Prevalence and impact of autism

The prevalence of autism in UK primary school-aged children is approximately 1–2%¹ and the effects of autism are well documented, including increased incidence of mental health disorders, most commonly anxiety. Approximately 40–90% of children with a diagnosis of autism spectrum disorder (ASD) also meet the criteria for at least one anxiety disorder.^2,3 The impacts on family life are significant. Parents report higher levels of stress⁴ and loss of earnings,⁵ and the lifetime cost to the UK economy of supporting an autistic individual is estimated at £0.92–1.5M.⁵ Relationship difficulties, particularly at school, are also commonly reported. Children with autism are significantly more likely than neurotypical peers to experience bullying⁶ and tend to have fewer friendships.⁷

Prevalence and impact of sensory processing difficulties in autism

Difficulty in processing sensory information and, in particular, extreme sensitivity or insensitivity to sensory input from the environment is common in autism,⁸ with prevalence estimates of 90–95%.^9–11 Such difficulties may exacerbate social communication deficits and increase the frequency of restrictive and repetitive behaviour, and may occur because of impaired regulation of central nervous system arousal.¹² Hyper-reactivity, reflecting the autonomic nervous system ‘fight or flight’ response, may result in behaviours such as aggression, hypervigilance or withdrawal (owing to poor tolerance of noise, touch, smell or movement), or additional ‘safe space’ needs. Hyporeactivity, in contrast, is characterised by reduced awareness of sensory stimuli within the environment.¹³ Impaired sensory processing may also result in poor motor control, affecting participation in daily life.

Sensory modulation difficulties (i.e. difficulty recognising and/or integrating sensory information) in children with autism probably pose substantial burden to children and families, limiting participation in leisure activities,^14,15 and are linked to problems with activities of daily living, such as eating, sleeping, dressing, toileting and personal hygiene.¹⁶ Such difficulties represent a long-term challenge for health services in terms of treating potential consequences, such as behaviours that challenge and mental health disorders. Awareness and management of sensory difficulties in mainstream educational settings is also likely to affect peer relationships and educational outcomes. The potential pathway of effect is unconfirmed [i.e. the mechanism(s) by which sensory difficulties affect key outcomes], but it is plausible that reducing sensory processing difficulties (SPDs) could lead to improvements across behavioural, social and educational domains.

Therapeutic approaches to treating sensory difficulties

The National Institute for Health and Care Excellence (NICE) guidelines on the management and support of children and young people with autism¹⁷ highlight parental perceptions of unmet need for occupational therapy (OT) input to address sensory and functional difficulties as part of the wider supportive network spanning education, health and social care. However, despite this clear unmet need, there is insufficient evidence to recommend any single therapeutic approach for SPDs. The NICE Guideline Development Group recommended further research to establish whether or not sensory integration therapy (SIT) improves SPDs in children and young people with autism across a range of contexts.

The primary goal of OT intervention is to enable participation in activities of everyday life¹⁸ by supporting people to engage in the occupations they need, want or are expected to do,¹⁹ considering their engagement and performance within relevant environments.²⁰ OT intervention for children with autism and SPDs is set within this context. There is a range of approaches used to achieve occupational goals, some of which focus on addressing underlying impairment (i.e. a ‘bottom-up’ approach²¹) and others on adapting the activity or environment for the individual to achieve these goals (i.e. a ‘top-down’ approach²²). Both approaches aim to improve occupational engagement, participation and performance, with the distinction that ‘bottom-up’ approaches address underlying SPDs to improve activities of daily living, whereas ‘top-down’ approaches focus on environmental support to accommodate sensory difficulties. These ‘top-down’ approaches include the use of sensory strategies that are not intended to address underlying neurological factors but focus, instead, on adapting the task or environment to accommodate the individual’s sensory needs, or on more targeted support (e.g. helping the child to use self-regulation to manage their sensory needs). ‘Bottom-up’ treatment approaches include Ayres Sensory Integration^® (ASI) therapy, which is a play-based intervention providing sensory–motor engagement in meaningful activities while adhering to fidelity principles,²³ and sensory-based interventions, which are adult-directed application of sensory input to effect change in behaviour linked to sensory modulation difficulties. These interventions are intended to influence how the child integrates sensory information to facilitate the development of adaptive responses in everyday life.

Sensory-based interventions are usually adult-directed sensory stimulation strategies applied to the child (i.e. without their active engagement) or made available to the child for regulation of their reactivity within the home or school environment. Sensory-based interventions typically focus on a single or narrow range of sensory modalities or techniques (e.g. use of weighted blankets, pressure vests, brushing and sitting on a ball). Adaptations to family routines and the environment may be suggested. Although some studies report positive effects of this approach, for example by coaching children and families to adapt routines and activities,²⁴ effectiveness evidence for functional performance is limited, particularly if strategies are not individualised to the child.¹³ However, sensory-based interventions, as well as sensory strategies and environmental adaptations, are currently reported by many professionals and carers as the most common form of ‘usual care’ in a UK setting. Treatment that meets the fidelity requirements for SIT (in terms of content and dose) is not often reported within the UK context, although evidence suggests that parents’/carers’ preference is for this more intensive type of intervention (see Chapter 2).

Sensory-based interventions

Case-Smith et al.¹³ systematically reviewed evidence (from 2000 to 2012) of sensory-based interventions and SIT for children with autism and SPDs. Nineteen studies were included in the review, including 14 studies of sensory-based interventions. Sensory-based interventions were defined as those that were based on individual assessment of the child’s sensory needs and functional performance, included explicit self-regulation goals and associated behavioural outcomes, and required the child to actively participate. Few positive effects were reported for studies evaluating sensory-based interventions.¹³ Thirteen of 14 studies eligible for review were multiple baseline single-case evaluations of weighted vests, therapy balls and different types of vestibular stimulation (e.g. swinging and bouncing). One study²⁵ found a positive effect for weighted vests on attention. There was limited evidence to support the use of therapy balls to increase sitting behaviours.^26,27 A further study²⁸ demonstrated a positive effect of a sensory diet (e.g. brushing, swinging and jumping) on self-regulation behaviours. Only one study²⁹ was a randomised controlled trial (RCT) (n = 30) of a ‘sensory diet’ protocol (i.e. exposure to a variety of different sensory experiences and practice of specific activities), which found positive effects in terms of a reduction in sensory difficulties overall, but did not assess fidelity or blind outcome assessors. The intervention also included several behavioural techniques (e.g. modelling, prompting and cueing) and, therefore, it is not possible to isolate the effects of sensory-based approaches.¹³ In summary, evidence to support the use of sensory-based interventions is limited in scope, methodology and generalisability.

Sensory integration therapy

Sensory integration therapy is a clinic-based approach that focuses on the therapist–child relationship and uses play-based sensory motor activities to address sensory–motor factors specific to the child to improve their ability to process and integrate sensation.³⁰ To distinguish SIT from other sensory-based interventions, a set of fidelity principles to guide delivery was developed and registered as ASI.^31,32 A fidelity measure has also been developed for use in research.²³ These principles ensure that underlying sensory–motor difficulties affecting activities of daily living are addressed through presenting a range of sensory opportunities and active engagement of the child in sensory–motor play at the ‘just-right’ level of challenge, within the context of a collaborative therapist–child relationship. Studies meeting ASI fidelity principles have been shown to lead to improvement in client-oriented goals,^33,34 but research is limited and, in some cases, interventions are poorly defined.¹³ Case-Smith et al.¹³ identified five studies specifically examining SIT, of which only two^33,34 were RCTs. SIT was described in these studies as ‘ . . . clinic-based interventions that use sensory-rich, child-directed activities to improve a child’s adaptive responses to sensory experiences’.¹³

Both RCTs^33,34 included in the review¹³ demonstrated positive effects of SIT on the Goal Attainment Scale (GAS). There were, however, methodological issues with both trials, including small sample sizes [i.e. no formal sample size calculation and use of convenience samples (n = 37³³ and n = 32³⁴)], lack of long-term follow-up^33,34 and limited description of usual care.³⁴ There are also well-documented methodological problems (with validity and reliability) related to the use of GAS as an outcome measure in clinical trials,^35,36 particularly in paediatric contexts.³⁷ However, more recent research indicates that GAS may be a promising approach to measuring effectiveness of psychosocial interventions in autism and some recommendations for optimising reliability, including the use of a standardised approach to writing GAS goals, have been made.³⁸ Nonetheless, uncertainty remains about the use of GAS as an objective outcome measure in the context of a clinical trial, despite the appeal of an individualised approach to measurement of what could be considered an individualised form of therapy.

The three remaining studies assessing the efficacy of SIT reported positive effects on behavioural outcomes linked to sensory difficulties, although it is difficult to draw meaningful conclusions given the use of non-randomised designs, very small sample sizes and insufficient descriptions of outcome measures.¹³ More recent systematic reviews^39,40 did not identify any additional RCTs, but reported positive effects on several functional, developmental^40,41 and play outcomes^39,42 in two small non-randomised pilot studies (n = 8⁴² and n = 20⁴¹) in which SIT was delivered with adequate fidelity.

Current evidence gaps and research priorities

The current evidence base to support use of SIT for children with autism is of low quality and insufficient to recommend treatment.¹⁷ Significant methodological issues are evident from studies conducted to date, including poorly described interventions that are unlikely to meet fidelity standards. Even in randomised studies with good fidelity of intervention delivery, of which several systematic reviews^13,39,40 have identified only two (i.e. Pfeiffer et al.³³ and Schaaf et al.³⁴), intervention protocols were variable in terms of dose and delivery period (e.g. 18 45-minute sessions delivered over 6 weeks in the Pfeiffer et al.³³ RCT and 30 sessions delivered three times per week for 10 weeks in the Schaaf et al.³⁴ RCT). Conclusions from all studies to date are limited by small convenience samples, poorly described comparators or definitions of what constitutes ‘usual care’ and a lack of long-term follow-up. The latter may be particularly important, given the focus of SIT on attention and learning rather than repetition of specific behaviours. Post-intervention treatment effects may be less specific, and we do not yet know whether or not SIT demonstrates sustained effects.¹³

There is also uncertainty around appropriate intervention targets and associated outcome measurement, given the focus of previous trials on goal attainment as the primary outcome of interest and associated psychometric challenges. Furthermore, aside from considerable variation in the intervention protocols, SIT is resource intensive and would require significant investment to be rolled out as a potential treatment option within the NHS. It is critical, therefore, to evaluate both the clinical effectiveness and the cost-effectiveness of SIT for children with autism and SPDs across a range of key outcomes, and to determine whether or not any effects are sustained in the longer term.

The SenITA trial

The main aims of the SenITA (SENsory Integration Therapy for sensory processing difficulties in children with Autism spectrum disorder) trial (see Chapter 3 for methods) were to (1) describe usual care in trial regions and clearly differentiate this from the proposed intervention (see Chapter 2); and (2) evaluate the clinical effectiveness of manualised ASI therapy (see Chapter 5) in a two-arm RCT for SPDs in young children with autism. The intervention was evaluated in terms of the impact on behavioural problems and adaptive skills, socialisation, carer stress and quality of life (see Chapter 6), and in terms of cost-effectiveness (see Chapter 7).

Participants with a range of autism and sensory symptom severity, as well as functional and cognitive ability, were recruited from NHS, educational and third-sector settings. The primary outcome time point was 6 months post randomisation, and was reassessed at 12 months to determine whether or not any observed effects were maintained in the longer term. An internal pilot (see Chapter 4) examined whether or not the intervention differed significantly in content or intensity from usual care and assessed recruitment and retention. Contamination, adherence and fidelity of intervention delivery were measured as part of the process evaluation (see Chapter 8). Experiences of the intervention and usual care were also explored via semistructured interviews with carers and therapists (see Chapter 9).