Maximizing treatment time for language disorders

The purpose of this study was to figure out the amount of therapy required for children with language disorders to make significant gains to their language skills. 

Business-as-usual therapy was provided to 233 Kindergarten, 1st, and 2nd-grade students with language disorders, by 73 SLPs, over the course of a school year. The SLPs were not told exactly what to do in therapy, but were simply asked to treat and take data as usual, and additionally record three videos of the kids (fall, winter, spring) to submit to the researchers to code. The researchers then coded these videos for what was being worked on, each moment of therapy. Nine language codes were used: grammar, vocabulary, listening comprehension, communicative functions, discourse, narration, abstract language, metalinguistics, and literacy. It took the researchers hours to code just a single therapy session, examining what was happening moment-by-moment.
 
From the coded videos and data logs, they found that the average language therapy dose per session was about 12 minutes (so, 12 minutes spent explicitly targeting one of the above-listed components of language, versus something else; which corresponded to just under half a session). Average frequency (how many times per week the child is seen) was just over one session per week (range was once every two weeks to daily).
 
More advanced analyses allowed the researchers to determine the intensity of therapy best-suited to making progress, as measured by language tests compared from the beginning to the end of the school year. That found the greatest gains are achieved with either high frequency/low dose or low frequency/high dose (treating about once per week) therapy, but not high/high or low/low. So, why not high/high? The researchers hypothesize a  “… value of distributing learning opportunities over time to allow periods of consolidation and to increase engagement…”. So, when looking for time-efficient options, a “low frequency/high dose treatment regimen warrants serious consideration”.
 
Now—how confident can we be in this conclusion? Well, it’s important to recognize that this is correlational data. Thus, we don’t know that certain intensities are causing the change, or if some other factor(s) are at play. Also, keep in mind that when trying low-frequency therapy, analyses revealed that the dose must be high. So, presumably, the therapy and therapist had better be good to pull this off. And I think that may be a sticking point for therapists looking to apply this data—ensuring that the dose is high enough during the session. The results also showed that, “…a significant 28% of the variance in children’s CELF-4 scores was attributable to the SLP…”. So, which SLP was providing the therapy made a big difference as well—emphasizing the need to look further into this good therapy thing.
 
Justice, L.M., Logan, J., Jiang, H., Schmitt, M.B. (2016). Algorithm-Driven Dosage Decisions (AD3): Optimizing Treatment for Children With Language Impairment. American Journal of Speech-Language Pathology. Advance online publication: doi:10.1044/2016_AJSLP-15-0058.