Purpose: Early language and communication interventions for children with language impairments have been shown to be effective in assessments administered immediately after treatment. The purpose of the current systematic review and meta-analysis was to assess the overall durability of those effects over time and whether durability was related to outcome type, etiology of child language impairments, implementer of intervention, magnitude of posttest effects, time between intervention and follow-up, and study risk of bias.
Method: We conducted a systematic search of online databases and reference lists to identify experimental and quasi-experimental group design studies. All studies tested the effects of early communication interventions at least 3 months post-intervention. Participants were children 0-5 years old with language impairments. Coders identified study features and rated methodological quality indicators for all studies. Effect sizes at long-term timepoints and associations with potential moderators were estimated using multilevel meta-analysis with robust variance estimation.
Results: Twenty studies with 129 long-term outcome effect sizes met inclusion criteria. Studies included children with developmental language disorders or language impairment associated with autism. The overall average effect size was small and significant (g = .22, p = .002). Effect size estimates were larger for prelinguistic outcomes (g = .36, p < .001) than for linguistic outcomes (g = .14, p = .101). Significant factors were the posttest effect sizes, the risk of bias for randomized trials, and etiology of language impairment for linguistic outcomes. Time post-intervention did not significantly predict long-term effect sizes.
Conclusions: Outcomes of early language and communication interventions appear to persist for at least several months post-intervention. More research is needed with collection and evaluation of long-term outcomes, a focus on measurement, and consistency of primary study reporting.
Supplemental material: https://doi.org/10.23641/asha.23589648.