To compare 2 short-term, community caregiver training interventions for preschool-aged children with Autism Spectrum Disorder who had low resources. Low resource was defined by the US Department of Housing and Urban Development low-income index or 1 “indicator,” (eg, Medicaid eligibility). Child outcomes focused on joint engagement, joint attention, and play.
Participants included 112 families of a child who had Autism Spectrum Disorder who met criteria for being low-resourced and who were randomly assigned to 1 of 2 3-month interventions, group caregiver education or individualized caregiver-mediated intervention (CMM). Children were assessed for social communication skills pre- and post-treatment, and followed up at 3 months.
All children improved in joint engagement and initiating joint attention, with significantly greater improvement by the CMM group. Outcomes on play skills were mixed, with improvement of symbolic play for the CMM group and no change in functional play skills. Joint engagement maintained over time for the CMM group, and initiating joint attention maintained for both groups over time.
This study is among the first randomized trials comparing 2 active interventions with a large sample of low-resourced families. Results suggest improvements in core autism deficits of joint engagement, joint attention, and symbolic play with relatively brief, caregiver-mediated interventions, but additional support is necessary to maintain and generalize these gains over time.
We appreciate the opportunity to comment on our recent published study. It is important to better understand both the promise and the limitation of these data. First, our sample was a diverse sample, reflecting differences in service access and families across the United States. The locations of the 5 centers are provided in the authors' list and in each author's contribution to the manuscript. All 5 sites are teaching clinics, and autism centers, and this is important to consider. However, these centers differ in size and diversity, area of the country, and large versus small cities. Furthermore, we did not find any significant site differences in the treatment effect for both primary and secondary outcomes. Thus, these factors should lend greater credibility to the generalizability of our findings.
We agree that primary caregiver being unemployed alone does not necessarily reflect a low resourced household. Primary caregiver in this case referred to the financial provider, and we should have been clearer in this description. The time period in which the study was conducted was a time of high economic stress in the US, with high unemployment rates. Given the diversity in study sites, some families had more economic resources but had limited access to autism-specific early intervention services; others had limited economic resources in which to access services. Within our sample, 61% of the families are considered as low income based on the US Department of Housing and Urban Development and 55% of the families are receiving financial assistance from the government. Sixty mothers in our sample worked from home, and 37 of them were receiving government assistance, and 42 of them were considered low income. About 12% of our families were not considered low income, but were in the study for several reasons including an inability to obtain services, family illness, or some other disadvantage. Study personnel at each site screened families and determined their eligibility according to the general criteria of "low resourced". As mentioned in the paper, none of these criteria significantly affected the outcomes.
We are happy to respond to the several comments about analyses. Post -hoc power calculation or sample size calculation was not provided since we obtained significant findings in both primary and secondary outcomes. The estimated marginal means for both entry and exit for both primary and secondary outcomes were provided within the manuscript along with the clinical relevance of the improvements for the children in the CMM group (pages 76 and 77). The reporting of the classical estimate plus or minus 2 SE confidence intervals would be misleading here, since these are not the appropriate quantities to report based on a mixed effects model with multiple covariates where effects would need to be interpreted adjusted for other covariates in the model.
We agree that the cut-off values for Cohen's f effect size for small, medium and large effects are somewhat arbitrary, however the reported effect sizes themselves are not statistical quantities in that they are free of sample size. In addition, these effect sizes were reported along with estimated marginal means on the primary measures for both entry and exit within the manuscript. The estimated marginal means are in the units of interest. Moreover, the clinical interpretation of the improvements was provided within the manuscript. For our primary outcome, children in the CMM group improved from being jointly engaged for 226 seconds to 333 seconds (difference of 107 seconds or almost 2 minutes within a 10 minute session) and the increase in our secondary outcome, joint attention, for the CMM group more than doubled during the treatment period. As was also stated in our paper, we find these changes to be clinically meaningful.
Finally, even though the current study is not a cost-effectiveness study, we do agree that these considerations are useful in implementation of the methods. Therapist time and community travel was similar between our two interventions; however, the cost of group versus individual therapy depended on the number of children in the group, ranging from one to four children. At times, non-research families were offered participation as well in order to compose a group. Thus we would expect the individual therapy to cost 2 to 4 times as much as group therapy, a consideration in generalizing findings.
Connie Kasari, Wendy Shih, Damla Senturk
Conflict of Interest:
None declared
The RCT by Kasari et al examining care-giver interventions for low resourced pre-schoolers with autism was a very interesting study but we question the study design and clinical applicability of their findings.
The inclusion criteria for "low resourced" families seems too broad and may not capture the group intended. A mother with a high school diploma or lower or the primary caregiver being unemployed alone does not necessarily reflect a low resourced household. In Australia a large proportion of primary caregivers of toddlers are "stay at home" parents, with the other parent working. These households may be very well- resourced. Furthermore, the study recruited participants from 5 study centres across the United States. However, the locations of these centres and indeed the type of facilities were not disclosed in the paper. These details are of key importance in determining generalizability of study data to other populations.
In the methods, there are no sample size calculations nor an expected or clinically relevant difference between the two groups indicated. Thus the clinical relevance of the study is hard to interpret. Furthermore, the mean and SD are given but there are no confidence intervals stated. We also found the results lacking in detail and specificity. The use of Cohen's f in this article is less helpful than simply reporting the results in the units of interest (for example, difference in minutes between the 2 treatments). Specifically, the standard notion that 0.1, 0.25 and 0.4 denote small, medium and large effect sizes refers only to the strength of the statistical association - there is no consideration of clinical relevance. The latter is impossible to gauge from Cohen's f and so the clinical importance of the results are opaque.
Finally, it would be helpful to view the differential costs of the study interventions and outcomes. Although follow up at three months appeared to show positive correlation between carer-mediated interventions and joint engagement, the cost of the more intensive intervention needs to be weighed up in terms of any sustained change in behavior. This data would have been useful to understand to further interpret the feasibility of such intensive home-based therapy.
Conflict of Interest:
None declared