JCPSLP Vol 21 No 2 2019 DIGITAL Edition

shown to be difficult to replicate in standard clinical settings (O’Brian et al., 2013; Rousseau, Packman, Onslow, Dredge, & Harrison, 2002; Van Eerdenbrugh, Packman, O’Brian, & Onslow, 2018). Speech-language pathologists have been found to alter the published treatment protocol to suit clients and the treatment setting, such as not adhering to the prescribed 45–60 minute treatment sessions and not observing parents demonstrate treatment procedures (O’Brian et al., 2013). Some of the reasons for these adaptations include the limitations imposed by service or organisational policies and restrictions, and funding or financial constraints (O’Brian et al., 2013; Donaghy & Smith, 2016; Rousseau et al., 2002). In addition to these practice context constraints, SLPs are commonly presented with the challenge of implementing stuttering treatment for clients also presenting with comorbidities. Stuttering can co-occur with communication disorders including (but not be limited to) those affecting articulation, language, and voice production. Stuttering can also co-occur with intellectual impairment, autism spectrum disorder, Down syndrome, attention deficit and auditory processing disorder, as well as literacy and learning disabilities (Arndt & Healey, 2001; Eggers & Van Eerdenbrugh, 2018; Harasym, & Langevin, 2012; Paul et al., 2005; Sisskin, 2006; Unicomb, Hewat, Spencer, & Harrison, 2017; Watanabe & Kenjo, 2015). Much of this evidence has come from surveys or file audits from SLPs working with children who stutter in the community because, as previously mentioned, these children may typically be excluded from efficacy research. This can mean that the clinical application of research findings from efficacy research can be problematic. One way to address this problem is through effectiveness research, which seeks to evaluate treatment in a “real world” clinical context (O’Brian et al., 2013). To date only O’Brian et al.’s 2013 investigation of the implementation and effectiveness of the Lidcombe Program in typical Australian community clinics has done this. Thirty-one Australian SLPs, working in private and public clinics, and their 57 paediatric clients participated. While the reported speech outcomes were largely comparable to published trials, particularly when SLPs were trained to deliver the program, several aspects of the program’s protocol appear difficult to replicate in typical clinical conditions. As previously mentioned, SLPs were shown to deliver shorter treatment sessions, omitted parent demonstrations of treatment and the reported frequency of clinic visits indicated clients were not attending weekly sessions (O’Brian et al., 2013). It was also noted that, as the SLPs volunteered for the study, they may not have been representative of typical clinicians and so further investigations of general community settings were needed. File audits were suggested as potentially being an appropriate methodology to investigate community caseloads. Research aims Given the need for more information about the nature of typical clinical caseloads, this research aims to: (a) establish key characteristics of clients seeking paediatric stuttering treatment; (b) compare these characteristics to the eligibility criteria used in published stuttering clinical trials; and (c) determine the service types and treatment that clients received in a typical clinical caseload. Method This file audit spanned March 2008 to September 2013. The study idea and design were conceived by the first

author following completion of employment as an SLP at the practice. Therefore, file entries were recorded as per standard clinical practice and not for the purpose of future data extraction or analysis. The (private) practice was located in a diverse socioeconomic inner north-west Melbourne suburb. Clients self-referred to the clinic or attended following advice from educators, maternal child health nurses, general practitioners or paediatricians. Active client records were stored securely in a filing cabinet on site files organised in alphabetical order. Archived files of clients who had been discharged from the service were securely stored in boxes organised by year of discharge and filed in alphabetic order. The first author was the sole SLP managing stuttering clients at this general speech pathology practice. The treating SLP had attended training workshops for both the Lidcombe Program and school-aged stuttering treatment prior to working with this caseload. The SLP used this education to inform clinical decision-making. The selection criteria for this file audit were: (a) reason for attendance was listed on the standardised client history form as being stuttering (either as the primary reason or as one of multiple areas of concern) and (b) client age at assessment was younger than 12 years 11 months. Ethics approval was provided by the La Trobe University ethics committee (reference number S17-189). Vasser and Holzmann’s (2013) file review methodology was employed to ensure appropriate rigor in the data extraction process. This methodology draws on the authors’ personal experiences and a review of file audit methodologies utilised in the field of health care. This included developing specific research aims, trialling the research aims on a sample of files, considering variables in the data extraction, using standard extraction methods, maintaining explicit criteria and addressing confidentiality and ethical considerations. A structured methodology was implemented as poor design and analysis are routinely reported as a barrier to the success or impact file audits can have in clinical research (Johnston, Crombie, Davies, Alder & Millard, 2000). After the research aims were established, the data extraction methodology was developed (see Appendix 1). This took the form of an online survey that was initially trialled on a sample of 10% of files to ensure utility. No questions in this survey were altered following this initial trial. Due to confidentiality concerns, it was agreed that only the first author as the treating SLP would extract data from the files. Data was extracted and entered independently for each file on two separate occasions (4 weeks apart) to ensure accuracy. The contingency for any anomalies was for the first author to check the data for a third time; however, this did not need to occur. Clinical trials selected for the participant eligibility criteria comparison were drawn from the list of clinical trials provided in Appendix B of the Speech Pathology Australia Clinical Guideline for Stuttering Management (SPA, 2017). Studies were selected if they were Australian, published between the years of 2000 and 2016, and investigated a treatment also offered at the participating clinic (as listed in table 2). RESTART-DCM was excluded from this study, despite it being listed in the clinical guidelines, as training is currently unavailable in Australia and it was not offered at the participating clinic. Treatment outcomes from the file audit are briefly captured in the results; however, stuttering measures are not reported as it is beyond the scope of this study, and file audit data does not satisfy the Australian

Rachael Unicomb (top) and Bernadette O’Connor

109

JCPSLP Volume 21, Number 2 2019

www.speechpathologyaustralia.org.au