Impairments or delayed development in skills concerning social interaction and communication are at the core of autism spectrum disorders (ASD) . The attempts to help children with ASD develop meaningful language and social communication skills cover a wide range of different approaches. Yet, for most of the various intervention methods available and reported within ASD that are designed to improve communication and social interaction, only insufficient evidence of effectiveness exists . Some favourable results can be found for approaches such as early behavioural interventions [3, 4] and augmentative communication . However, if methodologically rigorous standards are applied, statistically significant improvements in communication and speech could hitherto only be ascertained in a parent-mediated communication-focused intervention , parent-mediated behavioural interventions , and music therapy [2, 7]. Similarly, in a review of "novel and emerging treatments" for ASD  including several nutritional supplements, diets, medications, and nonbiological treatments, it was found that the only treatment options that reached the highest ranking in an evidence-based grading system were melatonin, acetylcholinesterase inhibitors, naltrexone, and music therapy. Considering that pharmacological treatments typically target symptoms such as hyperactivity, agitation, or sleep disorders rather than core symptoms of ASD, and may have adverse effects [2, 8], music therapy can be viewed as a promising, but not yet sufficiently evidenced treatment for improving social interaction and communication skills within ASD. Due to various methodological quality limitations of previous studies , further high quality randomised controlled trials (RCTs) on common interventions for ASD have been found to be urgently needed.
Music therapy has a long tradition within ASD [9, 10], and there are many clinical reports, case studies, and single group studies (e.g. [11–13]; for an overview, see ) suggesting that music therapy may enhance skills of social communication such as initiating and responding to communicative acts. In recent years, increased efforts have been made to conduct more rigorous studies in this area. A Cochrane review combining the findings of three small controlled studies of music therapy in children with ASD  concluded that this type of intervention may have positive effects on the communicative skills of children with ASD, but also noted limited applicability of the studies' results to clinical practice due to very short duration of treatment conditions and low flexibility in music therapy techniques applied.
Following this review, some RCTs were conducted that strived for improved clinical relevance by applying treatment durations of several months as well as improvisational, flexible, child-centred methods of music therapy provided by trained therapists [15–17].
Improvisational music therapy for children with ASD may generally be described as a child-centred approach making use of the potential for social engagement and expression of emotions occurring through improvisational music making. Instead of practising targeted skills in an abstract manner, improvisational music therapy has been noted for its potential to provide a meaningful framework that, similar to early mother-infant interaction, encompasses relevant features of social communication such as being embedded in a shared history of interaction, having a common focus of attention, turn-taking, and musical and emotional attunement [15–18]. In the first of these newer RCTs on child-centred music therapy methods , it was suggested that improvisational music therapy may facilitate skills fundamental to social interaction in children with autism and proves to be effective in improving lower levels of initiating joint attention and responding to joint attention bids. Despite this trial's significant results, some methodological constraints such as its small sample size (N = 10) and large number of outcome measures limited the generalisability of its findings. Recent RCTs with slightly larger sample sizes of N = 23 (unpublished report, Thompson, McFerran, and Wigram, 2011) and N = 24 , respectively, also investigated effects of improvisational, child-centred music therapy approaches on social communication skills of young children with ASD, but were still seriously limited in sample size and test power. A large pragmatic RCT is needed to decide if improvisational music therapy improves core symptoms of ASD in a generalised setting.
The objectives of this study are as follows:
To determine whether music therapy is superior to standard care in improving social communicative skills in children with ASD as assessed by independent clinicians at the end of the treatment period.
To determine whether music therapy is superior to standard care in improving social responsiveness in children with ASD as assessed by parents/guardians at the end of the treatment period.
To determine whether the response varies with variation of treatment intensity.
To determine how the development of social communicative skills proceeds until follow-up twelve months after the start of treatment.
It is predicted that children's social communicative ability will increase over time, that social communicative skills may be better in music therapy conditions than in the standard care condition, and that more frequent music therapy may intensify the improvement in the skills assessed. Assessing participants' social communicative skills seven months after ending of treatment (12 months after randomisation) will yield important information on whether any effects in the skills investigated will be sustained.