rather than the primary focus.

‘Train the trainer’

Health visitors and nurses reported that the 2-day ‘train the trainer’ event ‘felt amazing on the day’, but when they got round to delivering the course themselves for the first time they felt unprepared and had to undertake a lot of preparation in their own time to acquaint themselves with the materials and activities. There was an unintended long break (up to 4 months) between training and delivering the course owing to poor recruitment, which they reported added to their concerns about being prepared to deliver the sessions. They suggested that in future the ‘train the trainer’ event could allow greater opportunity to witness or attempt key activities or sessions themselves. All teams recognised the importance of trainers having the skills to manage, educate and respond to parents in order to facilitate their learning. Individual trainers perceived that they drew on their previous nursing experience to support parents. The trainers’ manual should be reordered so that everything for each week is in a single section, and that sections have different coloured borders for easy identification.

Location of course delivery

Health visitors and nurses recognised that children’s centres varied considerably, and suggested that community venues might on occasion be preferable to using children’s centres for delivery of the course. Rooms used in children’s centres were sometimes found to be small, hot, untidy or even unsafe. The managers of two children’s centres in Bristol reflected on the challenge of booking one room for the course and a second one for the crèche a long time in advance ahead, with the risk that neither would be used if that centre were randomised to the control arm of the study:

We said we can’t guarantee that as it was such a long way away . . . our programmes change quite quickly, so what we need to do is I think we need to make absolutely firm bookings . . . umm . . . nearer the time.

Bristol children’s centre 5

This issue led to one children’s centre having to give up a booked room owing to competing demands on space:

Our nursery provider needed to expand and increase their places and the only way we could do that was to give them our spare room . . . that was quite a priority for the most vulnerable families.

Bristol children’s centre 5

Therefore, a community venue had to be found at short notice that was able to host both the course and a crèche. In another location, the programme was asked to vacate the children’s centre mid-course and successfully relocated to a community venue. The amount of support offered by children’s centres to the teams during course delivery varied considerably. Two health visitor teams reported concerns that the children’s centre was perceived by parents to be a potentially stigmatising location.

Evaluation by Professor Elizabeth Towner

The parenting programme was evaluated against 16 health promotion criteria. The first 10 of these were adapted from McWhirter’s 10 principles of effective safety education. 108 These 10 principles were distilled from a literature review and are specifically targeted at the school setting and health promotion initiatives in school-aged children. The adaptations included more appropriate wording for the preschool setting; for example ‘Encourage the adoption of, or reinforce, a holistic approach within the wider community’ rather than ‘Encourage the adoption of, or reinforce, a whole school approach, within the wider community’; and ‘Involve parents in real decisions to help them keep their children safe’ rather than ‘Involve children and young people in real decisions to help them stay safe’.

The remaining six criteria were added by Elizabeth Towner, in consultation with Mariana Brussoni and Julie Mytton, to closely reflect the content of a preschool first aid training programme. For each of the 16 criteria, the trainers’ manual and participants’ handbook were examined to see whether or not the content was ‘limited or none’, ‘sufficient’ or ‘well addressed’.

The results of the evaluation are summarised in Appendix 10. Professor Towner concluded that, overall, the FAST parent programme was well structured and combined injury prevention and safety promotion messages with parenting advice. There was a strong emphasis on child development and injuries and this was illustrated using examples of different injury types. Safety was thus taught as part of anticipatory guidance and appropriate supervision was stressed. Particular attention was paid to psychosocial aspects of safety and parents were provided with both the confidence and the skills to anticipate and cope with injury events. The ethos of the programme was non-judgemental and parental contributions were encouraged and valued. There was a range of methods used: small and large groups and practical demonstrations, with an emphasis on practical examples of relevance to the context of the target audience. Professor Towner reported that it was not clear whether or not ‘working in partnership’ was sufficiently stressed, but acknowledged that this may be implicit, rather than explicit. It was also not clear whether or not parents had been involved in the development of the resource from the materials available. One additional concern was the inclusion of a parents’ handout from the resources used in the Solihull Approach parenting programme (www.solihullapproachparenting.com). This did not appear to be integrated into the rest of the package and the rationale for its inclusion was not clear.

Evaluation by Dr Mariana Brussoni

Dr Brussoni evaluated the parenting programme against the evidence statements contained in the Canadian Edition of the Child Safety Good Practice Guide. 109 The degree to which the evidence statement was addressed in programme was considered to be ‘limited or none’, ‘sufficient’, ‘well addressed’ or ‘not applicable’.

The results of the evaluation are summarised in a table in Appendix 11. Dr Brussoni concluded that the strengths of the programme included the curriculum’s language, which was simple, user friendly and engaging, and allowed for flexibility in adjusting for local parents’ interests and needs. In addition, the practical exercises that covered the safety topics were delivered in several different ways, helping to engage participants. Dr Brussoni made a number of specific recommendations:

1. The FAST parent programme curriculum included a question in the quiz on week 6 that touches on the importance of not carrying children in adults’ laps in the car, but there did not appear to be any other point in the curriculum where the importance of child safety seats or the need for children to ride in the back of the car were covered. Because the FAST curriculum was focused on home safety, it may not be appropriate to cover traffic safety. However, as this topic is introduced, it is important to cover it appropriately or consider removing mention of it altogether.

2. Consider discussing pool fencing. While most target parents will not have a pool in their home, there may be one in their housing block or in other venues they visit. It may be important to mention the need for pool fencing so that they are aware of the hazard that lack of pool fencing can be.

3. Mention the importance of supervision for drowning prevention for other bodies of water. It is discussed only in the context of baths. For the same reasons described above regarding pools, it may be important to include this.

4. Consider providing examples of the different types of window safety mechanisms available to prevent falls.

5. Consider providing information on obtaining and installing window safety mechanisms, stair gates and other home safety equipment.

6. Mention the importance of not using pressure-mounted stair gates at the top of stairs.

7. Provide information on fire safety products, such as child-resistant cigarette lighters, hearth gates and self-extinguishing cigarettes.

8. Mention the importance of storing poisons locked up and out of reach.

9. If applicable, provide numbers for poison control centres and inform parents of their purpose.

10. Mention the importance of not using baby walkers, bath seats and other injury hazard-producing equipment.

Key element 1: a clearly specified target population

Results are shown in Table 20.

On basis of questions 1, 3, 4 and 8 (questions 2, 5–7 disregarded as research study), a rating of three was awarded. A score of four was not awarded because additional information should be added to the trainers’ manual where this is currently only included in the commissioning brief or decision log (e.g. eligibility criteria, classification as a targeted prevention programme, ability to match content to intended level of need, and practitioner qualifications).

Key element 2: what is the theory underpinning the programme and how is this reflected in its content and delivery?

Results are shown in Table 21.

An overall rating based on questions 1–8 is 3 for the parenting component of the intervention. Rationale: the programme is based on psychological theories although there is lack of clarity as to how the theories translate into activities. The materials are interesting and engaging but feedback from parents and deliverers suggests that there are minor problems hindering parents’ ability to engage with the programme. It should be noted that this evaluation has been based on the parenting component of the programme as the injury prevention component has been evaluated above using different standards. It should also be noted that this research study does not formally assess mothers’ confidence in preventing injuries, only mothers’ well-being.

Key element 3: training and implementation support

Results are shown in Table 22.

Overall, the rating for this element, based on questions 1–5, would be 2, but is currently zero for questions 6 and 7. Further work would be required with regard to the ‘train the trainer’ programme and mechanisms to ensure intervention fidelity should the study progress to a main trial.

Key element 4: evidence that it works

The tool included seven questions relating to the evaluation of the effectiveness of the programme. These were not considered in this evaluation, as a formal trial of the effectiveness of the intervention has not yet been conducted.

In conclusion, the tool indicated that, from a parenting programme perspective, the eligibility criteria, the theoretical evidence underpinning the programme and the training support package are all reasonably robust, with areas for development acknowledged. It is recognised that the tool has been designed specifically to evaluate parenting programmes intended to support parents with difficulties, children with behavioural problems, or both. References throughout the tool relate to child behaviour being the underpinning reason for the parent engaging in the programme. For this reason, it is not directly applicable to the FAST parenting programme, which considers that neither parents participating in the programme nor their children have ‘problems’. The tool is also designed to evaluate programmes that are already published and available for commissioning. Element 4 of this tool could not be rated because the programme was at an early developmental stage.

Chapter 2, Theoretical basis for the intervention: lessons from two systematic reviews

• Parenting interventions that include home-based, one-to-one, multifaceted components can reduce parent-reported and medically attended child injuries, and appear to improve home safety measures.

• The mechanisms through which parenting programmes may reduce child injury are unclear but may include a generic change in parenting.

• There is no current evidence from RCTs of the effectiveness of solely group-based community-delivered parenting programmes in reducing child injury.

• Participants were interested in joining and completing parenting programmes if they believed that in doing so they would have the opportunity to learn new and specific skills, either for their own personal development or because they believed their skills would support their children.

• The relationship between the participant, the deliverer and the other group members was very important. Participants needed to feel safe both with the deliverer and within the group. A known or trusted deliverer of the programme was helpful, but the deliverer needed to have the skills to present the programme in a non-judgemental, empathic and supportive manner. Participants needed to be able to relate to the other members of the group.

• Practical issues such as the location, frequency and timing of the programme influenced parental engagement. Programmes needed to fit around existing commitments. Incentives such as childcare, travel expenses and refreshments were important.

• Those delivering the programmes emphasised the need to be able to respond to the needs of the group, that is to say to be able to tailor the programme where necessary. This is in potential conflict with the production of manualised programmes that support fidelity of intervention delivery. Deliverer training needs to include group facilitation skills in addition to the ability to deliver the programme materials.

• The potential difference in issues raised by participants and those delivering programmes indicate that both perspectives should be explored when evaluating programmes.

Chapter 3, Development of the parenting programme, and Chapter 7, Evaluation of the parenting programme

• Voluntary sector organisations working with the participant group were informed and valued partners in the intervention development process.

• The involvement of the voluntary sector organisations in the production of the 8-week parenting programme, informed by recommendations made by the PAG and a parents’ forum facilitated by Parenting UK, resulted in an intervention more likely to meet the needs of the participant group than if it had been developed from a theoretical perspective alone.

• Practical issues, such as the need for a wheeled case to transport programme materials and resources into and out of venues, need to be addressed in any future trial.

• Each programme delivery team should have its own set of resources in a future trial. A process to maintain and replace lost and broken equipment would need to be established.

• In a future trial, venues that host programme delivery should be encouraged to facilitate courses during the mornings when parents may be more engaged and less worried about collecting other children from school.

• Future ‘train the trainer’ events need to be timed more closely to the first delivery of the programme for new trainers. At the ‘train the trainer’ event, there should be more time allocated to practising delivery of some of the course materials. Training in group facilitation skills was acknowledged as very important, even for experienced health visitors.

• Children’s centres need to be able to plan their programmes well ahead. Therefore, in a future trial, fixed periods when a course would run need to be agreed in advance and should not be repeatedly put back if numbers of participants available are too small to run the course.

Chapter 4, Development of the injury calendar

• The acceptability of a new parent-reported injury outcome measure was improved by the use of a graphic designer who recommended changes to simplify the information recording process and to increase the familiarity of the design.

• The PAG provided feedback on designs that increased the likelihood that the recording of potentially sensitive information would be completed.

Chapter 5, Feasibility of evaluating the parenting programme through a randomised controlled trial design

• The need to admit that your child has sustained an injury is a significant barrier to recruitment. The faster rate of participant identification through the ‘open invite’ route [we engaged 11 parents (in two settings) over 2 weeks to attend the parenting programme when an ‘open invite’ policy was used, compared with 40 parents recruited over 10 months from four settings when using the criterion of having a child who had sustained a medically attended injury] suggests that rather than target families once an injury has occurred it would be better to target families before the injury event. Significantly more parents identified through the ‘open invite’ route completed the intervention compared with those recruited using the original criteria.

• At this point in time, health visitor teams do not have the capacity to support the identification of potential participants or the delivery of the parenting programme for any future trial.

• Using children’s centres in deprived areas would be an appropriate method to identify families, and children’s centres would be supportive of this approach in a future trial.

• The setting for delivery of the parenting programme should be flexible to respond to local capacity and facilities. The most important factors about the setting are that it is known to parents and easily accessible. Children’s centres may provide an appropriate setting but where necessary alternative community venues should be used.

• A short period of time between identification/recruitment and the start of the programme appeared to be important to avoid attrition of participants before the programme commences.

• Once parents commenced the programme, retention rates to the end of the course were good both for parents recruited against the original eligibility criteria (80%) and for those identified through ‘open access’ (79%), suggesting that the programme was well received by those attending.

• First aid courses are very commonly provided in children’s centre settings and can be considered to be part of usual care provided to users of these settings.

• Approximately two-thirds of parents sent injury calendars returned them, and 151 injuries were recorded on the returned calendar pages, suggesting that the calendar was relatively well accepted as a measure and that parents appeared comfortable reporting injuries in their children as evidenced by the number of injuries reported.

• Not all parents who were sent injury calendars returned them, and therefore checking the records of NHS providers is important in order to identify use of these services during the follow-up period.

• Some of the parents did not report major injuries in the calendar. In a future trial the use of an objective measure of NHS provider use would be necessary, although it would not be possible to validate all NHS provider use such as, for example, following up telephone calls to NHS 111 or ‘999’ calls.

• Many of the injuries reported by parents on the injury calendars appeared minor and may not have met the definition of an injury as provided. Support to complete the injury calendars correctly would be important in a future trial, for example amendments to the instructions on the calendar and to the way the calendar is introduced to parents to help parents in both arms know when and when not to record an injury.

• The list of injury types on the calendar should be reviewed (e.g. the addition of objects getting stuck in the nose and ears).

• An alternative measure for recording child behaviour needs to be identified that will allow a baseline post-test comparison even in young infants in whom significant developmental change between baseline and completion of follow-up would be expected.

Chapter 6, Determination of the parameters for a future cost-effectiveness evaluation of the programme

• It is possible to build a resource use checklist to determine the non-recurrent and recurrent costs associated with delivering this programme as part of a future trial.

• The injury calendar is able to indicate the initial parent-reported NHS provider use costs in intervention and usual care arms of any future trial although amendments to the calendar are required to add transfer by ambulance and allow the use of multiple NHS providers to be recorded.

• Injury management is often associated with a range of NHS provider resources beyond the initial action, and this series of NHS savings relating to an injury event needs to be captured for full costing savings to be estimated in a future trial economic evaluation.

• In a community setting, effective parenting programmes depend upon parental participation in the programme, suggesting that a social perspective on costs that are outside the public sector should be developed to capture these resources.

Study eligibility

The targeted identification and recruitment of parents whose preschool child had sustained an injury resulted in considerable recruitment challenges and is the primary barrier to the success of a trial of the parenting programme intervention. The stigma associated with admitting that your child has sustained a medically attended injury appears to be even greater than we initially suspected. Parents who have attended courses during the feasibility study have reported feeling stigmatised by being invited to join the programme. Children’s centre managers and staff, and health visitor teams, have consistently advised us throughout the study that the best way to engage the families most at risk of child injury (e.g. families with chaotic lives, large families, or families with poor support) would be to provide a universally accessible programme. We therefore conclude that it is not feasible to run a future trial of the FAST parent programme using the existing eligibility criterion that specifies families where a child has sustained an injury.

Recommendation: In a future trial, we would recommend that targeting all parents from a high-risk population would be a more feasible strategy.

Recruitment

Identification of potential parent participants via health visitor teams has been shown not to be feasible owing to their current limited capacity. As we are recommending that a future trial should not target parents of children who have sustained an injury, identification and recruitment via emergency departments would not be appropriate. We have demonstrated through our experience in two children’s centres in Nottingham that making the programme universally accessible to parents was more effective in engaging parents in these two settings than targeting parents of injured children. These ‘open access’ parents were not formally recruited to the trial, but invited to attend a group that was starting the following week. While our experience suggests that the ‘open access’ route is less stigmatising, a future trial would need to evaluate recruitment to a group using these eligibility criteria. Personal contact with the researchers or informed children’s centre staff at the time of identification of eligible parents may help increase recruitment, rather than using posters and leaflets alone.

Recommendation: On the basis that a future trial should target all parents in a high-risk population, we suggest using children’s centres in deprived areas as venues for the identification of potential participants. The engagement of the senior managers of participating children’s centres would therefore be important to the success of the trial.

Randomisation

Recruitment through children’s centres would favour cluster randomisation in a future trial. Discussion with the manager of the children’s centre hosting our PAG confirmed our impression that individual randomisation of parents attending the same children’s centre would be unacceptable to both parents and children’s centre staff. On the basis of equity it would be preferable that all parents recruited at the same time should experience the same arm of the trial. The cluster would therefore be either at the level of the group or the level of the children’s centre. As during the feasibility study we would recommend that, in a multicentre future trial, clusters would be randomly allocated to intervention or control arm via a trials unit using a remote automated system.

Recommendation: Randomisation should be at the level of the children’s centre. Allocation to intervention or control arms would be delayed until after recruitment of parents to avoid post-randomisation recruitment bias.

Intervention delivery: recruitment of delivery teams

Our experience indicates that it is currently not feasible to ask health visitor teams to deliver the intervention alongside their existing workload. The current investment in England in health visitor training has not yet resulted in the capacity within teams to participate with intervention delivery in a future trial. It is possible that by the time of completion of a future trial the workload pressures on health visitor teams may have eased such that should the intervention be shown effective and cost-effective, the wider distribution of the programme could be considered through health visitor teams at that time. Parent participants in this feasibility study reported that they valued the knowledge and experience of the nurses delivering the programme. Therefore, for a future trial we would recommend that nurses be employed by the study to deliver the intervention in each locality. We acknowledge that the potential effectiveness of a programme in a future roll out may not be as great as the effect when evaluated in a trial using dedicated programme deliverers.

Children’s centres we have worked with have indicated an interest in supporting the further evaluation of this intervention. We propose that in a future trial each participating children’s centre would be encouraged to put forward a member of the children’s centre staff to be trained to be the cofacilitator for the nurse trainer. Discussion with the community support manager of the children’s centre hosting our PAG leads us to understand that this would be well received by children’s centre managers as an opportunity to develop skills in members of their staff, without additional cost to the children’s centre. We also anticipate this would encourage children’s centres to be committed to participation in a future trial, and may have the benefit of encouraging retention of parents during the delivery of the course as a familiar member of staff would be involved.

Recommendation: For a future trial, we would recommend that nurses be employed by the study to deliver the intervention in each locality, and each that participating children’s centre provide a staff member to be the cofacilitator for the nurse trainer.

Equipment and facilities

The resource kits (burns dolls, colour-change heat mug, choking tube, resuscitation dolls, fire safety DVD and images for group discussions) were well received by delivery teams and parent participants in the feasibility study.

Recommendation: In a future trial, each delivery team should be equipped with its own set of resources, a wheeled trolley/case to facilitate transport of the resource kit, and a process for replacing broken or lost items.

We recommend that a crèche facility be provided to enable parents to participate in the programme during a main trial, and, therefore, venues need to be identified that have the physical space to provide a room suitable for delivering the course to parents, as well as a nearby room available for a crèche. Where possible, the crèche or nursery facilities at the children’s centre would be used as we know that parents will feel more confident to use a crèche provider and venue known to them. If a children’s centre wished to participate in a main trial but was unable to provide such facilities, its participation would be dependent on the identification of a suitable alternative local community centre as a venue that would be acceptable to parents.

Recommendation: A crèche facility should be provided to enable parents to participate in the programme during a future trial.

Where children’s centres choose to host the programme in their own facilities, an agreement would need to be made to (i) protect the rooms for the programme and not release the rooms to alternative providers, should there be sufficient interest from parents in the course; and (ii) release the rooms for alternative providers should an inadequate number of parents have been recruited to the course by an agreed date.

Outcome measures

If the parenting programme is effective in helping parents to keep their preschool children safe from injuries we would expect a reduction in injury occurrence in all preschool children in that family. The intervention may be more likely to prevent the more serious injuries than minor ones, and may enable parents to feel more confident to manage minor injuries at home rather than seek medical attention. Alternatively, increased knowledge of when to seek medical attention could increase the use of health care. Therefore, it would be necessary to record both medically attended and non-medically attended injuries during a future trial to evaluate the full effects of the intervention. In order to record non-medically attended injuries, a measure to capture parent-reported injuries would be required. The injury calendar was shown in the feasibility study to be an acceptable tool for collecting parent-reported injuries. We have undertaken a limited validation of the tool and found it to be a good record of both health service use and self-care.

Recommendation: The injury calendar developed during this feasibility study should be used as the tool for collecting the injury outcomes in a future trial. Minor amendments to the calendar are required and further validation of the injury calendar against NHS provider records should take place. In addition, NHS provider records will need to be used as a method to collect injury outcomes in a future trial as this feasibility study has indicated that not all medically attended injuries will be recorded on injury calendars.

In order to establish the mechanism through which any injury risk reduction was achieved, it would be necessary to collect a range of measures as secondary outcomes. As established in this feasibility study, measures for parental well-being and parenting, parental knowledge of first aid and parentally reported home safety practices would be appropriate secondary outcomes in a future trial. Some parents reported that the child behaviour questions asked when their child was an infant at entry to the study were no longer appropriate when their child was a toddler at the end of the follow-up period almost 1 year later.

Recommendation: The parent well-being, parenting, first aid knowledge and parent reported home safety practices should be used in a future trial as secondary outcomes. A new measure for child behaviour that enables behaviour change to be recorded over periods of rapid developmental progress will need to be identified for use in any future trial.

Usual care

In our original study protocol we proposed to compare the parenting programme intervention with usual care where this was described as the actions taken by health visitor teams upon receipt of a notification of attendance at an emergency department or NHS walk-in centre for an injury. We discovered that health visitor teams have variable and locally determined usual care, and that most teams are using paper-based records with no central register of receipt of notifications or action taken. Importantly, the practice of usual care varies with time within teams, depending on the workforce capacity and the number of vacant positions in that team. At times teams are required to prioritise only the families in greatest need of support for direct contact.

Recommendation: It is no longer appropriate to record usual care as the action taken by a health visitor team following receipt of a notification of injury in a preschool child because in a future trial we would not recommend an eligible parent to be one whose child has sustained a medically attended injury.

We have identified that it is very common for children’s centres to provide first aid courses for parents in response to parents’ interests and also because children’s centres are encouraged to provide injury prevention as one of a range of health promotion topics. We propose that usual care should be the courses that the children’s centre would usually provide for parents. If the children’s centre were randomised to the intervention arm of a future trial then we would ask that the FAST parent programme be delivered instead of a first aid course if one were due to be given. We recognise that a no-training arm may be unattractive to parents being recruited to a future trial, and therefore where children’s centre would be randomised to the control arm and would not be routinely offering a first aid course, we would provide parents with a booklet of the first aid advice pages from the FAST programme parents’ handbook.

Recommendation: In a future trial, we would recommend that usual care should be the courses that a children’s centre would normally provide if the centre were not participating in a trial. Where a children’s centre would not usually provide a first aid course, we would provide a booklet of first aid advice to reduce the potential for parents to be reluctant to engage in a trial where one option is to receive no training.

Sample size

We were unable to achieve sufficient numbers in terms of recruitment and completion in the feasibility study owing to several difficulties, many of which were anticipated. An aim of this feasibility study was to investigate these problems and identify different approaches to overcome them. The low recruitment and completion rate has meant the number of medically attended injuries reported by parents was too small to make an estimate of the sample size required for a future trial.

Recommendation: A future trial needs to be cluster randomised, and powered around the primary outcome of injuries requiring NHS provider use, data for which should be drawn from published estimates. We recommend that a future trial should be two-staged with a pilot period to formally test the proposed new eligibility criteria and potential for recruitment via children’s centres, and during which it should be confirmed that early data are suggestive of an outcome in the desired direction of a reduction in NHS provider use.

‘Train the trainer’

We developed a 2-day ‘train the trainer’ event for trainers and cofacilitators delivering the programme. We encouraged trainers to seek peer support during delivery and offered telephone support from the programme developer at Parenting UK. The ‘train the trainer’ event was evaluated using feedback forms immediately after the training and through interviews and a focus group with trainers and cofacilitators after the courses had been delivered. In a future trial, regional trainers could be trained to both deliver and cascade training across a geographical area and be responsible for ensuring adequate intervention fidelity, validation of trainer skills, and that session content is always delivered with a sense of optimism and a non-judgemental approach. Consideration would need to be given to turnover of staff through natural change (for both regional trainers and local deliverers). The regional trainers could be responsible for ensuring the maintenance of programme resource kits and would know how to obtain replacement materials and equipment if required. Regional trainers’ ongoing learning needs would need to be supported, for example by refresh training, by teleconference support or by a moderated online forum.

Recommendation: A programme to meet the training and support needs of regional trainers and local deliverers would need to be developed for a future trial to reflect the changes to the intervention from an 8-week to a 6-week programme, to ensure high-quality training of a larger number of trainers and cofacilitators, and to support an intervention fidelity assessment process.

Name of the intervention

Since the funding of the feasibility study, there has been a rapid rise in the prominence and use of a parenting programme that also uses the acronym FAST: the Families And Schools Together parenting support programme.

Recommendation: The First-aid Advice and Safety Training parent programme should be renamed for a future trial to avoid any possible confusion for parents and practitioners with the Families and Schools Together programme.

In this chapter, we have summarised our main findings and reported on the completion of the study objectives and previously specified criteria for success. We have then summarised the lessons learnt from the component parts of the study and brought these together in the section on implications for a future trial.

2.1 Research team roles

Recruitment and quantitative data collection will be conducted by the research fellows (RF) appointed in Bristol and Nottingham, employed specifically for this project. JM will line manage and provide supervision for the RF in Bristol and CM will provide similar support for the RF employed in Nottingham. JM, CM and PB will analyse the quantitative data with the RFs. JI will facilitate the Parent Advisory Group and support the health visitors delivering the programme. JI will conduct the qualitative interviews in Bristol and CM will conduct those in Nottingham. JI will analyse the qualitative data supported by CM and TD. JP will collect and analyse the data for the economic analysis of the programme.

Childhood injuries – the scale of the problem

Unintentional injury is the major cause of death in children over the age of 1 in the UK, and for each child that dies many more will suffer morbidity and possibly long term consequences. Over 2 million visits to accident and emergency departments and over 120000 admissions in children occurred in 2005 due to unintentional injury costing the NHS in the region of £146 million (1). Staying safe has been a fundamental component of child health policies such as Every Child Matters (2). The type and location of child injuries varies with age and the child’s stage of development. The majority of injuries occurring to preschool children occur in the home (3,4,5). Between 2000 and 2002 an average of 502,000 children aged 0–4 years attended hospital every year in the UK due to a home injury, representing 78% of all injuries occurring to children in this age group. The most frequent events leading to injuries in preschool children include, in order; falls, hitting/being hit/crushed by objects, poisoning, and burns/scalds (6). Inequalities in injury occurrence have been widely reported (7–12).

Risk factors for injury

A number of risk factors related to the family and the child have been associated with increased risk of injury. Single parents, step families and teenage parenthood, maternal life events and maternal depression were all associated with increased risk of medically attended child injury by age 2 in the Avon Longitudinal Study of Parents and Children (ALSPAC) (3). In a randomised controlled trial in Nottingham family factors including having a teenage mother or being in a single parent family were associated with increased risk of hospital attended injuries in preschool children (13). Family structure (14) and parental behaviours, such as excessive use of alcohol (15) have been associated with increased injury risk in children. Male sex and difficult behaviour in childhood, particularly that relating to antisocial, aggressive or overactive behaviour, have been associated with increased incidence of unintentional injuries in the UK (16, 17) and in other high income countries (18). Parental understanding of the relationship between injury risk and child behaviour and development is variable, and provision of educational anticipatory guidance has been recommended (19).

Parenting programmes

Parenting programmes are short term interventions to promote changes in the behaviour of parents and children that result in better health and wellbeing outcomes for both. They are usually delivered as face to face programmes, either individually or in groups. Parenting programmes have been increasingly recognised as an intervention to improve the life chances of children due to their effectiveness in reducing antisocial behaviour and improving educational and mental health outcomes in children, and the improved mental health and wellbeing of parents. Low socioeconomic status, unemployment, social exclusion or isolation, young or single parenthood and learning difficulties are known to adversely affect parenting. Consequently, parenting programmes have become a core component of child and family policy (20). Parenting programmes have been developed on the basis of two main theoretical approaches: behavioural and relational. Some programmes combine elements of both approaches. Behavioural approaches aim to develop parents understanding of the negative impact of attention to problem behaviour and lack of attention to positive behaviour, and teach positive discipline practices including praise and time out; relational programmes aim to improve interactions between parent and child, correcting misattributions and increasing understanding of developmental phases. Both have been developed to improve children’s mental health, the former with a particular emphasis on the prevention and treatment of antisocial behaviour and conduct disorder.

Analyses of longitudinal studies have shown the influence of parents on child outcomes that are related to injury risk. Research from the ALSPAC cohort has shown that positive parenting behaviour, parent–child interaction and a stimulating home environment were associated with enhanced development by the age of three (21) and improved cognitive and behavioural outcomes in children by age 5 (22). The ‘better’ the parenting, the more likely children are to be well adjusted and developmentally competent (23). Other studies, for mothers with learning difficulties, have shown that supportive parent training can improve childcare practices (24). Evidence suggests that enhanced carer supervision can help reduce injury risk to children (25,26). Parenting interventions have the potential to reduce poor maternal mental health and increase maternal self efficacy (27,28), to improve maternal–child interactions (29), and to change child behaviour, especially behaviour that is challenging or could place the child at risk of injury (27,30,31). Parenting interventions can reduce injury risk either through these mechanisms or through increased parental knowledge of safety practices (32), improvement in the quality of the home environment (33), or through the use of home safety practices such as having a fitted and functioning smoke alarm, using stair gates or keeping sharp objects safely (34,35). Parenting programmes have shown reductions in injury risk taking behaviour in primary school aged children (36). Health visitor interventions to support parents can reduce injury rates in both prospective studies (37) and meta-analysis of randomised controlled trials (38). Meta-analysis of parenting interventions, primarily conducted in high-risk or disadvantaged families, have demonstrated significantly lower risks of injury, as measured by parental self-report of either medically or non-medically attended injuries (39,40). Parents value programmes that enable the acquisition of knowledge, skills and understanding, and facilitate acceptance and support from other parents. Such outcomes reduce feelings of guilt and social isolation, increase empathy with children, and give confidence to cope with challenging child behaviour (41). A child’s medically attended injury represents a ‘teachable moment’ when parents are receptive to information regarding injury risk in their children (42).

The features of parenting interventions that are most effective are becoming clearer. There is strong evidence that home safety education and the provision of safety equipment are effective in increasing a range of home safety practices (35). A review of ‘what works?’ in parenting interventions has shown that interventions are more likely to be effective if they are delivered early in childhood, if intensity is proportional to need, if they include group activities where parents can benefit from the social aspect of working with peers, if they include formal programmes or manuals to maintain the consistency of the delivery of the intervention which should be delivered by trained staff, and if there is a focus on specific parenting skills and practical ‘take-home’ tips (43). A review of the effectiveness of parenting support programmes in European countries where universal early intervention approaches tend to be used, suggests that positive outcomes can also be achieved when the programme is delivered by non-health workers or agencies (44).

The cost effectiveness of parenting programmes has not been widely studied (45). A recent systematic review of economic evaluations of child and adolescent mental health interventions demonstrated that most evaluations were small scale, had short time horizons for assessing outcomes and had limited reporting (46). However, the cost effectiveness of parenting programmes has been established for group parenting programmes. A formal evaluation of Sure Start parenting programmes demonstrated improved child behaviour outcomes for modest cost and considered the programme value for money (47).

Justification for this proposal

Parenting interventions, usually delivered as part of a programme to improve a range of child and family outcomes, appear to be effective in reducing self-reported or medically attended injuries in young children (48). Due to the range of positive outcomes associated with the programmes, such as improved child behaviour, maternal self-efficacy or maternal–child interactions, it is unclear whether one of these outcomes is more effective in reducing child injury than another. We know that injury prevention education alone has not shown reduction in injury occurrence (35) but the hypothesis that injury prevention education in the context of a parenting programme may be effective remains to be tested. Furthermore, it is unclear whether group based programmes, delivered outside of the home can achieve reductions in injury occurrence similar to intensive one-to-one home based programmes. Evidence suggests that RCTs of home safety education can successfully recruit parents of recently injured children (49,50), and that parents are interested in learning first aid (51). We therefore propose to develop a parenting programme that provides injury prevention education through the delivery of first aid and safety training tailored to the stages of preschool child development and delivered to groups of parents in a community setting. We propose to test the feasibility of delivering that programme with a view to a future large scale randomised controlled trial.

4.1 Aim

To develop and test the feasibility of delivering a children’s centre based parenting programme to prevent recurrent unintentional home injuries in children aged 0–4 years; compared to normal care for such children.

4.2 Objectives

1. to develop a health professional delivered parenting programme

2. to assess the acceptability of the parenting programme to parents and professionals

3. to assess the feasibility of delivering the parenting programme

   1. to assess recruitment and retention of parents within the trial

   2. to assess compliance with the intervention during the follow up period

   3. to determine the training, equipment and facilities needed for delivery of the parenting programme

   4. to assess the collection of primary and secondary outcome measures

   5. to determine which information to collect on ‘normal care’

   6. to assess which relevant resource utilisation/costing data needs to be collected

   7. to produce estimates of effect sizes to inform sample size estimation for the full trial

5.1 Experimental group

The experimental arm of the trial will be ‘normal care’ plus a parenting programme. The parenting programme will be developed in collaboration with Parenting UK (a parenting programme development organisation). Engaging parents in a parenting programme following injury in their child may be difficult since the injury may result in feelings of stigmatisation, guilt or concern that the injury is believed to be intentional. Our programme will contain both home safety education (‘keeping your child safe’) and first aid training as methods of primary and tertiary injury prevention. Unpublished evidence from local injury prevention projects in Gateshead and Bristol suggest that parents are interested in learning first aid and willing to attend a group to do so. The emphasis on first aid and safety training is hoped to enhance acceptability and diffuse any negative feelings generated by being identified for participation in the programme. Interest in learning first aid is hoped to stimulate further interest in injury prevention and parenting to reduce injury risk. The programme will be delivered in children’s centres by Health Visitor teams, to groups of parents.

The parenting programme is likely to contain elements of existing parenting programmes that may reduce injury risk, for example, those intended to enhance parental self-efficacy and well being, improve parent/child communication, and improve child behaviour through the increased use of positive reinforcement, and the enablement of setting and maintaining boundaries (52). Evidence-based safety components are likely to include home safety education, assessment of home hazards, guidance on types, sources and fitting of home safety equipment and tailoring advice to both the home context and understanding how a child’s injury risks change as the child grows and develops (anticipatory guidance) (53).

Families in the experimental group will be invited to participate in 1 to 1 interviews after delivery of the parenting programme to explore parents’ views and experience of the programme.

5.2 Control Group

The control group will receive ‘normal care’. In both Bristol and Nottingham, Health Visitors are routinely sent details of children within their geographical area of responsibility that have sustained a medically attended injury, whether at A&E, or NHS Walk-in Centre. Usually the notification is sent by fax either on the same day as the injury or on the next working day. It is usual practice that following receipt of such notification the Health Visitor Team may take a range of actions determined by the circumstances of the event and the Health Visitor’s understanding of the needs of that family. Four different actions can be considered as part of ‘normal care’; telephone contact, face-to-face contact, referral to services, or no action. The use of each of these four possibilities will vary between locations and between Health Visitor Teams. For example, some Health Visitors may take every fax notification as an opportunity to make contact with a family, especially in deprived or multicultural communities, where parent initiated contact with Health Visitors is low. In other areas, workload or knowledge of families will mean that Health Visitors are much more selective of those families that they will contact.

5.3 Inclusion and exclusion criteria

5.4 Randomisation and allocation

The unit of randomisation will be the children’s centre . We will recruit a total of 8 children’s centres; four in Bristol and four in Nottingham. Those agreeing to participate will be stratified by study centre (two strata) and randomly allocated within strata to treatment arm using a remote automated system available through the Bristol Randomised Trials Collaboration (BRTC) based at the University of Bristol. Two children’s centres in each study centre will be randomly allocated to the intervention arm, and two children’s centres in each study centre will be allocated to the control or ‘normal care’ arm.

To reduce post-randomisation recruitment bias, informing the children’s centres and Health Visitor Teams of their allocation to intervention or control arms will be delayed until after recruitment of families in each children’s centre has been completed. If allocation to intervention or control arm is indicated to the Health Visitor Team/children’s centre at the time of randomisation, we anticipate two potential post randomisation recruitment biases; a) the Health Visitor Team could choose not to offer entry to the study if the team knew that the family were unlikely to participate or continue in the programme once commenced, and b) a family may be influenced in their decision to participate if they knew in advance that their children’s centre was, or was not, offering the FAST Parents programme. Therefore once recruitment has been completed Health Visiting Teams and children’s centres will be informed of their allocation. Health Visitors and Nursery Nurses from intervention children’s centres will be trained in delivery of the FAST Parents programme.

6.1 Ethics committee approval

Approval will be sought from a Type 3 committee; South West 3 REC (Bristol Central) based at University Hospitals Bristol.

6.2 Participant consent

Parents asked to participate in this research are entitled to choose whether or not to take part. Their decision will be voluntary and they will be competent to understand what is involved. Consent forms will be designed to assure the protection of their rights.

In the eight children’s centre areas participating in the study receipt of a notification of a medically attended injury by the Health Visitor Team will result in the team making a decision on ‘normal care’. The families will be contacted after completion of ‘normal care’ to advise them that their local children’s centre is participating in a study to follow up pre-school children who have had an injury and that some children’s centres will be offering first aid courses for parents. Parents will be asked if their details can be passed to the research team who will tell them more about the study. Refusal to participate in the programme will not prevent access to any other routinely available services, although will be recorded as an outcome of the feasibility study along with the reason for refusal where this is provided. The Health Visitor Teams will be given a list of the inclusion and exclusion criteria to enable them to determine eligibility of families for participating in the study.

Families that agree to be contacted by the research team will receive both written and verbal information. The research fellow will send a study information sheet and consent form to the family. One week later the researcher will contact the family and ask permission to visit. If the family agrees the researcher will send baseline measure questionnaires prior to the visit.

Parents living within the catchment areas of intervention children’s centres will be invited to participate in the FAST Parent programme. At the visit the researcher will verbally explain the study. The explanation will cover all the elements specified in the written information provided for the participant. The participants will be informed of the aims, methods and participation requirements of the study. They will be informed that the programme is intended to help reduce the risk of injuries in children, but because this is a new programme that has not previously been tested the likelihood of success is unclear. The researcher will explain that the programme will provide the opportunity to learn first aid and safety information in a ‘hands-on’ and interactive manner. Parents will be informed that we are interested in their experience of participating in the programme to help us understand how the programme could be improved. Parents will be advised that there are no anticipated risks to attending the programme.

Parents living within the catchment areas of control children’s centres will be invited to participate in an injury follow-up study, where the researchers are interested in children’s behaviour and whether a child who has had one injury has any further injuries. During the visit the researcher will verbally explain the study. The explanation will cover all the elements specified in the written information provided for the participant. The participants will be informed of the aims, methods and participation requirements of the study.

The participant will be given every opportunity to clarify any points they do not understand and if necessary ask for more information. At the end of the discussion the participant will be given time to reflect. The participant will be informed that they are at liberty to withdraw their consent to participate at any time, without prejudicing any future medical care.

The researcher will obtain the participants freely given written consent before participating in the study. The consent form will assure the participant of the confidentiality of the data collected. Participants attending children’s centres in the experimental arm of the study will be asked permission that the interviews conducted after the parenting programme can be audio recorded and to publish anonymised quotations from the study. With the agreement of the participants, the data will be anonymised and stored in accordance with data protection guidelines and University of the West of England, Bristol good practice. Both the researcher and the participants will retain copies of the signed consent forms.

Families who drop out after initially agreeing to participate will be asked about their decision and any information offered will be recorded as an outcome of the feasibility study.

7.1 Delivery of the intervention

A local Health Visitor from Bristol and Nottingham will be recruited to deliver the FAST parent programme in the two study centres in each city during the feasibility study. They will be trained by Parenting UK to deliver the programme supported by the Nursery/Children’s Nurse or Health Visitor in the Health Visiting Team at that children’s centre . Crèche facilities will be provided to enable attendance at the parenting group and refreshments will be provided for parents.

To ensure fidelity in delivery of the programme the researcher in each study centre will attend the training in how to deliver the programme and will observe and record fidelity of programme delivery. In addition, Health Visitors delivering the programme will participate in teleconference discussions with each other and Parenting UK after the delivery of each session to raise issues, concerns and ensure the programme is delivered in as consistent a manner as possible.

Furthermore, the health economist (JP) will observe a selection of sessions to ensure that all resource costs are correctly included in the economic evaluation.

7.2 Post intervention data collection

7.3 Interview conduct

Topic guides will be used in order to assist questioning during both face-to-face and telephone interviews. These guides are designed to direct but not dictate data collection and will incorporate considerable flexibility to allow participants to introduce new issues not anticipated by the researchers. The topic guides will be modified as necessary throughout the course of the study to reflect findings as they emerge. The researcher will use open-ended questioning techniques to elicit participants’ own experiences and views, and participants will be asked to provide examples. Face-to-face interviews for both parents and professionals are anticipated to last approximately 30 minutes. Telephone interviews may be briefer than 30 minutes.

7.4 Outcome measures

The outcome measures have been chosen to identify whether the parenting programme can reduce injury occurrence and the mechanism by which any change occurs.

7.5 Data analysis

7.6 Researcher safety

The researchers will follow the University of the West of England’s Researcher Safety Guidance when conducting any field work away from university premises. This will involve undertaking an assessment of risk prior to arranging an interview, prior informing a designated person from the study team the details of an interview, and calling in when the interview has been completed at an agreed time. If the designated person is not contacted at the agreed time, the designated person will contact the researcher’s mobile telephone. If there is no answer, the designated person will phone the participant’s house. If contact has still not been made, the designated person will phone the police and ask them to visit the participant’s house.

7.7 Research governance

The sponsor for the research will be the University of the West of England, Bristol. A research fellow (Project Manager) will be appointed in both Bristol and Nottingham to oversee the day to day running of the project. The Project Manager will be supervised by Julie Mytton in Bristol and Caroline Mulvaney in Nottingham.

Governance of the feasibility study will be through a management group comprising the co-applicants and collaborators. The management group will meet monthly at the beginning and end of the study and bimonthly during the study. Meetings will be in person where possible and by teleconference were necessary. The management group will oversee the progress of the study and adherence to timescales and the project plan. Not all co-applicants will be required for every management group meeting. The principle investigator (JM) will report to a Trial Steering Committee (TSC) that will meet four times during the course of the study. The requirement for a Data monitoring and Ethics Committee (DMC) will be the decision of the Chairperson of the TSC. As this is a feasibility study, and there are no plans for interim data analysis that could potentially lead to early closure of the trial, we consider the trial to be low risk and do not anticipate the need for a DMC.

7.8 Writing up and dissemination

Findings of the study will be made available to the participating families. They will also be disseminated to the Parents Advisory Group (see below) and to Health Visiting teams and children’s centre managers at intervention and control sites. The findings of the feasibility study will be written in non-specialist language so that they may be accessible to families and a range of professionals. Any quotations from participant or professional’s interviews will be anonymised prior to inclusion in the study report.

We anticipate that a subsequent main trial would generate outcomes of interest to the health community, local authorities, the public and to academics. Care would be required not to disseminate the findings too widely after the feasibility study before a main trial; otherwise recruitment to the main trial may be compromised. We would however provide local feedback in Bristol and Nottingham and would prepare papers for publication and a UK conference presentation.

2.1 Research team roles

Recruitment and quantitative data collection will be conducted by the research fellows (RF) appointed in Bristol and Nottingham, employed specifically for this project. JM will line manage and provide supervision for the RF in Bristol and CM will provide similar support for the RF employed in Nottingham. JM, CM and PB will analyse the quantitative data with the RFs. JI will facilitate the Parent Advisory Group and support the health visitors delivering the programme. JI will conduct the qualitative interviews in Bristol and CM will conduct those in Nottingham. JI will analyse the qualitative data supported by CM and TD. JP will collect and analyse the data for the economic analysis of the programme.

Childhood injuries – the scale of the problem

Unintentional injury is the major cause of death in children over the age of 1 in the UK, and for each child that dies many more will suffer morbidity and possibly long term consequences. Over 2 million visits to accident and emergency departments and over 120000 admissions in children occurred in 2005 due to unintentional injury costing the NHS in the region of £146 million (1). Staying safe has been a fundamental component of child health policies such as Every Child Matters (2). The type and location of child injuries varies with age and the child’s stage of development. The majority of injuries occurring to preschool children occur in the home (3,4,5). Between 2000 and 2002 an average of 502,000 children aged 0–4 years attended hospital every year in the UK due to a home injury, representing 78% of all injuries occurring to children in this age group. The most frequent events leading to injuries in preschool children include, in order; falls, hitting/being hit/crushed by objects, poisoning, and burns/scalds (6). Inequalities in injury occurrence have been widely reported (7–12).

Risk factors for injury

A number of risk factors related to the family and the child have been associated with increased risk of injury. Single parents, step families and teenage parenthood, maternal life events and maternal depression were all associated with increased risk of medically attended child injury by age 2 in the Avon Longitudinal Study of Parents and Children (ALSPAC) (3). In a randomised controlled trial in Nottingham family factors including having a teenage mother or being in a single parent family were associated with increased risk of hospital attended injuries in preschool children (13). Family structure (14) and parental behaviours, such as excessive use of alcohol (15) have been associated with increased injury risk in children. Male sex and difficult behaviour in childhood, particularly that relating to antisocial, aggressive or overactive behaviour, have been associated with increased incidence of unintentional injuries in the UK (16,17) and in other high income countries (18). Parental understanding of the relationship between injury risk and child behaviour and development is variable, and provision of educational anticipatory guidance has been recommended (19).

Parenting programmes

Parenting programmes are short term interventions to promote changes in the behaviour of parents and children that result in better health and wellbeing outcomes for both. They are usually delivered as face to face programmes, either individually or in groups. Parenting programmes have been increasingly recognised as an intervention to improve the life chances of children due to their effectiveness in reducing antisocial behaviour and improving educational and mental health outcomes in children, and the improved mental health and wellbeing of parents. Low socioeconomic status, unemployment, social exclusion or isolation, young or single parenthood and learning difficulties are known to adversely affect parenting. Consequently, parenting programmes have become a core component of child and family policy (20). Parenting programmes have been developed on the basis of two main theoretical approaches: behavioural and relational. Some programmes combine elements of both approaches. Behavioural approaches aim to develop parents understanding of the negative impact of attention to problem behaviour and lack of attention to positive behaviour, and teach positive discipline practices including praise and time out; relational programmes aim to improve interactions between parent and child, correcting misattributions and increasing understanding of developmental phases. Both have been developed to improve children’s mental health, the former with a particular emphasis on the prevention and treatment of antisocial behaviour and conduct disorder.

Analyses of longitudinal studies have shown the influence of parents on child outcomes that are related to injury risk. Research from the ALSPAC cohort has shown that positive parenting behaviour, parent–child interaction and a stimulating home environment were associated with enhanced development by the age of three (21) and improved cognitive and behavioural outcomes in children by age 5 (22). The ‘better’ the parenting, the more likely children are to be well adjusted and developmentally competent (23). Other studies, for mothers with learning difficulties, have shown that supportive parent training can improve childcare practices (24). Evidence suggests that enhanced carer supervision can help reduce injury risk to children (25,26). Parenting interventions have the potential to reduce poor maternal mental health and increase maternal self efficacy (27,28), to improve maternal–child interactions (29), and to change child behaviour, especially behaviour that is challenging or could place the child at risk of injury (27,30,31). Parenting interventions can reduce injury risk either through these mechanisms or through increased parental knowledge of safety practices (32), improvement in the quality of the home environment (33), or through the use of home safety practices such as having a fitted and functioning smoke alarm, using stair gates or keeping sharp objects safely (34,35). Parenting programmes have shown reductions in injury risk taking behaviour in primary school aged children (36). Health visitor interventions to support parents can reduce injury rates in both prospective studies (37) and meta-analysis of randomised controlled trials (38). Meta-analysis of parenting interventions, primarily conducted in high-risk or disadvantaged families, have demonstrated significantly lower risks of injury, as measured by parental self-report of either medically or non-medically attended injuries (39,40). Parents value programmes that enable the acquisition of knowledge, skills and understanding, and facilitate acceptance and support from other parents. Such outcomes reduce feelings of guilt and social isolation, increase empathy with children, and give confidence to cope with challenging child behaviour (41). A child’s medically attended injury represents a ‘teachable moment’ when parents are receptive to information regarding injury risk in their children (42).

The features of parenting interventions that are most effective are becoming clearer. There is strong evidence that home safety education and the provision of safety equipment are effective in increasing a range of home safety practices (35). A review of ‘what works?’ in parenting interventions has shown that interventions are more likely to be effective if they are delivered early in childhood, if intensity is proportional to need, if they include group activities where parents can benefit from the social aspect of working with peers, if they include formal programmes or manuals to maintain the consistency of the delivery of the intervention which should be delivered by trained staff, and if there is a focus on specific parenting skills and practical ‘take-home’ tips (43). A review of the effectiveness of parenting support programmes in European countries where universal early intervention approaches tend to be used, suggests that positive outcomes can also be achieved when the programme is delivered by non-health workers or agencies (44).

The cost effectiveness of parenting programmes has not been widely studied (45). A recent systematic review of economic evaluations of child and adolescent mental health interventions demonstrated that most evaluations were small scale, had short time horizons for assessing outcomes and had limited reporting (46). However, the cost effectiveness of parenting programmes has been established for group parenting programmes. A formal evaluation of Sure Start parenting programmes demonstrated improved child behaviour outcomes for modest cost and considered the programme value for money (47).

Justification for this proposal

Parenting interventions, usually delivered as part of a programme to improve a range of child and family outcomes, appear to be effective in reducing self-reported or medically attended injuries in young children (48). Due to the range of positive outcomes associated with the programmes, such as improved child behaviour, maternal self-efficacy or maternal–child interactions, it is unclear whether one of these outcomes is more effective in reducing child injury than another. We know that injury prevention education alone has not shown reduction in injury occurrence (35) but the hypothesis that injury prevention education in the context of a parenting programme may be effective remains to be tested. Furthermore, it is unclear whether group based programmes, delivered outside of the home can achieve reductions in injury occurrence similar to intensive one-to-one home based programmes. Evidence suggests that RCTs of home safety education can successfully recruit parents of recently injured children (49,50), and that parents are interested in learning first aid (51). We therefore propose to develop a parenting programme that provides injury prevention education through the delivery of first aid and safety training tailored to the stages of preschool child development and delivered to groups of parents in a community setting. We propose to test the feasibility of delivering that programme with a view to a future large scale randomised controlled trial.

4.1 Aim

To develop and test the feasibility of delivering a children’s centre based parenting programme to prevent recurrent unintentional home injuries in children aged 0–4 years; compared to normal care for such children.

4.2 Objectives

1. to develop a health professional delivered parenting programme

2. to assess the acceptability of the parenting programme to parents and professionals

3. to assess the feasibility of delivering the parenting programme

   1. to assess recruitment and retention of parents within the trial

   2. to assess compliance with the intervention during the follow up period

   3. to determine the training, equipment and facilities needed for delivery of the parenting programme

   4. to assess the collection of primary and secondary outcome measures

   5. to determine which information to collect on ‘normal care’

   6. to assess which relevant resource utilisation/costing data needs to be collected

   7. to produce estimates of effect sizes to inform sample size estimation for the full trial

5.1 Experimental group

The experimental arm of the trial will be ‘normal care’ plus a parenting programme. The parenting programme will be developed in collaboration with Parenting UK (a parenting programme development organisation). Engaging parents in a parenting programme following injury in their child may be difficult since the injury may result in feelings of stigmatisation, guilt or concern that the injury is believed to be intentional. Our programme will contain both home safety education (‘keeping your child safe’) and first aid advice as methods of primary and tertiary injury prevention. Unpublished evidence from local injury prevention projects in Gateshead and Bristol suggest that parents are interested in learning first aid and willing to attend a group to do so. The emphasis on first aid advice and safety training is hoped to enhance acceptability and diffuse any negative feelings generated by being identified for participation in the programme. Interest in learning first aid is hoped to stimulate further interest in injury prevention and parenting to reduce injury risk. The programme will be delivered in children’s centres by a Health Visitor and cofacilitator, to groups of parents.

The parenting programme is likely to contain elements of existing parenting programmes that may reduce injury risk, for example, those intended to enhance parental self-efficacy and well being, improve parent/child communication, and improve child behaviour through the increased use of positive reinforcement, and the enablement of setting and maintaining boundaries (52). Evidence-based safety components are likely to include home safety education, assessment of home hazards, guidance on types, sources and fitting of home safety equipment and tailoring advice to both the home context and understanding how a child’s injury risks change as the child grows and develops (anticipatory guidance) (53).

Families in the experimental group will be invited to participate in 1 to 1 interviews after delivery of the parenting programme to explore parents’ views and experience of the programme.

5.2 Control Group

The control group will receive ‘normal care’. In both Bristol and Nottingham, Health Visitors are routinely sent details of children within their geographical area of responsibility that have sustained a medically attended injury, whether at A&E, or NHS Walk-in Centre. Usually the notification is sent by fax either on the same day as the injury or on the next working day. It is usual practice that following receipt of such notification the Health Visitor Team may take a range of actions determined by the circumstances of the event and the Health Visitor’s understanding of the needs of that family. Four different actions can be considered as part of ‘normal care’; telephone contact, face-to-face contact, referral to services, or no action. The use of each of these four possibilities will vary between locations and between Health Visitor Teams. For example, some Health Visitors may take every fax notification as an opportunity to make contact with a family, especially in deprived or multicultural communities, where parent initiated contact with Health Visitors is low. In other areas, workload or knowledge of families will mean that Health Visitors are much more selective of those families that they will contact.

5.3 Inclusion and exclusion criteria

5.4 Randomisation and allocation

The unit of randomisation will be the children’s centre . We will recruit a total of 8 children’s centres; four in Bristol and four in Nottingham. Those agreeing to participate will be stratified by study centre (two strata) and randomly allocated within strata to treatment arm using a remote automated system available through the Bristol Randomised Trials Collaboration (BRTC) based at the University of Bristol. Two children’s centres in each study centre will be randomly allocated to the intervention arm, and two children’s centres in each study centre will be allocated to the control or ‘normal care’ arm.

To reduce post-randomisation recruitment bias, informing the children’s centres and Health Visitor Teams of their allocation to intervention or control arms will be delayed until after recruitment of families has been completed. If allocation to intervention or control arm is indicated to the Health Visitor Team/children’s centre at the time of recruitment, we anticipate two potential post randomisation recruitment biases; a) the Health Visitor Team could choose not to offer entry to the study if the team knew that the family were unlikely to participate or continue in the programme once commenced, and b) a family may be influenced in their decision to participate if they knew in advance that their children’s centre was, or was not, offering the FAST Parents programme. Therefore once recruitment has been completed Health Visiting Teams and children’s centres will be informed of their allocation.

6.1 Ethics committee approval

Approval will be sought from a Type 3 committee; South West 3 REC (Bristol Central) based at University Hospitals Bristol.

6.2 Participant consent

Parents asked to participate in this research are entitled to choose whether or not to take part. Their decision will be voluntary and they will be competent to understand what is involved. Consent forms will be designed to assure the protection of their rights.

In the eight children’s centre areas participating in the study receipt of a notification of a medically attended injury by the Health Visitor Team will result in the team making a decision on ‘normal care’. The families will be contacted after completion of ‘normal care’ to advise them that their local children’s centre is participating in a study to follow up pre-school children who have had an injury and that some children’s centres will be offering first aid advice courses for parents. Parents will be asked if their details can be passed to the research team who will tell them more about the study. Refusal to participate in the programme will not prevent access to any other routinely available services, although will be recorded as an outcome of the feasibility study along with the reason for refusal where this is provided. The Health Visitor Teams will be given a list of the inclusion and exclusion criteria to enable them to determine eligibility of families for participating in the study.

Identification of eligible families may occur through alternative routes when it is inappropriate for Health Visitor teams to undertake recruitment themselves (e.g. reduced capacity within the team). Identification may occur via Emergency Departments (ED) where generation of the notification letter occurs. Eligible families will be contacted either by telephone by a member of the ED team and asked if their details can be passed to the research team who will tell them more about the study, or by letter sent by the ED team to the parents with a reply slip to the research team. Parents will not be approached during their ED visit, but afterwards. The number of parents approached but declining to have their details passed to the research team, or failing to return a reply slip, will be noted, together with the reason for refusal where this is provided. The member of the ED team will be given a list of the inclusion and exclusion criteria to enable them to determine eligibility of families for participating in the study.

Identification may also occur via primary care as the notification letter will routinely be sent to the child’s General Practitioner. Eligible families will be contacted by letter sent out by a member of the primary care team who will be provided with a list of the inclusion and exclusion criteria to enable them to determine eligible families. General Practices linked to the participating Health Visitor teams will be invited to support the study. A template letter will be provided for sending to eligible families that allows the surgery to add their own header/logo. The letter, sent from their General Practitioner, will introduce the FAST study, enclose the parent information sheet, and ask that if the parent is interested in participating they should telephone the local FAST Research Fellow (a mobile telephone number will be provided) or return a reply slip in a reply paid envelop. Parents will be able to choose whether or not to respond to the letter.

Families that agree to be contacted by the research team will receive both written and verbal information. The research fellow will send a study information sheet and consent form to the family. One week later the researcher will contact the family and ask permission to visit. If the family agrees the researcher will send baseline measure questionnaires prior to the visit.

Parents living within the catchment areas of intervention and control children’s centres will be invited to participate in the study. At the visit the researcher will verbally explain the study. The explanation will cover all the elements specified in the written information provided for the participant. The participants will be informed of the aims, methods and participation requirements of the study. They will be informed that the study is intended to help understand children’s behaviour after injuries and reduce the risk of further injuries. The researcher will explain that some families will have the opportunity to attend a programme will provide the opportunity to gain first aid advice and safety information in a ‘hands-on’ and interactive manner, but because this is a new programme that has not previously been tested the likelihood of success is unclear. Parents will be informed that we are interested in their experience of participating in the study and the programme to help us understand how they could be improved. Parents will be advised that there are no anticipated risks to participation.

The participant will be given every opportunity to clarify any points they do not understand and if necessary ask for more information. At the end of the discussion the participant will be given time to reflect. The participant will be informed that they are at liberty to withdraw their consent to participate at any time, without prejudicing any future medical care.

The researcher will obtain the participants freely given written consent before participating in the study. The consent form will assure the participant of the confidentiality of the data collected. Participants attending children’s centres in the experimental arm of the study will be asked permission that the interviews conducted after the parenting programme can be audio recorded and to publish anonymised quotations from the study. With the agreement of the participants, the data will be anonymised and stored in accordance with data protection guidelines and University of the West of England, Bristol good practice. Both the researcher and the participants will retain copies of the signed consent forms.

Families who drop out after initially agreeing to participate will be asked about their decision and any information offered will be recorded as an outcome of the feasibility study.

We aim to recruit 12 families from each of four children’s centre areas in Bristol (48 families) and from each of four Children’s centre areas in Nottingham (48 families). We expect that between recruitment and start of the intervention some families will drop-out of the study. We hope that about 8 families per children’s centre will remain in the study by the start of the intervention period (total = 64 families).

7.1 Delivery of the intervention

Health Visitors from Bristol and Nottingham will be recruited to deliver the FAST parent programme in the two Children’s centres in each city during the feasibility study. They will be trained by Parenting UK to deliver the programme supported by a cofacilitator. The Nursery/Children’s Nurse or Health Visitor in the Health Visiting Team at that children’s centre will be invited to support delivery of the intervention, but if unable, a Health Visitor and cofacilitator independent of the Health Visitor teams participating in the study will be available to deliver the intervention. Crèche facilities will be provided to enable attendance at the parenting group and refreshments will be provided for parents.

To ensure fidelity in delivery of the programme the researcher in each study centre will attend the training in how to deliver the programme and will observe and record fidelity of programme delivery. In addition, Health Visitors delivering the programme will participate in teleconference discussions with each other and Parenting UK after the delivery of each session to raise issues, concerns and ensure the programme is delivered in as consistent a manner as possible.

Furthermore, the health economist (JP) will observe a selection of sessions to ensure that all resource costs are correctly included in the economic evaluation.

7.2 Post intervention data collection

7.3 Interview conduct

Topic guides will be used in order to assist questioning during both face-to-face and telephone interviews. These guides are designed to direct but not dictate data collection and will incorporate considerable flexibility to allow participants to introduce new issues not anticipated by the researchers. The topic guides will be modified as necessary throughout the course of the study to reflect findings as they emerge. The researcher will use open-ended questioning techniques to elicit participants’ own experiences and views, and participants will be asked to provide examples. Face-to-face interviews for both parents and professionals are anticipated to last approximately 30 minutes. Telephone interviews may be briefer than 30 minutes.

7.4 Outcome measures

The outcome measures have been chosen to identify whether the parenting programme can reduce injury occurrence and the mechanism by which any change occurs.

7.5 Data analysis

7.6 Researcher safety

The researchers will follow the University of the West of England’s Researcher Safety Guidance when conducting any field work away from university premises. This will involve undertaking an assessment of risk prior to arranging an interview, prior informing a designated person from the study team the details of an interview, and calling in when the interview has been completed at an agreed time. If the designated person is not contacted at the agreed time, the designated person will contact the researcher’s mobile telephone. If there is no answer, the designated person will telephone the participant’s house. If contact has still not been made, the designated person will telephone the police and ask them to visit the participant’s house.

7.7 Research governance

The sponsor for the research will be the University of the West of England, Bristol. A research fellow (Project Manager) will be appointed in both Bristol and Nottingham to oversee the day to day running of the project. The Project Manager will be supervised by Julie Mytton in Bristol and Caroline Mulvaney in Nottingham.

Governance of the feasibility study will be through a management group comprising the co-applicants and collaborators. The management group will meet monthly at the beginning and end of the study and bimonthly during the study. Meetings will be in person where possible and by teleconference were necessary. The management group will oversee the progress of the study and adherence to timescales and the project plan. Not all co-applicants will be required for every management group meeting. The principle investigator (JM) will report to a Trial Steering Committee (TSC) that will meet four times during the course of the study. The requirement for a Data monitoring and Ethics Committee (DMC) will be the decision of the Chairperson of the TSC. As this is a feasibility study, and there are no plans for interim data analysis that could potentially lead to early closure of the trial, we consider the trial to be low risk and do not anticipate the need for a DMC.

7.8 Writing up and dissemination

Findings of the study will be made available to the participating families. They will also be disseminated to the Parents Advisory Group (see below) and to Health Visiting teams and children’s centre managers at intervention and control sites. The findings of the feasibility study will be written in non-specialist language so that they may be accessible to families and a range of professionals. Any quotations from participant or professional’s interviews will be anonymised prior to inclusion in the study report.

We anticipate that a subsequent main trial would generate outcomes of interest to the health community, local authorities, the public and to academics. Care would be required not to disseminate the findings too widely after the feasibility study before a main trial; otherwise recruitment to the main trial may be compromised. We would however provide local feedback in Bristol and Nottingham and would prepare papers for publication and a UK conference presentation.

1. Introduction

Thanks. Introduce self. Re-state purpose of the interview and use of audio-recording

2. Background

Could you start by telling us your name, and the children’s centre that you work with/in?

3. Programme experience

Have you ever been involved with delivering/organising/hosting a programme for parents before? If so, what was it about? In what ways was it similar or different to this programme?

What was your experience of the FAST Parent Programme? (positive and negative features)

What were you hoping that parents would get out of the FAST Parent Programme? Do you think those expectations/needs were met? Why?

Were there features of the programme that you think were particularly enjoyed by the parents? Or were particularly helpful for the parents? What were they and why were they enjoyable/helpful?

Were there features of the programme that you think the parents did not enjoy or were not helpful? What were they and why do you think they did not work so well?

What did you think of the format of the programme? (number of sessions/frequency of sessions/duration of sessions?, the leadership required?, the materials used?)

What was your experience of having the course running in the children’s centre/Community Centre? (Positive and negative)

4. The future

Would you want to host/run another FAST Parent Programme in the future? Why?

5. Other issues

Any other issues that the participant would like to raise

Thank them for their time.