Cited By :92

|

Export Date: 8 November 2019

|

Funding details: Medical Research Council, MR/K005863/1

|

Funding details: Menzies Centre for Australian Studies, King's College London, University of London

|

Funding details: National Institute for Health Research

|

Funding details: National Institute for Health Research

|

Funding details: South London and Maudsley NHS Foundation Trust

|

Funding details: University of Manchester

|

Funding text 1: We sought to assess the long-term outcomes of the preschool PACT intervention on autism symptoms and other outcomes. The results of our investigation show a treatment effect to reduce autism symptom severity at treatment endpoint, which remained almost 6 years later, giving a clear averaged treatment effect over the total period. The effect was apparent across both autism social-communication and repetitive symptom domains. A similar treatment effect is also seen in parent-reported symptom measures at follow-up which, although unblinded, have the potential complementary strengths of being service-user outcome measures and being based on knowledge of the child in naturalistic settings. To our knowledge, this study is the first study to report long-term symptom outcomes to middle childhood (7–11 years) following a randomised controlled trial of early intervention in young children. As such, our results extend the findings from the smaller study of Estes and colleagues, with a shorter follow-up period of 2 years 11 that suggested follow-up effects on autism symptoms. Notably, the PACT intervention is substantially less time intensive than is Estes and colleagues intervention (ESDM) in terms of therapist treatment hours. Taken together, these results are encouraging and provide evidence that sustained changes in autism symptoms can be possible after early intervention, something that has previously been regarded as difficult to achieve. Our findings also have substantial implications for developmental science, given that a sustained effect of this type on targeted outcomes for a prolonged period after the end of treatment is very uncommon in developmental interventions. PACT is designed to work with parents to reduce autism symptoms through optimising naturalistic parent–child social communication in the home setting. The theoretical advantage of this approach over direct therapist–child therapy is that it has potential for change in the everyday life of the child, in which much social learning takes place, and might continue to have sustained effects after the end of the therapist-delivered intervention to the parent. Evidence from the initial trial mediation analysis 16 supported this theoretical model of action by showing that it was the treatment effects on parent interaction style that were responsible for the positive child change during the treatment period. The current follow-up analysis shows a loss of this original treatment effect on parental synchrony over time, but maintenance of the effects on child communication and symptom-change. This finding suggests that the improvement in child communication and autism symptoms during treatment could have become self-sustaining during the years following the end of treatment, independent of the initial parental behavioural change that mediated them. This finding lends support to the theoretical rationale behind a developmental approach in which targeting of pivotal precursor social communication skills can lead to improvements in developmental trajectories, further supporting the notion that a parent-mediated approach can lead to sustained effects beyond the end of treatment. Further empirical study of the mechanism behind such maintained change is now needed. Before the follow-up study, we hypothesised that the PACT intervention would lead to sustained reduction in restricted and repetitive behaviours of the child. Testing of this hypothesis in a prespecified analysis produced another important finding: the intervention, which was targeted at improving child social communication, also had marked cross-domain effects on the restricted and repetitive behaviours aspect of autism, seen both at endpoint 6 and at follow-up, with the follow-up result reflected in ADOS CSS and parent-reported RBQ scores. We argue that this finding supports a developmental account of the emergence of restricted and repetitive behaviours as at least partly dependent on a lack of effective social communication during the development of individuals with autism. However, our related hypothesis that levels of child anxiety, which are often linked to levels of restricted and repetitive behaviours in autism, 17 would also show a treatment effect at follow-up was refuted ( figure 2 ). Amelioration of the defining symptoms of autism is positive in its own right; but, in longitudinal autism cohort studies, ADOS symptom trajectories also predict concurrent adaptive functioning and later long-range adjustment through to young adulthood. 31,32 Estes and colleagues reported an effect on parent-rated adaptive function during their 2 year follow-up. 11 Our measures designed to address this aspect of autism were blinded teacher-rating of adaptive function in school, and non-blinded reports by parents of child adaptive function, peer relationships, and prosocial behavior. Point estimates of treatment effect on these measures range from modest to substantial size in favour of treatment, but CIs for the effect sizes were wide and included null or negative effects in each case ( figure 2 ). Consequently we cannot be sure at this time of the existence or extent of the effect of symptom changes on general adaptation. We found no evidence of an overall effect on comorbid mental health problems; the only apparent effect, a small point estimate of effect on anxiety or OCD symptoms, had a wide CI that included both some negative and positive effect. These findings suggest that, at the least, additional strategies will be needed if broader adaptive function and mental health in autism at later ages is to be further improved. The high rate of follow-up of what was already the largest randomised intervention cohort of its kind to address core autism features represents a key strength of this study; follow-up ascertainment was achieved for almost 80% of the recruited sample at 82 months after initial randomisation. Key measures were assessed in a blinded manner and consistently across time-points, although parent reports of repetitive behaviours and peer functioning were at follow-up only. Between-group demographic imbalances in the follow-up sample were adjusted for in the analysis. We note that our inclusion criteria were for core-defined autism rather than broader autism spectrum disorder; we cannot be sure how our results would generalise to young children with less severe symptoms. This study advances previous work by showing that a theoretically derived, developmentally targeted early intervention can have a sustained effect on autism symptom outcomes nearly 6 years after the end of treatment. In addition to replication, further research is needed to elucidate the developmental mechanisms behind such sustained change, as well as the extent of and barriers to wider developmental benefits, and the cost-effectiveness of such interventions over longer-term development. On the basis of these results, we are now able to support the use of the PACT intervention for reducing symptoms of autism in young children, a revision of our initial view 6 and consistent with the results of a subsequent UK NICE meta-analysis. 8 Contributors JG, AP, ALC, KL, SB, CA, VS, HM, PH, JP, and TC, conceived and designed the study. JG led the study. JG, TC, ALC, and HM led the research teams at each study site. KL, ES, RC-F, HT, IG, and JL traced and recruited the study participants, collected and coded data, and contributed to the write up. AP and GV analysed the data. JG and AP led the interpretation and writing with input from all authors. AP and JG had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Declaration of interests We declare no competing interests. Acknowledgments This study was sponsored by the University of Manchester. The study was principally funded by the Medical Research Council ( MR/K005863/1 ) with additional funding from the UK Department of Health and the National Institute for Health Research (NIHR) Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King's College London. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health. We gratefully thank all families who participated in the study and the referring professionals.