Blogging about Research: A Randomized Control Trial of DIR/Floortime in PK children with ASD

RCT of DIR in PK children with ASD

A pitot randomized controlled trial of DIR/Floortime parenting training intervention for pre-school children with autistic spectrum disorders

Citation:  Pajareya, K., & Nopmaneejumruslers, K. (2011). A pilot randomized controlled trial of DIR/Floortime™ parent training intervention for pre-school children with autistic spectrum disorders. Autism15 (5), 563-577. doi: DOI: 10.1177/1362361310386502

Purpose:  Purpose of this study was to compare DIR/Floortime parent training in a randomized control trial to determine effectiveness.  Success in this RCT would also validate results of Solomon’s pilot study from 2007 (see my review here).

Participants:  There were 32 participants in this study, however 1 subject did not complete the study due to refusal to use DIR/Floortime techniques.  Therefore data from 31 subjects were taken and analyzed.  Subjects were recruited via paper advertisement of DIR/Floortime model displayed at the National Institute for Child and Family Development, Mahidol University, Thailand.  Interested parents registered for more information.  Registrations were arranged in sequence and serially called into an office for screening and to confirm a diagnosis of ASD.  Children whose dx met the DSM-4 definitions of ASD  were admitted into this study and randomly placed in a control and experimental group.  All participants were between ages of 2-6 years of age.  Subjects with any other medical diagnosis, those who were geographically inaccessible for follow-up visits, or those who parents were not literate or had known to have chronic physical or psychiatric illness were excluded from this study.

Participants were divided equally into four strata (4 groups of 8 children):  mild ASD (24-47 mos), mild ASD (48-72 mos), severe ASD (24-47 mos) and severe ASD (48-72 mos), determined by CARS (Childhood Autism Rating Scale) scores (i.e. scores 30-40 points=mild, 41-60=severe).  Ten out of 16 children in control group, and 13 out of 16 children in intervention group were dx’d with ASD (the remaining were dx’d with PDD-NOS).  The difference between these ratios were statistically analyzed and no statistical difference between groups were identified.  Parents of intervention group tended to have lower educational levels as compared to parents in control group, however when analyzed this was not statistically significant either.

All subjects at the start of this study were also participating in their typical routine care ( in Thailand this includes early childhood and PK programs (from 10-15 hrs/wk) which implement ABA techniques (~ 3hrs/wk) as well as one-on-one ST, OT).  In order to control for the effects of additional services outside of DIR, the researchers allowed continuation of all current routine care.  By the end of this study 5 families in the control group had decreased the number of hours their child spend in PK programs or took their children out of their PK programs so they could focus on implementation of DIR at home.


Parent Training and Intervention:  Parents of control group were trained by a home consultant-HC (who had been trained in DIR and working as a HC in DIR for two years prior to this study), by participating in a one-day workshop on DIR, including a 3 hour DVD explaining the concept of DIR model, sensory processing of children with ASD, and the primary 6 FLDs (Functional Developmental Levels) of DIR model.  This was followed by a one-on-one visit where parents were trained for 1.5 hours on ways to identify and follow their child’s cues and implement techniques of DIR.  Goals were determined and parents were given several ideas of way to target these goals in semi-structured activities using DIR techniques (these activities were ID’d in a pocket handbook given to each parent for reference throughout the week).  Progress observed and goals updated upon each HC visit.

Goal of this study was to have parents provide at least 20 hours of DIR intervention at home. Parents were given log sheets to estimate the amount of time they spent providing the semi-structured DIR activities throughout each week.


FEAS:  The primary outcome measurement was based on the FEAS (Functional Emotional Assessment Scale) and was given at the beginning and end of this 3 month study.  A 15 minute videotape of parent playing with their child was taken an assessed using the FEAS pre- and post-study.

Intra-class correlation coefficient:  Was determined by comparing FEAS scores of the two raters using 20 random videos. Intra-class correlation coefficient was used to assess reliability between two raters and judged to be very high ( at 0.96).

CARS:  The CARS was used pre- and post-study to assess the severity of symptoms of ASD participants were exhibiting (rated on a scale from 15-60).

FEDQ:  The Functional Emotional Developmental Questionnaire was also used to provide a rating pre-and post-study via this parent questionnaire.  The FEDQ has to be translated into Thai for parents to answer then translated back to English for full assessment.  Validity of Thai version was determined by 3 other health professionals who had worked in the field for at least 3 years and one parent of ASD.  Intra-class coefficient (to determine validity) ranged from .75-1 showing strong validity.

Contamination:  Two families of control group were friends with families of experimental group and were inadvertently taught some DIR techniques that they also began using at home with their children.  Therefore, overall comparison between control and experimental groups may not be as accurate due to this contamination.

Results/Statistical Analysis:

FEAS:  Change in FEAS score for subjects in the control group from pre- to post-study scores was averaged at 1.9 as compared to the change in FEAS scores for subjects from the intervention group pre- and post test which was averaged at 7.0.  The differences between scores of both groups was in fact statistically significant.

CARS: The changes in the CARS scores demonstrated higher rate of decreased scores for intervention group (on average 2.9) as compared to control group (on average 0.8), suggesting by the end of this 3 month study the intervention group was exhibiting less symptoms of ASD overall as compared to the control group.

FEDQ:  The intervention group also demonstrated greater improvements on this parent questionnaire (average of 7.7) as compared to control group (on average 0.8)

Even when taking into account the one subject who dropped out of the study and giving that child scores of 0, analysis demonstrates statistical difference between scores on CARS and FEDQ when comparing intervention and control groups.

Intensity of Intervention:  When analyzing the number of hours of parent intervention provided, researchers determined that children whose parents provided more than 10 hours of DIR intervention per week made greater gains on FEAS as compared to children whose parents provided less than 10 hours of intervention a week.

Limitations of this study:

1.  Short period for study with out a follow-up assessment over time:  This study is the first RCT of it’s kind that I am aware of and the number of subjects (31) who completed the study is a fair number to compare the results.  However the short period of time with which this study was performed is promising but not a full view of the use of DIR parent intervention over time.  In addition, the lack of a follow-up assessment AFTER the end of this study (possibly 3 months later) would be very helpful to see if current gains remained or if more gains were made with parent intervention without home consultation.

2.  Contamination:  It does appear that contamination did not affect the statistical significance of this study, however having subjects who are not familiar with each other would of course ward against this in the future.

3.  Intensity level was not recorded well:  As in Solomon’s study, the parent logs were estimations of the time they spent with their child in DIR activities weekly.  More likely parents overestimated the amount of time they spent with their child and the intensity levels may not actually be what they were recorded to be.

4.  Cultural differences:  Researchers noted the HC spent much time explaining and teaching how to interact with their child with ASD rather than talk at them.  It may be typical in Thai culture to see children in this way, and improvements in skills may be due to the fact that this is the first experience these children had interacting socially with another adult.  Cultural differences should be taken into account when comparing results of this study to others.

5.  Lack of other assessments:  Further comparison of cognitive and communication improvements would be helpful in determining the implications of the use of DIR on other skills areas.

Thoughts and Implications:  It appears this 2011 study does validate the results of Solomon’s 2007 study that parent intervention via DIR does improve social skills in PK children with ASD.  As this is the first RCT of it’s kind, that I am a aware of, I look forward to seeing more research conducted in this area.  I would love to see a RCT with several different intervention groups and an appropriate follow-up over time to compare different types of interventions. With that said, this RCT did control for the additional PK intervention, 1:1 ST, OT, and ABA and still there were significant improvements on all measures of the intervention group as compared to the control group.