ASSIST: Treatment for Childhood Apraxia of Speech (ASSIST)

Childhood apraxia of speech (CAS) is a pediatric motor speech disorder that impairs the planning of movements needed for intelligible speech. Children with CAS often show little or slow progress in standard speech therapy. This research is a Phase 1 study that tests initial efficacy and optimal parameters of a theoretically based integral stimulation treatment called ASSIST (Apraxia of Speech Systematic Integral Stimulation Treatment). In three small randomized group design studies, children (N=20 per study) receive 16 hours of individual ASSIST. The three studies systematically investigate treatment intensity (2 vs. 4 weeks) and two critical aspects of target selection: complexity (simple vs. complex target) and lexicality (words vs. nonwords). Each study also systematically examines the effect of treatment on functional outcome measures, including parent ratings of intelligibility and communicative participation, and objective intelligibility measures obtained from unfamiliar listeners.

This research involves a hypothesis-driven series of Phase 1 studies using comparative group designs to investigate a new integral stimulation treatment program for children with CAS, called ASSIST (Apraxia of Speech Systematic Integral Stimulation Treatment). The overall objective of this research is to test the initial efficacy of ASSIST and determine its optimal parameters. Specific Aim 1: To test the initial efficacy of intensive ASSIST for children with CAS, by examining short-term improvements on the primary outcome measure of speech accuracy. Forty children (combined across two studies) will be randomly assigned to a treatment or a control group (who will receive the treatment). The central hypothesis is that ASSIST enhances speech motor skill, and predicts greater improvements after 2 weeks (16 hours) for the treatment group than for the control group (who have not yet received treatment). Specific Aim 2: To determine the optimal parameters of ASSIST for children with CAS. In three studies, children (N=20 each) will be randomly assigned to condition, to address three critical treatment parameters: Aim 2a: COMPLEXITY: simple vs. complex targets (Study 1). Simple targets may enhance motivation through greater success during treatment, but more complex targets may facilitate greater transfer. Aim 2b: LEXICALITY: real word vs. nonword targets (Study 2). Real words may enhance motivation through functional relevance, but nonwords may facilitate transfer through a focus on underlying speech motor skills rather than existing lexical or speech motor representations. Aim 2c: INTENSITY: massed vs. distributed ASSIST (Study 3). Massed practice may enhance learning through neuroplasticity principles, but motor learning principles predict greater learning for distributed practice. Specific Aim 3: To determine changes on functional outcome measures following ASSIST treatment. In particular, for each of the three studies, the investigators will examine changes on the following outcome measures: Aim 3a: Parent-rated measures of intelligibility and communicative participation. Aim 3b: Objective measures of intelligibility based on transcriptions by unfamiliar listeners. This research includes three small-scale group treatment studies (Studies 1-3), conducted during summer to minimize confounds from concurrent treatment provided at school. In all three treatment studies, 20 children will be randomly assigned to conditions and will receive 16 hours of ASSIST in two-week blocks (weeks 2-3 and/or 5-6). Data will be collected in weeks 1, 4, and 7 (T1, T2, and T3, respectively). Assessment will take place 2-8 weeks prior to T1 to allow time to develop individualized sets of potential treatment targets for each child. For each treatment study, the investigators will conduct a corresponding experiment with unfamiliar listeners (Studies 4-6) to evaluate potential impact of ASSIST on intelligibility. Studies 2 and 3 will be informed by, but are not contingent on, preceding studies. For example, if Study 1 shows an advantage for simple targets, then Study 2 will compare simple words vs. simple nonwords. Thus, each study independently contributes to the overall objective of this proposal and addresses an independent aim. Children will attend a summer camp with 19 other children. Camp days (4 days per week during treatment block 1 [weeks 2-3] and treatment block 2 [weeks 5-6]) are 8:00 am to 3:00 pm and involve group activities and individual treatment sessions. All children participate in child-appropriate group activities in both treatment blocks (e.g., arts and crafts projects, games). Group activities are not treatment and have no focus on communication, but serve as the control condition. Treatment follows the pull-out model in which children attend individual treatment sessions in a separate room and then return to the group room. Treatment will be delivered by SLP graduate students under supervision of certified, licensed, and experienced SLPs. Treatment involves 32 individual sessions of 30-minutes (16 hours total) in which the child works with an SLP graduate student clinician to work on individualized speech goals. Children in intensive ASSIST conditions (all children in Studies 1 & 2, and Massed ASSIST groups in Study 3) will receive four 30-minute sessions per day (2 in the morning, 2 in the afternoon) separated by at least one hour to minimize fatigue, in either the first treatment block (camp weeks 2-3) or the second treatment block (camp weeks 5-6). Children in the Distributed ASSIST condition (Study 3) will receive two 30-minutes sessions per day (1 in the morning, 1 in the afternoon) separated by at least one hour to minimize fatigue, in both treatment blocks (camp weeks 2-3 and 5-6). Treatment words or phrases will be tailored to each child based on pre-treatment testing and personal relevance, as well as study aim. From each child's list of potential targets, two sets will be created and matched on relevant variables except for those under study. One set will be treated; the other set will serve to measure generalization (randomly assigned). Each target will be practiced in several frames (e.g., more _ , I want _ , _ is yummy) to bolster communicative relevance, enable variable practice, incorporate prosody, and engage language. Each 30-minute session targets several items (active set) from the treatment set developed for each child; new items are introduced as previous items are mastered and moved to maintenance practice. ASSIST uses step-up and step-down criteria for adaptive practice and pre-practice and practice components. Each session starts with pre-practice to ensure motivation, establish a reference of correctness and ability to identify a correct response, and establish ability to produce an accurate response. During pre-practice, the clinician elicits targets via integral stimulation on the first attempt, provides feedback and elicits additional attempts using cues and supports as needed. The clinician also solicits the child's self-evaluation (e.g., "Did that sound right?") and provides feedback on this self-evaluation. Criteria to begin practice are either ≥1 accurate response per target and correctly identifying ≥1 accurate response, or 15 minutes have elapsed (to ensure that sessions include at least 10 minutes of practice). During practice, the clinician aims to elicit as many attempts as possible. Treatment will follow a structured protocol with event sequences called teaching episodes. A teaching episode involves an initial target elicitation by the clinician, followed by feedback and/or a cue and another attempt by the child. Thus, each teaching episode involves at least two attempts by the child. Practice occurs in practice runs, which are bursts of practice separated by 1-minute breaks to minimize risk of fatigue. A practice run will continue for 4 minutes or until each target in the child's active set has cycled through 5 teaching episodes. The last practice run of a session includes all previously targeted items (active set + maintenance items). Conditions are systematically adapted to facilitate learning. Adult listeners will transcribe recordings of speech samples produced by children with CAS. Listeners will hear speech samples (randomized with respect to time of recording) and type what the child said. They will be able to control the volume to ensure a comfortable loudness level. They will type their responses to enhance data quality (legibility). These procedures will take up to 2 hours, which may be completed in a single 2-hour session or two 1-hour sessions.

Parallel groups with delayed treatment control group (Study 1 and 2). Parallel groups with distributed or massed intervention (Study 3)

Integral stimulation based treatment ("watch me, listen to me, say what I say") for CAS that is based on principles of motor learning and neuroplasticity and includes a systematic approach to adaptive practice. Treatment includes a focus on prosody, use of visual, tactile, auditory cues, slowing speech rate, and large amounts of practice.

Perceptual accuracy of treated items, judged by blinded listeners. Will be used to address Aim 1 (initial efficacy of ASSIST) and Aims 2 (optimal parameters of ASSIST).

Perceptual accuracy of untreated items, judged by blinded listeners. Will be used to address Aim 1 (initial efficacy of ASSIST) and Aims 2 (optimal parameters of ASSIST).

Parent rating of intelligibility in context using the Intelligibility in Context Scale (McLeod et al., 2012, 2015). This will be used to address Aim 3 (functional outcomes of ASSIST). The ICS has 7 items which are rated on a scale from 1 to 5. The Average Score is the mean of the ratings across the 7 items (possible score range = 1 to 5). Higher scores reflect better performance. The ICS captures the ease with which children are understood in their lives by different interaction partners (e.g., teachers, family members, strangers).

Parent rating of communicative participation using the Focus on Outcomes of Communication Under Six (Thomas-Stonell et al., 2015). This will be used to address Aim 3 (functional outcomes of ASSIST). The FOCUS-34 has 34 items which are rated on a scale from 1 to 7. The Total Score is the sum of the ratings across the 34 items (possible score range = 34 to 238). Higher scores reflect better performance. The FOCUS-34 captures the degree of a child's communicative participation (e.g., my child uses words to ask for things).

Percentage of words correctly understood from the audio recording by blinded unfamiliar listeners, based on the Test of Children's Speech (Hodge et al., 2009). This will be used to address Aim 3 (functional outcomes of ASSIST).

Inclusion Criteria: Age between 4;0 and 9;11 (years;months) at enrollment, based on parent report. From homes where the primary language spoken is English, based on parent report. Verbal output (50+ words) and communicative intent, as determined by the clinician and parent report. Speech sound disorder, as determined by a score <16th percentile on the Diagnostic Evaluation of Articulation and Phonology (DEAP) Articulation Assessment (Dodd et al., 2006) and/or the Goldman-Fristoe Test of Articulation (Goldman & Fristoe, 2015). CAS as a primary speech diagnosis, based on the following criteria: An average rating > 1 across three expert speech-language pathologists (SLPs), who will independently rate presence of CAS in children live or from video recordings of the assessment using a 3-point scale (0 = no CAS, 1 = possible CAS, 2 = CAS). These judgments will be based on perceptual speech features of CAS (inconsistent vowel and consonant errors, difficulties achieving and transitioning into articulatory configurations, abnormal prosody). An Apraxia Score of 1 or 2 on a Maximum Performance protocol in which they sustain vowels and fricatives as long as possible and repeat syllables as fast as possible (Rvachew et al., 2005; Thoonen et al., 1999). Normal hearing based on parent report or passing a standard pure-tone audiometry hearing screening at 500, 1000, 2000, and 4000 Hz (ASHA, 1997). Typical nonverbal cognition as determined by a T-score within 1.5 standard deviation (SD) of the mean on nonverbal subtests of the Reynolds Intellectual Assessment Scales (Reynolds & Kamphaus, 2003). Exclusion Criteria: Diagnosis of disorder that significantly affects communication and/or social interactions (e.g., autism), as per referral diagnosis. Uncorrected vision impairments that may interfere with ability to process visual cues used in treatment, as per parent report. Significant impairments of oral structure (e.g., cleft palate) as judged by the SLP based on an oral mechanism exam (Robbins & Klee, 1987). A primary diagnosis of dysarthria, as judged by the SLP. Unrelated health concerns that prevent children from participating, per parent report. Inability to meet toileting needs independently or separate from parent for a full day (as needed for camp), based on parent report.

Data spreadsheets containing finalized outcome measures for each child (e.g., speech accuracy data and intelligibility data) will be made available upon request to the PI to interested researchers, speech-language pathologists, and members of the general public. These data will be de-identified.

Data access will only be provided upon submitting a completed and signed data sharing agreement to the PI.