The microsegmentation of the autism spectrum: research project

10 The Escapable Costs Of Autism: Microsegmentation And Future Research And Provision For ASD In Scotland

10.1 In taking forward the recommendation of the Scottish Strategy for Autism from which this research arose – that previous work on the economic costs of autism (Järbrink & Knapp, 2001; Knapp, Romeo, & Beecham, 2009) should be analysed and applied to the Scottish context – the aim was to inform strategy and planning on what interventions might ‘lead to positive impacts both for individuals and for the economy as a whole’ (Scottish Government, 2011, p.77). That is, a primary purpose of the research was to provide a reliable foundation for identifying those costs of autism which may be ‘escapable’ and which would not be incurred with appropriate interventions for individuals on the spectrum.

10.2 This report has provided an economic analysis of the cost of autism in Scotland, informed by the most accurate estimates of prevalence of autism spectrum disorders and distribution of intellectual ability and disability across the spectrum. The construction of an extensive dataset from a large-scale sample of individuals with ASD and their parents and carers, together with adapting findings from the world literature, has made it possible to have detailed economic costings with high potential utility not only at national level but also at Council and Health Board level to assist budgetary planning and the planning of service provision.

10.3 On the basis of the data collected and from all relevant current research literature on autism, we have also been able to construct a meaningful microsegmentation of the autism spectrum, presented in the form of a microsegmentation matrix, using intellectual status and symptom severity as stable moderators and co-occurring conditions and associated features as variable moderators of outcome in terms of life trajectories and economic costs. This provides an evidence-based approach to understanding the spectrum which could not be achieved through current understandings of diagnostic subgroups, assessment profiles or co-occurring conditions.

10.4 From the foundation provided by this research we are now able to consider the question of the escapable costs of autism not only with particular application to the Scottish context, but also with a focus on those costs which are bigger and potentially easier to address, both in terms of economic impacts and in relation to the quality of life of people with autism and their parents and carers.

Quality of life

10.5 While the focus of this research has been on economic impacts, we emphasise the importance of also considering quality of life alongside costs, both in terms of the overarching aims of the Scottish Strategy for Autism and in terms of the relationship between costs and quality of life.

10.6 All of the recommendations of the Scottish Strategy for Autism, including those that focussed mainly on economic impacts, were designed with quality of life in mind. In the opening paragraph of its introduction, the Strategy document referred to ‘a series of 26 recommendations about how to improve support in order to improve the quality of life of people with autism. The recommendations honour the vision and values which underpin the autism strategy and have the wellbeing of people with autism as central and fundamental’ (p.20).

10.7 The strong links between economic impacts and quality of life have been demonstrated over a long period in a wide variety of studies. They may be illustrated here with reference to two examples which have relevance to the autism spectrum, namely, the academic study of ‘happiness’ (or wellbeing) and the study of the economic impacts of mental disorders.

10.8 It is generally recognised that quality of life is a multidimensional concept which integrates both objective and subjective indicators, and which includes the domains of physical, mental, material and social wellbeing (see, for example, Felce & Perry, 1995). While happiness is a subjective state of wellbeing, mental disorders include objective aspects and are defined by the World Health Organization (1992) as ‘a clinically recognisable set of symptoms or behaviour associated in most cases with distress or interference with personal functions’ (p.92).

10.9 With regard to happiness, one important field of research relating to its economic benefits has been employment. For example, Gavin and Mason (2004) consider happiness in terms of its positive impact in reducing occupational burnout, boredom, disillusionment and sabotage. They examine specific occupational contexts, illustrating the complementarity of economic impacts and quality of life. Oswald, Proto and Sgroi (2015), in a series of four experiments with a large sample (n=713), reported that employee happiness led to productivity increases of 12%. Conversely, in studying the impact of major real-world shocks (bereavement and family illness), they reported that lower happiness is systematically associated with lower productivity. They proposed a causal link between wellbeing and work performance. These factors are of central relevance to ASD in view of the high frequency of mental health problems and the difficulties many people on the spectrum have in maintaining employment.

10.10 Mental disorders, and more broadly the wide range of issues linked to low mood and a general sense of poor mental wellbeing, have major economic impacts. These are estimated at £70-£100 billion per year for the UK ( OECD, 2014). The relationship between mental health and economic circumstances is recognised by the World Health Organization (2001) in its definition of mental health as ‘a state of well-being in which every individual realises his or her own potential, can cope with the normal stresses of life, can work productively and fruitfully, and is able to make a contribution to her or his community’ (p.1).

10.11 The highly elevated occurrence of mental health issues across the autism spectrum, particularly in relation to the prevalence of anxiety and depression, are discussed in Chapter 3 and are highlighted throughout this report. These and other issues relating to the individual’s overall wellbeing clearly have economic impacts which were recognised by parents, carers and the individuals themselves in the thematic data analysis reported in Chapter 7.

Autism: the inescapable costs

10.12 There are some costs of autism which may be viewed as inescapable. These relate to matters which are of a sufficiently fixed or static nature that they cannot be modified by any intervention framework available at the present time. We list here three factors which may be considered as the inescapable costs of autism.

10.13 First, the prevalence of autism may be viewed as representing an inescapable cost. There are at this stage no ‘cures’ for autism, in the sense that it is a biologically-based neurodevelopmental disorder which in not preventable. While we have noted that there are reports of a small number of people with ASD who later ‘lose their diagnosis’, or who otherwise have such favourable outcomes that they are no longer autism service users ( para. 5.6), for all practical purposes autism may nevertheless be viewed as a lifelong condition. Budgetary and service planning should therefore be based on the recommended prevalence figure we have proposed as being a stable factor.

10.14 Second, the occurrence of intellectual disability within the autism spectrum may be viewed as representing an inescapable cost. Again it is recognised that intellectual disability at its various levels is determined by thresholds, and that it also depends on practical judgements on the level of adaptive behaviour an individual is able to demonstrate. People with intellectual disability may vary across the lifespan at the margins of these thresholds, or their adaptive behaviour may decline, or it may be enhanced through interventions. However, intellectual disability may also for all practical purposes be viewed as a fixed factor in terms of budgetary and service planning.

10.15 Third, there are other conditions which co-occur with autism which may be viewed, in whole or in part, as representing an inescapable cost. Common co-occurring conditions, as discussed in detail in Chapter 3, include epilepsy, attention deficit hyperactivity disorder ( ADHD), schizophrenia, obsessive compulsive disorder ( OCD), Tourette’s Syndrome and anxiety and depressive disorders. These vary in the extent to which they are biologically determined or reactive to life circumstances. They also vary in the extent to which they can be ameliorated by pharmacological or psychotherapeutic interventions. Nevertheless, the co-occurring conditions include factors which may be viewed as being of a fixed nature.

Autism and evidence-based interventions

10.16 The question of interventions for autism has been the subject of a very extensive literature comprising studies and reviews at every level from exploratory single case studies to meta-analyses of randomised controlled trials ( RCTs). The overall field has been analysed and summarised in an iterative way across multiple research reviews and good practice guidelines. These include: the Comparative Effectiveness Reviews of the US Effective Health Care Program (Taylor et al., 2012; Warren et al., 2011) and in a UK context the NICE Guidelines (National Institute for Health and Clinical Excellence, 2012; National Institute for Health and Care Excellence, 2013), and most recently the National Autism Project Report (Iemmi, Knapp, & Ragan, 2017).

10.17 Internationally recognised research reviews providing good practice guidance have also been produced specifically in the Scottish context. The first SIGN Guideline on autism (Scottish Intercollegiate Guidelines Network, 2007) covered assessment, diagnosis and clinical interventions in children and young people with autism spectrum disorders. This was extensively revised to provide a new review of research covering not only children and young people, but also adults across the whole age span including older adults with ASD (Scottish Intercollegiate Guidelines Network, 2016).

10.18 In terms of the research criteria by which evidence is judged in reviews and guidelines at this level, it must be acknowledged that at the present time the evidence for autism interventions is weak and does not provide a basis for making clear recommendations on cost-effectiveness comparable to other fields of intervention. For example, The Lancet in 2011 published a series of four papers that examined what is known regarding the impact of obesity internationally. On the basis of all the evidence gathered, the focus of the final paper was on ‘changing the future of obesity’ through coordinating science, policy and action around promoting evidence-based and cost-effective interventions (Gortmaker et al., 2011). Eight interventions were found to be both health-improving and cost-saving at a level that met pre-determined criteria, while a further six showed evidence of benefits at a lower level. Evidence of this nature – that is, evidence of interventions that can be specifically shown both to improve ASD outcomes and also to be cost-saving – is very limited for autism research.

10.19 In terms of the Comparative Effectiveness Reviews, Warren et al. (2011) reported on a search of 4,120 nonduplicate citations for autism interventions for children aged 2-12 years. This yielded 159 unique studies which met final inclusion criteria. These covered behavioural interventions, educational interventions, medical and related interventions, allied health interventions and CAM (complementary and alternative medicine) interventions. Their conclusion was that some pharmacological interventions such as risperidone and aripiprazole show benefit for reducing challenging behaviours in some children with ASD, but side-effects are significant. Some behavioural and educational interventions that vary widely in terms of scope, target, and intensity have demonstrated effects, but the lack of consistent data limits understanding of whether these interventions are linked to specific clinically meaningful changes in functioning.

10.20 Behavioural interventions represented approximately half of the studies considered. However, few RCTs of sufficient quality had been conducted, no studies directly compared effects of different treatment approaches and little evidence of practical effectiveness or feasibility existed. While studies of UCLA (University of California Los Angeles)/Lovaas-based interventions reported greater improvements than broadly defined eclectic treatments available in the community, strength of evidence was low. Although positive results were reported for the effects of intensive interventions using a developmental framework, such as the Early Start Denver Model, evidence for this type of intervention was insufficient because few studies had been published to date. The evidence base for less intensive behavioural interventions focussing on providing parent training remained insufficient. Social skills interventions reported some positive results (see, for example, Gates, Kang, & Lerner, 2017, for a recent review) but strength of evidence was insufficient to assess effects on core autism outcome for older children or play- and interaction-based approaches for younger children. Similarly, while cognitive behavioural interventions seemed effective in reducing anxiety symptoms, strength of evidence was viewed as insufficient in terms of number and quality of available studies.

10.21 Specific educational interventions such as TEACCH had insufficient evidence because of too few studies or inconsistency in the outcomes measured. Most of the TEACCH research was conducted prior to the cut-off date of the Warren et al. (2011) study, and newer studies continued to report improvements. Although no current medical interventions demonstrated clear benefit for social or communication symptoms, a few medications showed benefit for repetitive behaviours or associated symptoms, and the clearest evidence favoured the use of medications to address challenging behaviours. However, their usefulness was limited by large side-effects. Allied health interventions had little support for their use. Studies of sensory integration and music therapy were of poor quality, and auditory integration studies showed no improvements associated with treatment. Some language and communication interventions (Picture Exchange Communication System [ PECS] and Responsive Education and Prelinguistic Milieu Training [ RPMT]) demonstrated short-term improvements and were considered worthy of further study. No evidence was found in support of interventions based on complementary and alternative medicine.

10.22 In a further Comparative Effectiveness Review for autism interventions for adolescents and young adults aged 13-30 years, Taylor et al. (2012) reported on a search of 4,855 non-duplicate citations. From this number, only 32 studies met their final inclusion criteria. They noted that even most of these remaining studies were of poor quality. Five studies, mainly of medical interventions, were of fair quality, and none was rated as good. They concluded that few studies have been conducted to assess treatment approaches for adolescents and young adults with ASD, and as such there is very little evidence available for specific treatment approaches in this population. This was especially the case for evidence-based approaches to support the transition of youth with autism into adulthood. Behavioural, educational, and adaptive/life skills studies were typically small and short-term and suggested some potential improvements in social skills and functional behaviour. Small studies suggested that vocational programmes may increase employment success for some individuals.

10.23 The lack of high-level evidence for the efficacy of ASD interventions may be illustrated by reference to the grades of recommendations given in the first of the SIGN Guidelines (Scottish Intercollegiate Guidelines Network, 2007). A recommendation at Grade A represents an intervention for which there is at least one meta-analysis, systematic review or RCT rated as 1++ (high quality meta-analyses, systematic reviews of RCTs, or RCTs with a very low risk of bias), and which is directly applicable to the target population; or, a body of evidence consisting principally of studies rated as 1+ (well conducted meta-analyses, systematic reviews or RCTs with a low risk of bias), directly applicable to the target population and demonstrating overall consistency of results. Only four intervention recommendations were made at Grade A – and all of these were for what there was strong evidence not to use: comprehensive applied behaviour analysis programmes on the pretence that they should lead to ‘normal functioning’, auditory integration training, ‘facilitated communication’ and secretin (a hormone hypothesised to ameliorate autistic behaviour). The majority of positive recommendations were at the lowest level of evidence, namely, ‘recommended best practice based on the clinical experience of the guideline development group’.

10.24 The issue of assessing the level of evidence required to support the effectiveness of an intervention is of particular importance in relation to autism interventions. The methodologies underlying the pursuit of evidence-based practice have generally concurred in assessing what constitutes the meaning of evidence in terms of a hierarchy which may be summarised broadly from highest to lowest level as: systematic review of RCTs, high-quality RCT, low-quality RCT, outcome evaluations, controlled single case studies, case series and expert opinion (see for example, Reynolds, 2000; Scottish Intercollegiate Guidelines Network, 2016).

10.25 While interventions such as medications lend themselves to evaluations based on large-scale, double-blind, RCTs, too few good quality RCTs have been conducted for ASD. The same issue is experienced across this type of intervention in general, in which ‘there are relatively few randomly controlled trials, large quantitative studies or evaluations of experimental interventions’ and where ‘reviews of “effective practice” through visits, case studies and reports (the “grey literature”) will also assist in identifying research priorities’ (Davies, Nutley, & Smith, 2000, pp.242-243). As Barrett and Ollendick (2004) have noted, ‘Our present overview of empirically supported psychosocial treatments… reveals that our armamentarium is relatively “light” and… we really do not have very many psychosocial treatments that possess well-established status in research settings let alone clinical settings’ (p.21). While that observation was made a number of years ago, and there has been a helpful accumulation of new studies since that time, the evidence continues to be relatively light (see, for example, Iemmi et al., 2017).

10.26 It was this recognition that guided the work which took forward Recommendation 10 of the Scottish Strategy for Autism, ‘It is recommended that agencies and services develop a menu of interventions including advice, therapeutic interventions and counselling for children, young people and adults with an ASD, that are appropriate and flexible to individual need. This menu should identify advice and support that is immediately available, and set out the referral and assessment process for all other services and interventions’. The group which prepared the ‘menu’ noted that the difficulties surrounding this area arose ‘not just from the proliferation of interventions on offer but also, and most particularly, from the lack of interventions which have a good evidence base’ (Neil-MacLachlan and Members of Group 3, 2013, p.28). For that reason a different approach was taken, one of emphasising the need to move from practice into theory, beginning by looking at the challenges people on the autism spectrum face, the needs arising from these challenges, the types of service provision required to address these needs and the gaps in existing services.

Making an economic case for interventions

10.27 The high and wide-ranging costs of autism represent a mix of what could be called appropriate and inappropriate economic impacts: ‘appropriate’ in so far as evidence-based interventions are utilised by the right people at the right time, and ‘inappropriate’ in that some costs result from avoidable crises or because interventions are made available to autistic people too late, or what is provided is simply not effective. Because public and private resources are always scarce relative to the range of uses to which they could be applied (i.e. relative to the many demands for them), it is important to understand not only the costs of different interventions for autistic people and their families, but also their cost-effectiveness.

10.28 First and foremost, interventions that are funded from public resources must be effective in the sense that they meet needs, or improve personal functioning or improve quality of life. For interventions also to make economic sense, they need to be feasible in that they only employ resources that are available (such as there being enough professionals trained in the right therapeutic approaches). They also need to be affordable within current budget constraints. Third, they need to be cost-effective, which means that their outcomes are sufficient to justify the resources that must be spent to generate them. This does not mean that an intervention needs to be cost- saving, but rather that if the intervention costs more than its best alternative (or more than what is currently provided), then the higher costs are considered by decision-makers to be ‘worth’ incurring because of the scale and nature of the effectiveness gains.

10.29 There is now a small body of evidence in the international literature on the cost-effectiveness of interventions for autistic people and their families, although far less is known than is needed. Some of that evidence has been generated from UK studies, and some is broadly applicable to the UK even though the research has been carried out abroad. We can pull out some of the main findings; the recent report from the National Autism Project provides fuller details (Iemmi et al 2017). What these studies suggest is that available public and private resources could be better used than they are currently, if more funds were directed towards interventions that have been shown to be effective and cost-effective. However, the overwhelming message is that there is still not enough known about what works for autistic people or whether these interventions represent a good use of public finances or private expenditure.

10.30 The National Audit Office in England (2009) carried out simulation modelling to explore the potential economic benefits of multi-disciplinary teams to identify and assess autistic adults, concluding that substantial economic gains might be achievable even with modest increases in identification rate (National Audit Office, 2009). There have been no other cost-effectiveness or related economic studies of approaches to identification or diagnosis of autism, or of ways to carry out assessments of needs, strengths and preferences. The Scottish Strategy for Autism made the identification and assessment of the autistic population the focus of five recommendations and one of its strategic outcomes (Scottish Government, 2011, 2015). The Autism ACHIEVE Alliance ( AAA) mapped the services that provide diagnostic assessment of ASD in Scotland (Autism ACHIEVE Alliance, 2012) and identified long waiting times for diagnostic assessment for both children and adults (Autism ACHIEVE Alliance, 2014). Timely identification and diagnosis could benefit not only children, but also adults with ASD.

10.31 A great many approaches to early intervention have been proposed in the autism field. Most have been evaluated, although relatively few from an economic point of view, and rarely in the UK. Some programmes require quite intensive inputs from skilled therapists over quite long periods, such as the ESDM approach, which may make them appear unaffordable in a constrained fiscal context. Nevertheless, a Canadian modelling study concluded that ESDM could be cost-effective. Modelling studies for another intensive intervention, the Early Intensive Behavioural Intervention ( EIBI), were reviewed by NICE and found to be methodologically weak. The Preschool Autism Communication Trial ( PACT) for autistic children in the UK showed significant effectiveness gains at both 13-month and 6-year follow-ups, but the within-trial economic evaluation at the 13-month point did not find it to be cost-effective (£4,105 per child, 2006-2007 price levels) (Green et al., 2010; Pickles et al., 2016). Therefore, while there might appear to be an overall prima facie case for effective early interventions being capable of heading off later costs and improving longer-term quality of life, there is as yet no clearly demonstrated evidence of economic gains. The Scottish Strategy for Autism recommended the use of early interventions built upon the four principles identified in the Early Years Framework (Scottish Government 2008, 2013). The recent SIGN guideline on assessment, diagnosis and interventions for ASD recommended parent-mediated early interventions (Scottish Intercollegiate Guidelines Network 2016). Early interventions could benefit autistic children, both with and without ID.

10.32 Employment is a major challenge for many autistic people, as shown very clearly from the survey. Supported employment schemes provide individualised training and workplace support through job coaches, and often involve a range of stakeholders whilst aiming to take account of the individual strengths and preferences of autistic people. UK research shows that supported employment can be both effective and (strongly) cost-effective from a societal perspective, and has important economic benefits for autistic people themselves. NICE concluded that supported employment for autistic adults without ID was cost-effective from a health and social care perspective, costing £5,600 per quality-adjusted life year ( QALY) (Mavranezouli et al., 2012). The intervention was found to be cost-reducing from a societal perspective, when also considering productivity gains for both autistic people and their carers (Iemmi et al., 2017). Already highlighted by the Scottish Strategy for Autism as a potentially promising intervention (Scottish Government, 2008) and recommended in the delivery plan for A Fairer Scotland for Disabled People (Scottish Government, 2016), supported employment schemes could benefit autistic adults without ID, who could then need less support from other services and contribute to the economy. Examples of employment support schemes in Scotland have been described by the Autism Initiatives (2013).

10.33 Support for employment should include support to enable people with ASD to travel independently. The data from the Scottish Autism Survey indicated that those who were able to travel independently were several times more likely to be in employment than those who lacked ability for independent travel. Of those with a diagnosis of Asperger’s Syndrome or high functioning autism who were in employment, only 16% were unable to travel independently.

10.34 Another area where there is some economic evidence from the UK is in relation to parent training and support programmes, albeit from relatively small studies. The evidence suggest that there are inexpensive group interventions for parents of autistic children (such as Cygnet, ASCEND and Riding the Rapids) that can be effective, at least for the short time periods over which they were evaluated, and probably cost-effective. The cost per person varied widely across and within programmes: Cygnet at £351 (ranging from £141 to £663, 2009/10 price levels) (Stuttard, Beresford, Clarke, Beecham, & Morris, 2016), ASCEND at £615 per person (ranging from £201 and £2,543, 2009/10 prices) (Stuttard, Beresford, Clarke, Beecham, & Morris, 2012), and Riding the Rapids at £407 (ranging from £80 to £685, 2009/10 prices) (Stuttard, Beresford, Clarke, Beecham, & Curtis, 2015). Parent training and support programmes could benefit families of autistic children and adults, both with and without ID.

10.35 Cognitive behavioural therapy ( CBT) to treat anxiety problems experienced by autistic adults has been found to be effective, and – when delivered on a group basis – also cost-effective. NICE calculated group-based CBT for autistic children without ID to be cost-effective from a health and social care perspective, costing £13,910 per QALY (National Institute for Health and Care Excellence, 2013). Group-based CBT was found to be even more cost-effective when viewed from a societal perspective, i.e. after additionally taking account of productivity gains for their carers (Iemmi et al., 2017). Under the same perspective, individual CBT was also likely to be cost-effective, costing £31,050 per QALY (Iemmi et al., 2017). CBT could benefit autistic children and adults without ID.

10.36 A limitation of cognitive behavioural therapy for many people with autism is that the standard CBT protocol relies heavily on skills which are generally weak in ASD, including normal levels of empathy, ability to differentiate emotions in oneself and others, theory of mind in terms of being able to reflect on the thoughts, behaviour and intentions of others and strong central coherence in having an ability to generalise from specific situations to the wider context. For this reason it is necessary that therapists should have expertise in adapting the standard CBT protocol to meet the needs of people with autism. The NICE Guideline on recognition, referral, diagnosis and management of adults on the autism spectrum recommended that the adaptation of CBT for autism could make effective interventions more widely available (National Institute for Health and Clinical Excellence, 2012). Specifically they recommended a more concrete and structure approach with greater use of written and visual information, making rules explicit and explaining their content, avoiding excessive use of metaphor, ambiguity and hypothetical situations and incorporating the individual’s special interests into therapy.

10.37 There are also interventions that emphasise personalised approaches – such as positive behavioural support, Circles of support and personal budgets – for which there is some short-term evidence of effectiveness in wider populations in the UK (such as for people with learning disabilities) and also evidence of cost-effectiveness. One of the values underpinning the Scottish Autism Strategy is choice according to which ‘care and support should be personalised and based on the identified needs and wishes of the individual’ (Scottish Government 2008, p. 9). Personalised approaches could benefit the entire autistic population, but even if research shows that an intervention can have such wide-ranging benefits for many autistic people, and could also possibly be economically attractive, it is still necessary to tailor the nature of the action to reflect differences in need and strengths, and to respond to individual preferences.

10.38 Regular health checks can help to address the issue of premature mortality for autistic people, for example from cancer or coronary heart disease, because of poor access to healthcare and limited service provision. There is little autism-specific evidence of the benefits of health checks, but studies of people with learning disabilities are promising, and show clear cost-effectiveness gains. Two of those studies evaluated nurse-led heath checks for adults with ID in Scotland (Romeo et al., 2009; Cooper et al., 2014). Another value underpinning the Scottish Autism Strategy is equality and diversity, according to which ‘people should have equal access to information assessment and services; health and social care agencies should work to redress inequalities and challenge discrimination’ (Scottish Government 2008, p. 9). Health checks could benefit the entire autistic population.

10.39 There are many interventions for which there is, as yet, no robust economic evidence. This applies to social skills interventions (such as LEAP or TEACCH) and pharmacological interventions for treating co-occurring mental health problems. There have also been no robust economic studies yet in the rapidly developing area of assistive devices and technologies, even though this field offers considerable promise in the longer term. Economic evidence on efforts such as campaigns to address stigma or to prevent bullying, whilst now accumulating in the mental health field, has not yet been gathered for autism-specific interventions.

10.40 Finally, it is not known whether people with ASD are over-represented in the criminal justice system or in the prison population, but they do have a number of predisposing features which lead a significant number to commit a wide range of offences (see King & Murphy, 2014, for a comprehensive review of this field). While again there is a lack of economic evidence for interventions in relation to criminal justice and ASD, it is known that the criminal justice system and life in prison incur high public costs. A number of interventions have shown promise here, such as the use of autism alert cards, and also the provision of autism-specific support to individuals who have already become involved with criminal justice, and these merit further investment and research.

10.41 It is not possible in terms of the current evidence base to quantify the savings that might be achieved in relation to any particular intervention with potential economic benefits. By way of illustration, a number of examples may serve to indicate what savings would be achieved annually in Scotland in terms of several different scenarios involving cost-effective interventions for children and for adults, with and without intellectual disability, and for the total autistic population.

10.42 In terms of children with autism, for each percentage point by which evidence-based interventions reduced total costs there would be potential savings of more than £1.5 million annually in Scotland (£886,846 for children with intellectual disability, and £619,421 for those without intellectual disability). A reduction in costs by five percentage points would bring annual savings of more than £7.5 million (£4,434,230 for those with intellectual disability, and £3,097,103 for those without), while if a 10% reduction could be achieved there would be annual savings of more than £15,000,000 (£8,868,459 for those with intellectual disability, and £6,194,207 for those without).

10.43 In terms of adults with autism, for each percentage point by which evidence-based interventions reduced total costs there would be potential savings of around £21,000,000 annually in Scotland (£9,573,545 for adults with intellectual disability, and £11,211,538 for those without intellectual disability). A reduction in costs by five percentage points would bring annual savings of around £104,000,000 (£47,867,727 for those with intellectual disability, and £56,058,191 for those without), while if a 10% reduction could be achieved there would be annual savings of around £208,000,000 (£95,735,454 for those with intellectual disability, and £112,116,383 for those without).

10.44 In terms of the total autistic population, for each percentage point by which evidence-based interventions reduced total costs there would be potential savings of around £22,000,000 annually in Scotland. A reduction in costs by five percentage points would bring annual savings of around £111,000,000, while if a 10% reduction could be achieved there would be annual savings of around £223,000,000.