Sorting the "mixed bag" of semantic tasks in aphasia therapy: a scoping review

BackgroundSemantic therapy tasks are popular in aphasia rehabilitation, with many different treatments reported that claim to target semantic processing and to effect improvement. While most tasks involve some semantic processing, the term "semantic therapy task" generally refers to tasks with an explicit focus on meaning (e.g. semantic feature analysis, odd-one out), rather than tasks that only require implicit semantic processing (e.g. repetition in the presence of the picture). Despite the focus on meaning, the active ingredients underpinning explicit semantic tasks remain unclear.AimThis research aimed to examine the prevalence and change over time in the use of different explicit semantic therapy tasks, identifying the steps within each task and the variability of these tasks and steps across studies.Method and proceduresScoping review methodology was used to systematically search for studies published before November 3, 2022. Studies were included if they reported primary data that included participants with stroke aphasia and described semantic therapy tasks. We documented explicit semantic tasks and their component steps, and examined variability across semantic therapy studies.Outcomes and results8,571 records were screened, 631 full-text articles assessed for eligibility, and 244 articles were included. We identified 67 distinct semantic tasks. Over time, there was a significant increase in the studies using semantic feature analysis (i.e. semantic feature generation protocol +/- semantic feature review), and a corresponding decrease in the studies using other tasks.There was considerable variability in the implementation of semantic tasks, with different combinations of steps. We examined two tasks in detail (semantic feature generation, cueing hierarchy). Variability in the implementation of semantic feature generation included differences in the features used, whether or not feature choice was provided, whether feature repetition was required, and the inclusion of additional tasks (e.g. response elaboration training). Cueing hierarchies varied in the type, number, and order of cues provided.Outcomes and results8,571 records were screened, 631 full-text articles assessed for eligibility, and 244 articles were included. We identified 67 distinct semantic tasks. Over time, there was a significant increase in the studies using semantic feature analysis (i.e. semantic feature generation protocol +/- semantic feature review), and a corresponding decrease in the studies using other tasks.There was considerable variability in the implementation of semantic tasks, with different combinations of steps. We examined two tasks in detail (semantic feature generation, cueing hierarchy). Variability in the implementation of semantic feature generation included differences in the features used, whether or not feature choice was provided, whether feature repetition was required, and the inclusion of additional tasks (e.g. response elaboration training). Cueing hierarchies varied in the type, number, and order of cues provided.ConclusionThis scoping review demonstrates the continued and increasing use of explicit semantic tasks in aphasia rehabilitation. By detailing the steps used in explicit semantic tasks, we highlight variability in their implementation. Semantic tasks are often multicomponent, multimodal and/or combined with additional tasks, and the vast majority include a word-form component (e.g. picture naming, repetition in the presence of a picture). The active ingredients of semantic tasks are obscured by this variability. Future research is required to further our understanding of the active and essential ingredients of semantic therapy provided to people with aphasia.