The influence of quantitative intervention dosage on oral language outcomes for children with developmental language disorder: a systematic review and narrative synthesis

Purpose The aim of this study was to examine the degree to which quantitative aspects of dosage (dose, dose frequency, and total intervention duration) have been examined in intervention studies for children with developmental language disorder (DLD). Additionally, to establish the optimal quantitative dosage characteristics for phonology, vocabulary, and morphosyntax outcomes.

Method This registered review (PROSPERO ID CRD42017076663) adhered to PRISMA guidelines. Search terms were included in seven electronic databases. We included peer-reviewed quasi-experimental, randomized controlled trial or cohort analytical studies, published in any language between January 2006 and May 2020. Included articles reported on participants with DLD (M = 3–18 years); oral language interventions with phonology, vocabulary, or morphosyntax outcomes; and experimental manipulation or statistical analysis of any quantitative aspect of dosage. Studies were appraised using the Cochrane risk-of-bias tool.

Results Two hundred forty-four articles reported on oral language interventions with children with DLD in the domains of interest; 13 focused on experimentally/statistically manipulating quantitative aspects of dosage. No article reported phonological outcomes, three reported vocabulary, and eight reported morphosyntax. Dose frequency was the most common characteristic manipulated.

Conclusions Research is in its infancy, and significant further research is required to inform speech-language pathologists in practice. Dosage characteristics are rarely adequately controlled for their individual effects to be identified. Findings to date suggest that there is a point in vocabulary and morphosyntax interventions after which there are diminishing returns from additional dosage. If dose is high (number of learning opportunities within a session), then the literature suggests that session frequency can be reduced. Frequent, short sessions (2/3 × per week, approximately 2 min) and less frequent, long sessions (1 × per week, approximately 20 min) have yielded the best outcomes when composite language measures have been used; however, replication and further research are required before clinicians can confidently integrate these findings into clinical practice.