Sequential meta-analysis to determine the sufficiency of cumulative knowledge: The case of early intensive behavioral intervention for children with autism spectrum disorders

Abstract

Meta-analysis has become a popular tool to statistically integrate results across studies in order to formulate more general conclusions on treatment effectiveness. Unfortunately, traditional meta-analytic applications fail to answer the question whether enough cumulative knowledge is available to draw convincing statistical conclusions. Leaving questions regarding the sufficiency of cumulative knowledge unaddressed may lead to inefficient use of limited resources or to the dissemination of spurious treatment benefit. Sequential meta-analysis or SMA provides a statistical framework to determine the sufficiency of cumulative knowledge in a meta-analysis, but is relatively unknown in mental health or disability fields. In this article, we introduce SMA and demonstrate its application by resynthesizing research findings on the effectiveness of early intensive behavioral intervention (EIBI) for children with autism reported in five published meta-analyses. The results illustrate the additional information that can be gained by including a sequential approach in research synthesis. SMA may serve as a valuable tool to systematically build and interpret a cumulative knowledge base on treatment effectiveness in the field of developmental disabilities.