Treatment Efficacy for Children with Childhood Apraxia of Speech

Abstract

This thesis addresses treatment outcomes for idiopathic childhood apraxia of speech (CAS), a significant and persistent disorder of speech motor planning and programming. There is a need for greater external evidence to guide clinical decision making. A systematic review of CAS treatment research published from 1970 to 2012 found there are no published randomised control trials (RCTs) for CAS. Three treatments showed strong preliminary evidence to warrant inclusion in future RCTs: Dynamic Temporal and Tactile Cueing, Rapid Syllable Transition Treatment (ReST) and Integrated Phonological Awareness intervention. A constraint on treatment outcomes is the lack of a validated, clinically replicable procedure for differentiating CAS from all other speech sound disorders (SSDs). A methodological protocol was developed for differential diagnosis of idiopathic CAS and a parallel group RCT. The diagnostic study examined 47 children referred with suspected CAS. CAS was determined from other SSDs using the current gold standard, expert judgment using the ASHA (2007) and the Shriberg, Potter and Strand (2009) feature lists. 100% diagnostic reliability was achieved. Assessment measures were analysed using discriminant function analysis and four measures in combination predicted the original diagnosis with 91% accuracy. Replication is required. The RCT compared the ReST and NDP3 treatments for 26 children aged 4-12 years in an intensive block (1 hour sessions, 4 days a week for 3 weeks) delivered by supervised speech pathology clinicians. The NDP3 showed significantly greater treatment gains within 1 week post treatment and the ReST treatment demonstrated greater maintenance and generalisation to pseudo-words. Overall ReST and NDP3 have strong evidence for treatment efficacy for CAS. The discussion argues that children with CAS can demonstrate strong treatment and generalisation effects when treatment considers best evidence, appropriate diagnosis and plans for generalisation.