Open Research Online What works in community health education for adults with learning disabilities: A scoping review of the literature

Background: Research suggests there is insufficient good quality information regarding the effectiveness of health education aimed at adults with intellectual disabilities. By analysing the literature, this review aimed to identify what constituted effectiveness in this context. Method: Relevant evaluations were extracted from bibliographic databases according to pre-specified criteria. Papers were analysed using QSR NVivo 11 by developing a narrative synthesis and analytic framework that identified and explored text addressing the research question. Results: Twenty-two studies were included. The review identified two broad components of effective health education: mechanisms and context. Mechanisms included embedded programme flexibility, appropriate and accessible resources, and motivational delivery. An effective context included an accessible and supportive environment and longer term opportunities for reinforcement of learning. Conclusions: Important gaps in the literature highlighted a need for further research addressing community learning experiences of adults with intellectual disabilities as well as the effectiveness of infection prevention programmes. physical activity and diet; general health and health advocacy; wom-en's health; hygiene, infection prevention and oral health; and diabetes. Studies covered a range of different geographical areas: nine were conducted in the United States, seven in the UK and the remainder geographically dispersed across the world. Programme participants represented a wide range of demographic characteristics.

aimed at adults with intellectual disabilities (Bergström, Elinder, & Wihlman, 2014;Naaldenberg et al., 2013). Studies are diverse and explore a range of health issues but many have methodological weaknesses (Frankena, Naaldenberg, Cardol, Linehan, & Valk, 2015;Gerber et al., 2012;Scott & Havercamp, 2016). There is currently no overall analysis of these weaknesses in a health education context nor synthesis of the ways in which programmes are effective.
Research suggests that adults with intellectual disabilities bring different abilities and motivations to a learning environment and effective education needs to account for this (Bergström et al., 2014;Scott & Havercamp, 2016). However, there is currently no analysis of the ways in which this had been addressed in health education programmes nor in subsequent evaluation findings.
To address these issues, the focus of this review is effectiveness in health education for adults with intellectual disabilities particularly regarding self-care, infection prevention and the ongoing management of good health. Specifically, it aims to review and synthesize findings in the literature in this context, and answer the research question: What are the components of effective health education for adults with intellectual disabilities?

| ME THOD
The Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews, PRISMA-ScR (Tricco et al., 2018), was used to guide this review. The PRISMA-ScR is a checklist of items specifically developed to ensure effective reporting within a scoping review. An initial review of the literature sug- Citation tracking and manual searching of reference lists enabled a search for additional papers that were excluded in the bibliographic search results.
Studies were included in the search if they: • evaluated a health education, training or health promotion intervention for adults with intellectual disabilities aged 16 + that aims to address a change in health behaviour; • appeared as peer-reviewed articles in academic databases published in the English language between database inception and January 2019. International studies were included if they were published in English; or • included people with mild, moderate or profound intellectual disabilities and included papers with an additional focus on learning difficulties such as dyslexia, autism and Asperger's syndrome.
These were included to ensure full coverage of potential components of effective health education.
Studies were excluded if they: • focused on children (under 16 years) because the focus of the present review is adults or young adults learning in community settings; • focused solely on caregivers, supporters or staff because the research questions of the present review focus on adults with intellectual disabilities; or • only used physical outcome measures because physical outcome measures alone do not give an indication of education effectiveness. Figure 2 illustrates the search and selection process. All included papers were imported into QSR NVivo 11 for narrative synthesis of the literature (Popay et al., 2006). Narrative synthesis uses words F I G U R E 1 Database search example

Moderate
Results of interview addressing women's health knowledge, healthy behaviour beliefs, problem-solving and coping strategies regarding medical procedures.
Significant gains in health knowledge and behaviour beliefs (p < .05). Retained some of these at follow-up (p < .01).

Weak
Questions from various established instruments used to measure knowledge. Tested and refined standardized tools with two women with intellectual disabilities.
Moderate knowledge gain regarding breast cancer screening (p < .05).
16 Moderate and text to explore findings and can be an effective approach when meta-analysis is inappropriate (Campbell, Katikireddi, Sowden, & Thomson, 2019;Noyes et al., 2019). A meta-analysis is often used for reviewing groups of intervention studies (Higgins & Green, 2011) and usually includes randomized controlled trials. The body of literature on health education relating to intellectual disability is not developed enough to conduct this type of analysis. Therefore, this review includes non-randomized studies and by using narrative synthesis evaluates the evidence for each intervention individually.
Narrative synthesis was conducted using the research question as the basis of the analysis; the whole of each paper was analysed. First, text was identified that addressed the research question and therefore related to the delivery, experience and outcomes of the interventions undergoing evaluation. Secondly, themes in this text were identified inductively by reviewing findings in the studies that were relevant to the research question. Ongoing review and refinement of the analytic framework was conducted to ensure that earlier papers were further analysed in the light of new themes arising from subsequent articles.
This thematic analysis provided insight into the educational experience of participants as reported by the authors, and reviewed in detail the context of the intervention and of the study results.
The included studies were also summarized (see Table 1), and the following data were extracted into a Microsoft Excel spreadsheet: general study characteristics and evaluation results; intervention characteristics; quality assessment scores; and evaluation findings.

| Quality assessment
The Cochrane Centre acknowledges the difficulty in assessing the quality of public health and health promotion studies (Higgins & Green, 2011). This is largely due to the range of designs used and the complexity of appraising qualitative studies that often form part of such evaluations.
As Higgins and Green (2011) suggest, appraisal criteria will depend on the type of study undergoing review. Since different types of study were included in this review, two methods of quality assessment were used:

| Eight criteria of quality assessment developed by Naaldenberg et al. (2013)
Using this method, studies were scored according to the following characteristics: clear description of aim(s) and research question(s); description and discussion of rationale for sample size chosen, research population, attrition rate and measurements used; discussion of study limitations; and description of intervention development and content. In total, 16 points can be assigned to each paper, two points per criteria if information was provided and elaborated, one point if marginal information was provided and discussion or elaboration was lacking and zero points if no information was provided.

| The quality assessment tool for quantitative studies (Effective Public Health Practice Project, 2007)
This tool was developed to appraise any quantitative study design and involves assessment of selection bias, study design, confounders, blinding, data collection methods and withdrawals/dropouts. Most studies accepted a definition of intellectual disability that was used by the service provider from which the participant was recruited. This is defined in most papers as "mild to moderate intellectual disability," and further detail is generally not provided.

| Main study characteristics
However, Clark, Espie, and Paul (2001)  Eleven programmes focused specifically on health education.
The remainder included a health education component as part of a wider programme, which also included behaviour change techniques. Sixteen papers described community settings, four a group home environment and two educational establishments.
Most were small-group interventions, although one programme  studies contacted participants 12 months after completion of the programme (Table 1, studies 1 and 8). One study (Sandjojo et al., 2019) conducted several follow-up assessments.
The methods used to measure outcome within the studies included in this review are outlined briefly in Table 1. Most researchers used more than one outcome measure, and a combination of standardized and specifically developed tools. One paper did not specify how it measured outcome (Codling, 2015). These measures, however, provided some insight into the outcomes of the programmes evaluated (Tables 1, 3 and 4), and quality assessment indicated that most outcome measurement tools were valid and reliable. Nevertheless, sample sizes were small and several authors commented that further research was needed with larger samples.

| Research quality
Overall, the quality of the research is mixed. Seven were "moderate" and eight "weak." Five studies were not appropriate for this measure because they used qualitative methods or did not provide sufficient information to assess adequately. Those that rated as moderate or weak often did not provide sufficient information relating to key assessment criteria and therefore scored as "can't tell" which equated to a weak rating.  2,3,4,5,6,7,8,9,10,11,14,20,21 Creates a sound basis for education programme development.
Built-in flexibility and adaptability of course.
Participatory methods in course development.
Combination of accessible resources and techniques.

Education delivery
Motivational, sensitive and perceptive delivery.
1, 2, 3, 20 Participants enjoy learning and get the best out of it.
Appropriate methods and techniques of delivery.
Course fidelity (but a need for flexibility too).
Visual and interactive mechanisms (including tools to take away). assessment; the average score was 13.9 out of a total of 16. Seven studies scored the maximum 16. The lowest score was seven.
No studies were excluded on the basis of quality. There are two reasons for this. First, even those that rated as weak on the quality assessment displayed some interesting methodological issues or rel- A limitation of many studies in this review is the lack of longterm follow-up. Evidence suggests that people with intellectual disabilities learn best through regular reinforcement and repetition (Dunkley et al., 2017). None of the studies followed up participants beyond a year, and therefore, evidence of longer term effectiveness is not available. There is also little evidence in the papers about the process and experience of education from perspectives of the participants. This is addressed further in the Discussion.

TA B L E 4 Components of an effective learning context for adults with intellectual disabilities
Context Identified in study index number (see Table 1)

Why/how context supports effective health education for adults with intellectual disabilities
A positive physical environment-comfortable and opportunities for refreshments and breaks.
2, 5, 7, 15, 20 Encourages a good frame of mind and physical comfort for learning.
A social and supportive learning environment. 1,2,7,9,15,19,20 Encourages a good frame of mind and mutual support for learning, sharing ideas and asking questions.
Participatory and interactive ethos but not be intimidating. issues Most of the studies in this review provide data on learning outcomes without reflecting on the process of education and learning that produced this outcome. It is therefore difficult to fully evaluate the education and its effect. However, the studies do provide some interesting information regarding components of effective health education. These are presented as mechanisms for education delivery and the preferred contextual issues that accompany these mechanisms. Table 3 shows that mechanisms include an appropriate theoretical basis, built-in flexibility and adaptability, goal setting and action planning and motivational, sensitive and perceptive delivery.
The studies also highlight the importance of achieving the right context for effective health education to encourage and reinforce learning. Table 4 illustrates the important components here, which include a social and supportive learning environment, contextual adaptation appropriate to the course and participants, and opportunities for post-learning support.

| Mechanisms for effective health education accessible to adults with intellectual disabilities
Cognitive models are recognized as being important for education and behaviour change interventions (Naaldenberg et al., 2013). Many of the health education programmes in this review were developed using learning or cognitive development theories. However, most authors did not refer to theoretical underpinnings in the analysis of results and it is therefore difficult to determine the role of theory in the evaluation of these studies. It is possible to make some inferences. For example, the activity programme in Heller et al.'s (2004) study was based on a social learning model. Group activities were reported to be enjoyable. Relating health to participants' feelings led to better understanding of their health conditions, and the implication is that a social model can support health learning in this context.
The literature suggests that an effective health education programme for adults with intellectual disabilities needs to be either specifically developed or carefully adapted to meet learning needs (Heller et al., 2004;Taggart et al., 2018;Witton et al., 2017). Most of the programmes were specifically developed to address the needs of adults with intellectual disabilities although some were adapted from existing programmes aimed at a wider population. Adaptations included accessible resources such as easy read materials, videos and scenario-based role playing.
Flexibility is also required in terms of both the learning and the support environment. Taggart et al. (2018) and Witton et al. (2017) identified Learning reinforcement mechanisms such as certificates of attendance or completion (Dunkley et al., 2017;Eley et al., 2018), action planning and goal setting (Dunkley et al., 2017;Eley et al., 2018;Sandjojo et al., 2019), self-monitoring opportunities and activities such as diary-keeping (Dunkley et al., 2017) were effective in supporting knowledge retention and behaviour change. Use of repetition and recapping on sessions were also effectively applied (Dunkley et al., 2017;Eley et al., 2018;Feldman et al., 2016). For example, Hartwig et al. (2017) maintained weekly motivation to ensure oral hygiene was part of daily routines in four sequential sessions.

| 13
Published for the British Institute of Learning Disabilities OWENS Et al. Witton et al. (2017) found that regular demonstration and repetition enabled effective toothbrushing and therefore supported independent oral hygiene practice. Tools to take away, including diaries and homework, were also effective in embedding learning because they provided the opportunity to practise what was learnt (Dixon-Ibarra et al., 2017).

Mechanisms of effective education delivery include small-group
sessions, one-to-one support, a computer-assisted learning approach Some health topics may need a more specific approach. Whilst role play worked well for participants in Lunsky et al.'s (2003) study of a women's health education programme, Swaine et al. (2014) found that some concepts taught in education of breast and cervical cancer screening were not appropriate for role play. Instead, a "Building Skills" activity worked better, which involved scenario-based questioning. Lunsky et al. (2003) suggested that role play would not suit women with profound and multiple disabilities however, and staff intervention combined with small-group or individual therapy would be more appropriate.
Many of the papers in this review illustrate that a combination of resources such as videos, pictorial instructions, worksheets, role play, games and other interactive activities could increase health knowledge gain (Bodde et al., 2012;Dunkley et al., 2017;Feldman et al., 2016;Sandjojo et al., 2019;Swaine et al., 2014) Facilitators in the programme evaluated by Sandjojo et al. (2019) ensured that each participant's training was tailored to their abilities and preferences. They achieved this by continuously consulting with participants about how they would like to be trained. For example, if a participant was unable to read, sessions were presented more orally or visually with demonstrations, role play or video material. Witton et al. (2017) and Heller et al. (2004) included a peer approach to learning in their programmes. Peers were considered effective because participants could relate to them. In this context, Witton et al. (2017) described high levels of engagement and enjoyment and good acquisition of skills and knowledge.
The extent to which caregivers and supporters are effective facilitators in health education for people with intellectual disabilities can vary considerably. Some studies reported they had a positive effect on acquiring and embedding learning (Bergström et al., 2013;Dunkley et al., 2017;Hartwig et al., 2017;Sandjojo et al., 2019). Dunkley et al. (2017) found that care workers had a key role in helping to motivate and support participants to make and sustain changes to their diet and physical activity. Hartwig et al. (2017) reported that targeting caregivers could be a successful strategy to improve oral health.
However, supporters can also impede progress. Course leaders in Bergström et al. (2013) suggested that a course without caregivers' presence provided an opportunity for the participants to increase autonomy. The authors identified several instances where caregivers did not support participants in the programme. They suggested that more frequent communication about how best to support participants would have improved this. Bergström et al. (2013) also suggested that good collaboration with caregivers was important, however. This is also illustrated by findings in Dixon-Ibarra et al. (2017). Here, the authors refer to the importance of obtaining "buy-in" from staff and caregivers and suggested this was critical for behaviour change in the context of the programme they evaluated.
Achieving this is often difficult however; Lunsky et al. (2003) found that support workers offered an opportunity to support and embed learning but there were often challenges regarding regular staff turnover and shift work. Several studies identified fidelity as an important mechanism in supporting a consistent and effective programme (Bodde et al., 2012;Dixon-Ibarra et al., 2017;Feldman et al., 2016;Hartwig et al., 2017;Marks et al., 2013;Parish et al., 2012;Taggart et al., 2018). Fidelity is the extent to which a course is delivered as planned and adheres to its theory. In the context of health education, it is important because it enables consistent delivery as well as understanding of the causal mechanisms regarding behaviour change.
Educational follow-up was identified by Dunkley et al. (2017) and Taggart et al. (2018) as a mechanism that could support longer term embedding of learning although most of the interventions evaluated within this review only included follow-up for research purposes.

| A context for effective health education
A participant's personal context affects the knowledge, experience and motivation they bring to a learning environment (Bergström et al., 2014). Additionally, the learning environment provides context that influences participants' experience (Bergström et al., 2014;Bodde et al., 2012). In a study of barriers and facilitators in health education for adults with intellectual disabilities, Bergström et al. (2014) found that support from within and outside the formal learning environment was important, and supportive post-education context was subsequently important in embedding learning. In this respect, context can also support the reinforcement of learning and behaviour change.
Individual motivations and abilities are another important contextual component of successful learning, and this is reflected in the learning theory used by some of the programmes in this review. Several studies suggested that the motivations, abilities and interests of participants can affect learning outcomes (Bergström et al., 2013;Dixon-Ibarra et al., ,,2017Witton et al., 2017).
In a review of a dental ambassador programme, Witton et al. (2017) concluded that it was critical to design a programme that was realistic about the abilities and skills of the participants and could account for diverse needs.
Other important contextual components included a comfortable and safe learning environment, an ethos of participation, collaborative working and enjoyment (Heller et al., 2004;Witton et al., 2017). These generated high levels of engagement and subsequent acquisition of knowledge and skills. Additionally, familiarity and consistency can support effective learning (Dunkley et al., 2017). Dunkley et al. (2017) found that combining use of the same venue with a core group of educators ensured continuity and developed rapport. The authors suggested that educators should also gain an understanding of each participant prior to starting a course. This enables familiarization and development of plans to meet individual needs. It also enables the establishment of mutually agreed guidelines to support group functioning.

| CON CLUS IONS
This review aims to identify the effective components of health education for adults with intellectual disabilities that are raised in the existing literature in the context of self-care, infection prevention and the ongoing management of good health. Although the range of studies in this review is broad, authors identified similar conclusions regarding the components of effective health education. These include fidelity alongside flexibility of delivery, accessibility of materials, and a supportive and motivational learning context. The means of measuring the effectiveness of these components varied considerably however, and the focus tended to be on outcome evaluation. Further research is needed to explore the learning context as experienced by people with intellectual disabilities, particularly the education environment, the personal context of course participants and how this relates to their learning outcomes and behaviour change. Whilst the studies included in this review described some components of effectiveness and outcomes, few addressed in depth the learning experience of participants and its effect on outcomes. Further research is needed in health education evaluation to explore the learning context as experienced by adults with intellectual disabilities, particularly the education environment, the personal context of course participants and how this relates to their learning outcomes and behaviour change.
In this review, only Eley et al.'s (2018) pilot study addresses the substantive issues of self-care, infection prevention and antibiotic use. In order to ensure adults with intellectual disabilities can understand and address these concepts, more evidence is needed on the effectiveness of health education in improving knowledge and behaviour in this respect. This includes evidence on whether such programmes work in the short, medium and longer term and on the development of appropriate outcome measures. Additionally, the fact that few of the studies in this review used collaborative, inclusive or participatory research methods highlights a need for a better understanding of how such methods can shape good quality health education evaluation.

| Strengths and limitations
A major strength of this review is that it systematically analyses important issues in public health education accessible to adults with intellectual disabilities. By objectively selecting papers and analysing narrative text, it has been possible to identify what is effective in this context. This review also enabled comprehensive understanding of methods of measuring outcome, their strengths and limitations.
A potential limitation is that the data extraction and analysis was conducted by just one person. Whilst every effort was made to search for and include all relevant studies, the volume of papers was large and it is possible that a relevant study may have been overlooked.

ACK N OWLED G M ENTS
This work is conducted as part of a PhD studentship funded by the ESRC Grand Union Doctoral Training Partnership, co-funded by Public Health England and hosted by the Open University.