Commentary: Glass half full or half empty? Testing social communication interventions for young children with autism – reflections on Landa, Holman, O’Neill, and Stuart (2011)

Impairments in social communication are the corediagnostic features of autism spectrum disorders(ASD). In the past two decades ndings from impor-tantclinical research studies(manypublishedin thisjournal) have translated into improvements inunderstanding and practice, for example leading to areductionintheageatwhichautismiscommonlyrstrecognised and diagnosed in many communities.One area of research that has not, perhaps sur-prisingly, progressed as much as would have beenexpected is that of psychological interventions. How-ever, it was not the case that intervention was notdiscussed by practitioners and researchers. To cut along story short, there has been an ongoing (and notalways civilised or fair-minded) debate over severaldecades about the effectiveness or otherwise of oneparticular approach - ‘applied behavioural analysis’(ABA), sometimes now called ‘early intensivebehavioural intervention’ (EIBI). There is not space toreview this sorry saga but I do think that the greenshoots of a rapprochement might be visible. What isclear is that both sides were debating a handful ofsmall, mostly poor-quality studies and almost norandomised controlled trials (RCTs) (Ospina et al.,2008). This contrasts with many other areas of childpsychiatryandpsychologywherethetrialliteratureofpsychological interventions is considerably moresecure. These are not mere esoteric considerationssince parents ask clinicians to be referred to treat-ments that ‘work’ and service commissioners andinsurance companies ask the same question too.One paper (Landa, Holman, O’Neill, & Stuart,2011) in the current issue of the Journal is thereforenotable in that it forms part of a ‘new wave’ ofintervention studies that are characterised both by afocus on specically enhancing social communica-tion outcomes and by using an RCT design. Landaand colleagues compared two kindergarten pro-grammes for 29-month-olds with an ASD. The pro-grammes differed only in that one focused on‘interpersonal synchrony’ (IS) more than the other(N = 50, two groups). IS includes a range of socialcommunication activities and constructs, includingjoint attention, imitation, turn-taking, non-verbalsocial communicative exchanges, affect sharing andengagement. The programme was delivered bytrained kindergarten staff and delivered for2.5 hours per day, 4 days per week for 6 months(overall around 200 hours of intervention). Parentsalso attended education classes focusing on thesame strategies implemented in the kindergarten(38 hours) and a monthly home coaching visit.Landa et al. found that the IS group differed from thenon-IS group on one variable only: ‘socially engagedimitation’, dened as the proportion of imitationspaired with eye contact to the examiner across aseries of modelled actions (which more than doubledfrom 17% to 42%). The groups did not differ inamount of initiated joint attention or shared positiveaffect when interacting with an examiner; nor didtheir scores on a standardised language measureimprove. This introduces the rst note of cautionafter my congratulations to the authors for havingrun and reported a ne study. What are the limits tothe effects we might expect for time-limited inter-ventions aiming to improve social communicativeoutcomes for young children with autism?In another recent trial Kasari and colleagues con-ductedashort-termkindergartentherapist-deliveredintervention.Threegroupsofchildren(N = 58intotal;mean age 42 months) were randomised to daily 30-minute sessions for 5 to 6 weeks where activitiesfocused on promoting either joint attention or sym-bolic play skills (and a control group). After 6 weeksthere were improvements in both the interventiongroups in aspects of child joint attention and play ininteraction with experimenters and with their moth-ers (Kasari, Freeman, & Paparella, 2006). One yearlater, both intervention groups had signicantlyhigher scores on structural language measures thanthe controls (Kasari, Paparella, Freeman, & Jahromi,2008).Both the Landa and the Kasari studies focused onpromoting early social communication skills butused a wide range of techniques in the delivery of thetherapy (‘from ABA to milieu teaching’ in Kasari et al.and ‘a continuum of adult-imposed structure fromdiscrete trail teaching … to pivotal response training… to routines-based interactions’ in Landa et al.)