Tidier descriptions of speech and language therapy interventions for people with aphasia; consensus from the release collaboration

Background: Speech and language therapy (SLT) interventions for people with aphasia are complex – for example, interventions vary by delivery model (face-to-face, tele-rehabilitation), dynamic (group, 1-to-1) and provider. Therapists tailor the functional relevance and intervention difficulty to the individual’s needs. Therapy regimes are planned at a specific intensity (hours per week), frequency (number of weekly sessions), duration (time from start to end of therapy intervention) and dose (total number of therapy hours). Detailed and transparent description of interventions for people with aphasia facilitates replication in clinic, between-study comparisons and data-syntheses. Incomplete intervention reporting and inconsisten-cies in the use of terminology have been observed (RELEASE: REhabilitation and recovery of peopLE with Aphasia after StrokE Collaborators, 2015-2018; Brady, Kelly, Godwin, Enderby, & Campbell, 2016; Pierce, O’Halloran, Togher, & Rose, in press). Even when similar terms are used there may be little agreement on their use (Pierce et al., in press). Our RELEASE Collaboration includes 72 multidisciplinary, multilingual aphasia researchers from 28 countries. In preparation for planned meta-analyses (HS&DR 14/04/22) we sought to extract and synthesise information on SLT interventions for aphasia. Description of interventions in research reports have benefited from the Template for Intervention Description and Replication (TIDieR; (Hoffmann et al., 2014). The TIDieR checklist supports transparent reporting, data extraction and synthesis in aphasia research. It has facilitated an exploration of the contribution specific parameters (for example intensity) may make to the effectiveness of (or tolerance to) an intervention (Brady et al., 2016). However, while the variables described above are readily summarised, other aspects of therapeutic interventions such as the theoretical approach, the materials used and the procedures employed (the “Why” and “What” within TIDieR) are more challenging to summarise in manner supporting data synthesis and meta-analyses. The World Health Organisation’s (“International Classification of Health Interventions (ICHI),”) also seeks a framework which supports the synthesis and statistical analysis of healthcare interventions based on (a) the treatment target, (b) the intended action to the target and (c) the processes and methods required to carry out the action (ICHI, 2018). However a framework which complements these initiatives and which supports greater consistency in the description of SLT interventions for aphasia is required. Aims: We sought to develop international consensus on a framework to support the description of SLT interventions for people with aphasia. Methods and procedures: Two researchers independently extracted information about the SLT interventions in our RELEASE database (Hoffmann et al., 2014). Information on therapy approaches, materials and procedures were extracted, where possible, as direct quotes from published reports. Using the narrative descriptions, similar approaches were grouped and assigned to one or more category labels by an experienced speech and language therapist. These preliminary groupings were shared with RELEASE Collaborators for review. Each reviewed interventions included within up to four category labels and responses via email were requested. There followed an opportunity for group discussion on the proposed categorisation via videoconference. Outcomes and results: Therapy interventions were categorised based on three per-spectives (a) the role of the intervention within the study design e.g., usual care as a comparison control (b) the intervention target e.g., rehabilitation of spoken language production and (c) the theoretical approach e.g., semantic therapy. We identified 15 SLT approaches. Categories were not mutually exclusive; rather they represented different ways of categorising a complex intervention. Inadequate reporting of therapy interven-tions, procedures and materials hampered some classifications. The Collaboration facilitated knowledge sharing relating to emerging treatment category definitions such as multimodal treatment (Pierce et al., in press) which reflected interventions aimed to utilise a range of learning mechanisms and neural networks to facilitate language recovery. Conclusions: Our collaboration agreed on a framework which supports transparent description, data synthesis and meta-analyses of SLT interventions for people with aphasia after stroke.