JCPSLP Vol 20 No 2 July 2018

is worthwhile noting that although all participants reported home practice, only 9 of the 17 participants returned their homework logs at some point throughout the study. The participants attended between 5 and 10 appointments of the 10 scheduled intervention appointments with attendance affected by familiar factors including illness, travel, and forgetfulness. Following intervention and post-assessment the participants were discharged from the project but were eligible to return to the referring speech pathologist for Mid-point review assessments occurred at the beginning of the sixth intervention appointment, with the exception of one participant who was unable to attend, and therefore the review occurred at session seven. Children were assessed only on the target sound word list relevant to their treatment target at mid-point review. Only accuracy of the target phonemes was considered when determining the percentage of target words correct. No formalised measures of inter- or intra-rater reliability measures were undertaken, although the authors did meet weekly for case discussion and watched a sample of recorded sessions with parent/carer consent and as approved by external ethics committee. The participants’ speech sound skills were reviewed again using the same measures as those used in the pre- assessment appointment (PCC attained on completion of the DEAP, percentage of target words produced correctly, perceptual measures) one week following the tenth scheduled intervention appointment. Analyses of the data recorded at pre-assessment, mid- assessment and post-assessment were performed by an independent statistician from the Hunter Medical using Stata v13.1software (StataCorp, 2013). Only children who completed the course of treatment were included in the analysis (n = 17). Dichotomous variables and categorical variables were compared using Fisher’s exact test, while continuous variables were compared using a Wilcoxon rank sum test. Multiple linear regression, with post-intervention outcomes as the dependent variables were used to control for differences in pre-intervention variables. Non parametric statistical methods were used throughout because of the small numbers of participants. Two sided p -values < 0.05 were considered statistically significant. Results Pre-assessment Nineteen of the 20 participants recruited met the language assessment criterion of attaining scores within age expected levels, as per CELF-4 normative data. Seventeen participants’ scores fell within average limits, while two participants’ scores were above age expected levels. One participant did not meet age expected levels and was excluded from the study as described above. That person returned to their local speech pathology clinic for management. Twelve of the 19 remaining participants were diagnosed as having a mild–moderate functional speech disorder, with percentage of consonants correct (PCC) ranging from 72.5 to 84.0, and seven were diagnosed as having a mild functional speech disorder, with PCC ranging from 87.0 to 98.0. The percentage of target words produced correctly ranged from 0% to 80% in the treatment groups, and 0% to 90% in the control groups. further speech sound intervention if required. Outcome measurement and analysis

treatment target, which was determined by performance on the DEAP, referral information, and the preference of the participant and their family. Only accuracy of the target phonemes was considered when determining the percentage of target words correct. Perceptual measures were also collected including clinician devised intelligibility scales at a connected speech level (using a Likert scale of 1 to 5) as completed by the speech-language pathologists and parents/carers, a rating of concern scale (using a Likert scale of 1 to 5) as completed by the parents/carers only, and satisfaction ratings where the participants and their parents/carers were asked how they feel about their/their child’s talking, using three faces (happy, neutral, sad) to represent this measure. The assessments were conducted by the two authors (the speech-language pathologists) who also conducted the intervention appointments and were therefore not blind to the planned intervention or the aims of this study. Intervention Intervention was delivered by the two authors (the speech- language pathologists), who delivered the treatment on alternate weeks. That is, participant 1 with their parent/ carer attended with speech-language pathologist A one week, and speech-language pathologist B the next. This meant that both the treatment group and the control group received intervention from both authors, controlling for this variable in interpreting group outcomes. Following pre-assessment, participants and parents/ carers attended the clinic weekly for a 30–40 minute clinic-based appointment. A maximum of 10 therapy appointments were offered. Attendance and participation in sessions was recorded in clinical documentation added to the participants’ medical records. Therapeutic service was provided as per the facility’s policies and procedures. Each initial therapy appointment included a perception component targeting auditory discrimination skills and a metalinguistic component where the place of movement was described and articulators drawn with the participant. A traditional articulation hierarchy approach was then used, starting with the sound in isolation, then syllable level, word level, phrases, sentences, story and conversational levels (Bowen, 2015). Principles of motor learning (Maas et al., 2008) were applied, with two phases implemented in each session; a pre-practice and a practice phase. Feedback known as “knowledge of performance” was provided by the speech-language pathologist to teach the participant how to make the sound in the pre-practice phase. The ultrasound was also used to provide visual feedback to the treatment group. In this phase 30 attempts of the target sound were produced. In the practice phase, feedback known as “knowledge of results” was provided by the speech-language pathologist and simply advised the participant as to whether the sound was correctly produced or not. Ultrasound was not used in this phase. The speech-language pathologist elicited 70 productions of the target sound/structure in the practice phase. The number of targets elicited in each phase of the intervention sessions was recorded on, now stored, data collection sheets and reported in clinical documentation added to the participants’ medical records to ensure adherence to the determined production ratio. Home practice was then discussed with the participants and the parents/carers and activity sheets provided. A homework log was provided to record home practice and it was asked that this be returned each session. The results of these homework logs were not analysed for the purpose of this study; however, it

78

JCPSLP Volume 20, Number 2 2018

Journal of Clinical Practice in Speech-Language Pathology