Reliability of perceptual measurement of Apraxia of Speech characteristics

Background: Diagnostic accuracy and reliability of acquired apraxia of speech (AOS) in the presence of co-occurring aphasia and/or dysarthria is crucial for appropriate treatment selection and clinical decision making. However, overlapping symptomology and lack of operationalization of AOS assessment methods have contributed to inadequate interrater reliability of perceptual measures differentially diagnostic of AOS. Purpose: This study investigated factors influencing the operationalization of AOS assessment methods, primarily interrater reliability of perceptual characteristics of differentially diagnostic (i.e., phonetic and prosodic errors) measures in order to inform assessment methods in AOS with concomitant aphasia. In addition, several other factors influencing the operationalization of AOS assessment methods were explored including: the utility of a pre-existing stimulus readily available in a standardized aphasia assessment (WAB-R), interrater reliability of non-discriminatory characteristics of AOS (i.e., auditory groping and false starts), the influence of alternating motion rates (AMRs) and sequential motion rates (SMRs) on a diagnosis of AOS, and the influence of the WAB-R subtests on error production by diagnostic group. Methods: Forty participants presenting with varying aphasia subtypes and severities and potential motor speech impairment were included. Speech production errors were analyzed by four raters using narrow transcription methods in response to the WAB-R spoken language subtest stimuli (Naming, Repetition, and Spontaneous Speech subtests) of the WAB-R. Interrater reliability of perceptual measurement of both differentially diagnostic and non-discriminatory features of AOS when using consistent stimuli (WAB-R), measures (Apraxia of Speech Rating Scale) and trained raters using narrow transcription methods were examined. In addition, percentage agreement of AOS diagnoses with and without the inclusion of AMRs/SMRs, as well as the influence of WAB-R subtest on error production across groups with AOS with concomitant aphasia and those with aphasia only were also examined. Results: Both differentially diagnostic as well as non-discriminatory speech characteristics were shown to demonstrate adequate interrater reliability across a variety of aphasia subtypes and severities of both AOS and aphasia. Adequate agreement between a diagnosis of AOS with and without the inclusion of AMRs/SMRs was reported as well as a lack of significant differences of phonetic and prosodic error production between subtests. Conclusion: The current work provides preliminary evidence of adequate interrater reliability of perceptual features of AOS using consistent stimuli (WAB-R), measures (Apraxia of Speech Rating Scale), and trained raters using narrow transcription. Findings from this work also support the inclusion of the AMRs/SMRs in AOS assessment and highlight the importance of their role when assessing individuals with borderline/mild motor speech impairments. These preliminary results support the consistency and operationalization of assessment methods through the investigation of reliability of perceptual measurements of differentially diagnostic characteristics of AOS in the presence of aphasia