A collaborative approach to adopting/adapting guidelines - The Australian 24-Hour Movement Guidelines for the early years (Birth to 5 years): an integration of physical activity, sedentary behavior, and sleep

The GRADE-ADOLOPMENT process followed the framework described in detail by Schünemann et al. 2017 [11]. In addition, several steps that were identified in the Appraisal of Guidelines for Research & Evaluation II (AGREE-II) instrument were included [12]. A summary of the timeline and sequence of steps we used is shown in Fig. 1.

Step 2: Formation of a Consensus Panel. A guideline development Consensus Panel was also formed which included expert researchers, representatives from key stakeholder groups (including parents and Indigenous communities), and methodology experts (Table 1). The role of this Panel is described in detail in Step 5. Efforts were made to achieve geographical representation across Australia within the confines of the budget.

Step 3: Identification of credible existing guidelines and definition of criteria for selection of the guidelines. We were aware of two sets of 24-h integrated movement guidelines for the early years. These were from Canada and New Zealand and, at the time (February 2017), were identified as being in development. The Canadian integrated Guidelines (then in development) were considered along with other existing integrated or physical activity and sedentary behaviour guidelines that met the following criteria: 1) published in the past 5 years (or in the process of being published); 2) addressed clear research questions (contained all Population, Intervention, Comparator and Outcome [PICO] elements); 3) followed the GRADE process; 4) allowed for updating (provided access to full systematic reviews, which were registered with the Prospective Register of Systematic Reviews (PROSPERO) and provided full access to the search strategy); 5) existing and accessible GRADE tables and summaries of findings; and 6) completed a risk-of bias assessment [11]. Table 2 contains a summary of the five sets of national physical activity and sedentary behaviour guidelines in the early years that the leadership group was able to find and the evaluation of these against these criteria. Using our existing networks we were unable to identify any other 24-h guidelines for the early years. Only the 2017 Canadian 24-Hour Movement Guidelines for the Early Years met these criteria and were therefore chosen as the guidelines to be adopted or adapted following the GRADE-ADOLOPMENT process.

The AGREE-II tool was used to determine the credibility of the 2017 Canadian Guidelines (as per Stage 1 of the suggested GRADE-ADOLOPMENT Protocol – see Appendix 1 [11]). As the AGREE tool was developed to enable end users to assess the quality of a completed guideline, some of the items are not relevant to guidelines in development. This includes AGREE Domains 4 (clarity of presentation) and 5 (applicability). Although the Canadian guidelines were not yet published at the time the Australian Guidelines were being developed, and the AGREE-II form could not be completed in its entirety, we were confident based on the nature of our relationship that the ratings for each of the AGREE-II criteria would be high. Following the credibility assessment the ADOLOPMENT framework moves on to the evaluation and final selection of the guidelines that will be adopted or adapted. It was agreed by the Leadership Group that it would be appropriate to adopt the Canadian Guidelines as they were determined to be of appropriate quality, their scope/applicability was appropriate for Australia, the topic was a priority for Australia and the research questions and PICOs (Population, Intervention, Comparators, and Outcomes) for the systematic reviews that served as the evidence base were relevant.

In most cases, assessed against the stated GRADE approach to evidence synthesis (60% of RCT studies statistically significant and positive), the evidence base was graded “Low” or “Very Low”. The Consensus Panel then made a decision to support the draft 2017 Canadian Guidelines or not based on the evidence and other criteria used to make recommendations including values and preferences; feasibility, acceptability and equity issues; resources; balance of benefits and harms; and quality of the evidence [12]. Parts of the EtD framework able to be followed during the Consensus Panel meeting included presenting the evidence and keeping track of the discussion and judgments. Following the Consensus Panel meeting, a transparent record of the discussions was communicated to those who attended.

Step 5 Determine availability, completeness & currency of information about EtD criteria. The next component in the general stages of GRADE-ADOLOPMENT (see Appendix 3 [11]) was to determine the availability, completeness, and currency of the information about the EtD criteria. For this, the criteria for updating reviews found in Appendix 4 of the GRADE-ADOLOPMENT paper [11] was used (see Table 3). Based on this information, the Leadership Group made a decision to update the Canadian systematic reviews focusing only on the critical outcomes (see [10] for a list of these for each systematic review) for randomized controlled trials and cohort study designs because the sources of these reviews were older than three months (i.e., they had an end date before November 2016) [11]. We decided not to update the reviews for non-critical outcomes (see [10] for a list of these) or for cross-sectional studies because the consensus was that even if an update was to uncover new studies, they would be graded low quality and as such, would not result in a change to the final guideline. The Australian Leadership Group made the PICOs that guided the four systematic reviews for the 2017 Canadian Guidelines available for comment by the Australian Consensus Panel prior to the Consensus meeting. The Panel was asked to comment on the appropriateness of each of the PICOs for the Australian context. Some of the initial comments sought clarification on the selection of specific search terms for some of the outcomes. These were resolved by indicating that these would be or were captured in the Australian or Canadian searches, respectively, although this information was not clear in the PICOs. Other queries related to the inclusion of information in the summary tables or in the PROSPERO registration or to definitions of specific terms. Where changes were suggested, these were discussed by the Leadership Group and agreement reached. None of the proposed changes were substantial enough to warrant changing any of the existing PICOs.

The updates to the four systematic reviews performed for the Canadian Guidelines were conducted with searches completed up to the end of March, 2017. These updated reviews were also fed back to the Canadian Guideline Development Panel to consider as part of their guideline development process. For each systematic review, the quality of evidence was assessed by outcome/indicator and study design, and age group, using the GRADE approach [13, 14]. Each systematic review used the same PICO as the corresponding systematic review completed for the 2017 Canadian Guidelines [15‐18].

The results of these systematic review updates were presented at the Consensus Panel meeting from 10 to 12 April 2017. The specific objectives of this meeting were to review, discuss, debate and interpret findings from the Canadian systematic reviews and Australian updated searches, including compositional analyses that were performed on data from Canada, the results of which appear elsewhere in this supplement [19]. Other objectives were to review and adopt/adapt the Preamble and the actual Canadian 24-Hour Movement Guidelines for the Early Years; discuss proposed stakeholder consultations; identify research gaps; and plan the launch, dissemination, promotion, integration, and evaluation activities for the Australian 24-h Movement Guidelines for the Early Years (Birth to 5 years).

The process at the Consensus Panel meeting involved reviewing the evidence for each movement behaviour (physical activity, sedentary behaviour, and sleep) individually, starting with the 2017 Canadian systematic reviews and integrating the Australian updates into these reviews. The evidence for each behaviour, including the conclusions of the Canadian review and how this process informed their guidelines was then discussed. The Consensus Panel then followed the GRADE-ADOLOPMENT process to make a decision to adopt or adapt the 2017 Canadian recommendations for each behaviour or create de novo recommendations. In addition, the Panel examined the results of the integrated behaviours systematic review and compositional data analyses from Canada, infused expert opinion into the evidence (such as feasibility, acceptability, equity issues, values and preferences, resources, and balance of benefits and harms), combining evidence of absolute effects across multiple outcomes [20‐23], leading to an informed assessment of whether the panel either agreed or disagreed with the judgments made by the Canadian Guideline Development Panel. If the Australian Consensus Panel agreed with the judgments, the recommendations were adopted and the Panel moved on to discuss the guidelines wording. If the Panel disagreed with the judgments, the recommendations were adapted and the Panel moved on to describe the reasons for deviation in the EtD framework. It was noted during the Consensus Panel meeting that a recommendation could be adopted and still added to or translated for adoption in the wording and adjusted if necessary based on this detailed discussion.

The next three sections of the Guideline Development Process [24] are not parts of the GRADE-ADOLOPMENT process but were important components in testing the appropriateness of the adopted guidelines with key stakeholders and in developing plans for the Australian Government (owner of the Guidelines) to consider in the activation of the Guidelines and their potential monitoring and surveillance.

The online survey developed as part of the 2017 Canadian Guidelines [10] was modified for the Australian context to seek feedback from stakeholders on their level of agreement with the draft Australian Guidelines emanating from the Consensus Panel meeting. The Human Research Ethics Committee of the University of Wollongong approved administration of the survey and use of a passive consent process (HE 2017/164). The survey sought assessments of the clarity of the title, preamble, and guidelines as well as levels of agreement with the text. Basic demographic information was requested and respondents had an opportunity to provide comments on any aspect of the guidelines. Consensus Panel members were asked to disseminate the survey through their networks, and used a snowball sampling methodology to optimise reach and input from relevant stakeholders. The survey was open from May 18 to June 13, 2017. After the survey closed, numerical responses from participants were tabulated and analysed. Written comments were consolidated into themes and summaries were prepared. The stakeholder survey also allowed respondents to express their interest in publicly disclosing their support for the guidelines pending their review of the final draft. To facilitate this, interested respondents were asked to provide an email address where the final guidelines could be sent.

In addition to the online stakeholder survey, focus groups targeting early childhood educators and parents from different socioeconomic and cultural backgrounds – including Indigenous groups and those who work with children with additional needs – were conducted. These were supplemented with key informant interviews held with a paediatrician; general practitioner with a diploma in Child Health; paediatric physiotherapist; the authority leading the implementation of the National Quality Framework for early childhood education and care services in Australia (ACECQA); and a not-for-profit organization representing long day care owners and operators (Australian Childcare Alliance). These focus groups and interviews sought to understand stakeholders’ perceptions of the Australian 24-Hour Movement Guidelines for the Early Years. The focus groups and interviews asked key stakeholders who were difficult to reach through the online survey specific questions about the acceptability and perceived importance, clarity of the guidelines and preamble, facilitators and barriers to implementation and dissemination, and dissemination and implementation recommendations for the Guidelines) [25, 26]. A total of 35 individuals participated in five focus groups (6 participants per focus group) and five interviews (1 participant per interview). Recruitment occurred through existing partnerships and connections. Focus groups and interviews lasted between 30 and 90 mins and were conducted from late-May to mid-September 2017 in the Illawarra region of New South Wales, Sydney, Melbourne, and Perth. The focus groups and interviews were audio-recorded and transcribed verbatim and inductive and thematic data analyses by two researchers were employed and consensus reached on any discrepancies through discussion [27]. Ethics approval was obtained from Human Research Ethics Committee of the University of Wollongong (HE 2017/164). A subcommittee of the Consensus Panel reviewed the survey, focus group and key informant interview results, and suggested revisions to the Guidelines based on the stakeholder feedback, ensuring changes remained true to the available evidence base. Revisions agreed upon by the Leadership Group were then circulated to the entire Consensus Panel for comment and final revisions. Consensus was achieved on the final Guidelines.

A sub-group of the Consensus Panel developed a detailed dissemination, implementation and integration strategy. This included key communication strategies leading up to and beyond the official launch of the Guidelines with government and non-government support for the integration of the guidelines into early childhood education and care settings and into other support services and whole-of-government approaches. Consultation with key users of the guidelines during the Consensus Panel meeting indicated that the most beneficial methods of dissemination, implementation and integration of the guidelines to facilitate behaviour change were: enhancing parent education through the development of an app; using social media to promote actively persuasive messages and materials; training for anticipated end-users (webinars, online, and face-to-face professional development); and a comprehensive social marketing campaign. Expected population health behavioural change and health system impacts and return on investment were modelled based on pilot evidence from previous child and youth guidelines [28].

A phased evaluation plan was also developed. This included initial evaluation of the dissemination and reach of the guidelines and the integrated communications strategy. Ongoing evaluations of community ownership, implementation and integration plan, as well as awareness and knowledge of the guidelines as pathways to long-term improvements in 24-h movement behaviours across early childhood populations were proposed as part of the plan.

Research gaps were identified through the updates of the systematic reviews and during discussions at the Consensus Panel meeting. This included thinking about how surveillance and monitoring of the new guidelines would occur. The full set of research gaps were distributed to the Consensus Panel after the meeting for further feedback and agreement. A sub-committee met several times via teleconference to make initial recommendations around the monitoring and surveillance of the new Australian 24-Hour Movement Guidelines for the Early Years using the same approach as the Canadian Surveillance Sub-committee and having access to the surveillance recommendations table from the Canadian sub-committee [10].

The full systematic reviews from the 2017 Canadian guidelines are available in this special journal issue [16‐18, 29]. The results of the updates to the Canadian systematic reviews by the Australian Leadership Group are described below.

For physical activity, 2458 studies were identified from a search of databases, with 24 studies included after screening title and abstracts. Of these, one additional study met the criteria to be included in the update. This was a longitudinal study that assessed physical activity using accelerometry at ages 19 months (n = 185) and 42 months (n = 116) and motor development at age 5 years [30]. Results showed that moderate- to vigorous-intensity physical activity (MVPA) at 36 months predicted locomotor skills at age 5 years but not object control skills. MVPA at age 19 months was not associated with any motor development outcome at age 5 years. The assessed quality of overall evidence using GRADE criteria for this outcome (“Very Low”) did not change by including this additional study from the updated review.

The sedentary behaviour updated systematic review captured 1820 studies with 99 studies remaining after titles and abstracts were screened. Three additional studies met the inclusion criteria and were included in this update. These comprised one RCT [31] and two longitudinal studies [32, 33]. In the RCT, 70 preschoolers were randomized to watching a fast or slow-paced film to examine the subsequent effect on children’s attention. This was measured by the number of behaviour changes during free-time play. The children who watched the fast-paced film had a higher number of negative behaviour changes compared with those who watched the slow-paced film [31]. The two longitudinal studies comprised 2432 participants aged between 3 years and 5 years at baseline. The first study (n = 2411) assessed hours of TV viewing at age 5 and subsequent adiposity and mental health at ages 8, 10, 14, 17, and 20 years [33]. Those children who had lower levels of TV viewing at age 5 had a lower percentage body fat at age 20. There were no associations with mental health. The second study (n = 111) used an objective measure of sedentary time and found no association with body composition over a 12mo period among children aged between 3 and 5 years [32]. The assessed GRADE quality of overall evidence did not change for longitudinal studies examining adiposity (“Very Low”) or for RCTs examining psychosocial health (“Moderate”).

For the updated sleep systematic review, 142 studies were identified from the search of databases, with six studies remaining after screening title and abstracts. Three additional studies met the inclusion criteria for the update. These comprised two RCTs [34, 35] (n = 108) and one longitudinal study [36] (n = 1192) examining the association between sleep duration and cognitive development. The two RCTs found that infants in the nap condition (who were visited after they had a naturally occurring nap) performed better on selected cognitive tasks compared with those in the no-nap condition (who were visited shortly before they were scheduled to take a nap) [34, 35]. The longitudinal study assessed sleep trajectories annually using parent-report from age 2.5 years to 10 years with follow-up at age 10. Results showed that compared to 11-h sleepers, the odds ratio of having poor receptive vocabulary at age 10 was 2.67 [95% confidence interval (CI): 1.24–5.74, P = 0.012] for short persistent sleepers and 1.66 (95% CI: 1.06–2.59, P = 0.026) for 10-h sleepers [36]. The assessed quality of overall evidence using GRADE criteria for this outcome (“Moderate” for RCTs and “Very Low” for longitudinal studies) did not change as a result of including these additional studies.

The final systematic review update included studies that investigated combinations of physical activity, sedentary behaviour, and sleep and their association with health indicators. The updated searches yielded 518 studies, with five remaining after screening titles and abstracts. No studies met the inclusion criteria and as such, these results were identical to those from the 2017 Canadian review [18].

Agreement in the interpretation of the evidence was reached for each behaviour and for the integration of the three behaviours. Based on the evidence from the Canadian systematic reviews and their GRADE tables and recommendations, in combination with the updated systematic reviews in Australia, the Consensus Panel agreed to adopt the Canadian recommendations. Once it was decided that Australia would adopt the recommendation from the modified EtD framework, the Consensus Panel then decided if they wanted to keep the guideline wording of the Canadian Guidelines. There were a number of minor changes to the wording of the guidelines, preamble and title that were made by the Australian Consensus Panel. Changes were not made to the guideline recommendations per se but rather to the wording of good practice statements [14]. When a change was suggested, the rationale for the change was put forward by the Panel member and discussed. The Panel determined if the proposed change would be consistent with the strength of the evidence recommended and ensured it would not unintentionally alter the interpretation of the guideline. Consensus was required for a change to be accepted. The changes in wording between the Canadian and Australian Guidelines are shown in Table 4. All Consensus Panel members endorsed the draft guidelines and preamble to be used for the stakeholder consultations. As the Canadians were undertaking stakeholder engagement during the period immediately post the Australian Consensus Panel meeting, further minor changes were made to the Canadian Guidelines [10]. Each time this occurred, the Australian Panel was informed and asked if they supported the same change for the Australian Guidelines. In all cases, consensus was reached to either accept or not accept the edit. There were also several instances where changes in wording made to the Australian Guidelines were considered by the Canadian Guideline Development Panel and the same process was followed to determine if this change would be made to the Canadian Guidelines, which it was in several cases.

The draft guidelines developed and approved by the Consensus Panel at their April 2017 meeting were used to seek broader consultation through the online stakeholder survey, focus groups and key informant interviews. At the close of the survey, responses from the 302 participants were tabulated and analysed. The number of responses varied by question with 181 to 249 responses for close-ended questions and 8 to 143 responses for open-ended questions. Respondents were from every state and territory in Australia except Tasmania with 36% from New South Wales, 6.1% from Victoria, 4.4% from Queensland, 28.2% from Western Australia, 5.6% from South Australia, 1.0% from the Australian Capital Territory, and 0.5% from the Northern Territory. Approximately one in five respondents were from outside Australia (18.8%). Respondents identified as being from the following sectors: research/academia (26.7%), early childhood education and care (24.6%), public health (17.1%), healthcare (12.3%), education (8.6%), government (5.4%), sport (2.1%), and physical activity/fitness (1.1%).

The proportion of respondents who strongly agreed or somewhat agreed that all sections of the Guidelines (title, preamble, guidelines) were clearly stated was very high, ranging from 89 to 97%. The proportion who strongly agreed or somewhat agreed with the message in these sections ranged from 91 to 96%. Table 5 provides a detailed breakdown of the responses from the stakeholder survey. Regarding the open-ended questions, the most frequent concerns and suggestions were in relation to identifying the key people or groups for implementing and activating the 24 Hour Guidelines, as well as identifying the support they would require. Changes were made accordingly as described in the Methods section. Many respondents (n = 85) indicated interest in publicly supporting the Guidelines once released.

Results from the focus groups and key informant interviews supported the findings from the online survey. There was a low awareness of current guidelines but consistent support for the integrated nature of the new Guidelines. Respondents found the Guidelines clear, simple and specific and felt they contained useful and practical information that was perceived, at least by educators, to be already occurring. There was awareness that educators, health workers and parents/carers all play an important role in dissemination and implementation. However, parents felt there needed to be clear messaging to minimise feelings of guilt that may be associated with not meeting the Guidelines. It was also suggested that a glossary of terms be included to provide examples and definitions for some of the terms used in the Guidelines such as sedentary screen time, sleep hygiene, energetic play, and tummy time.

A small number of changes were made to the draft Guidelines as a result of the stakeholder consultation. This included the addition of an age range in the title. Around 80% of respondents to the online survey thought the Guidelines should include a specific age range in the title, with nearly 50% of these suggesting they would like this to be expressed as “Birth to 5 years”. Being placed in a car seat was also suggested to be included as an example of restrained sedentary behaviour. In the preamble, it was suggested that the word “family” be included along with “parents, educators, and caregivers” as these are who young children should receive support from to meet the Guidelines. This was based on feedback from the focus groups of the need to be inclusive of all family structures, such as older siblings and grandparents. The final guidelines, including the title and preamble, are provided in Figs. 2 and 3.

The comprehensive plan to disseminate, implement, integrate and evaluate the guidelines was presented to the Australian Government on 28 June 2017 by a sub-committee of the Leadership Group, including a social marketing expert (Melanie Randle), health economist (SE), and knowledge translation expert (TS). Key aspects of the plan included:

Research gaps were identified through the updated systematic reviews and during discussion at the Consensus Panel meeting and are summarised in Table 6. These were determined independent of the research gaps identified in the Canadian Guideline development process. As a result, there may be some overlap between the two countries. As 24-h movement guidelines are new in the early years, there are many gaps in the research and evaluation, providing fertile ground for researchers in the future.

A sub-committee examined the surveillance recommendations made by the Canadian Guideline Development Panel (see [10]) and considered these for adoption in the Australian context. The Australian sub-committee adopted the Canadian recommendations and agreed with the rationale for those guidelines which were not recommended for surveillance until further research has been completed (see Table 7).

The Australian sub-committee recommended using a representative day (e.g., previous day) for surveillance of each of the behaviours rather than an average day (as recommended by the Canadian Surveillance Sub-committee). The rationale for recommending a representative day was that it would provide a more accurate recall and hence better estimate the prevalence of the guideline in a population representative sample [37, 38]. It would also allow direct comparison with previous national representative data collected using the same approach as part of the Australian Health Survey [39].

The new guidelines were co-released with Canada on November 21, 2017. Supporting the release by the Australian Government Health Minister were a social media strategy, and web resources housed on the Australian Department of Health website. Professional development for educators and other professionals delivered by ACECQA and state health departments will support their training post-release. At a minimum, the impact and success of the launch of the new guidelines will be assessed using indicators of dissemination reach. A plan has been developed for a comprehensive evaluation of the subsequent dissemination, implementation, and integration activities to assess community ownership and population-level community impacts on early childhood movement behaviours over time.

The final stage in the guideline development process is the planning of updates and revisions [24]. The Australian Consensus Panel recommends that these guidelines be reviewed, and updated if necessary, at least every 10 years or when significant new research emerges warranting change.

A strength of our hybrid guideline development approach was the prospective collaboration that resulted in two sets of National Guidelines being developed independently while using the same body of evidence. It allowed some cross-pollination and provided governmental co-operation between countries to enable co-release of the Guidelines.

A limitation of the Australian Guidelines is that they are for ages birth to 5 years, however, the systematic reviews from Canada upon which the evidence was based, covered the ages birth to 4 years. This reflected differences in schooling systems in each country. It was the opinion of the Consensus Panel that it would be highly unlikely that we would have missed any studies among 5 year-olds that that would have changed the strength of the recommendation and we are confident that the same recommendations would apply. Most research on which this is based is on children before they reach school – this structural distinction of school versus not school is more important for the guidelines than the biological age of the children.

Because of the recent development of the ADOLOPMENT approach, there was some uncertainty among the leadership group as to whether it was being followed correctly. While we do believe there is merit in the approach, we would make the following suggestions: