Researchers from MGH’s Institute for Health Professions and Harvard’s Program in Speech and Hearing Bioscience and Technology published a study comparing behaviors and outcomes in persons with chronic aphasia who practiced lexical retrieval using tablet-based therapy materials. During practice, half the users were discouraged from using the app’s ‘cue button’ – which spoke the answer – and to focus instead on naming items through mindful, effortful engagement (‘trained group’); the others were not told to avoid using the ‘cue button’ (‘untrained group’). The goals were to study – during practice – subjects’ response latencies, cue uses, performance accuracy, and outcome changes.
Six subjects, pseudo-randomly assigned to the 2 groups, completed the study. Four were males, and 2 – females. Subjects’ mean age was 62.3 y, mean time post-onset – 11.2 y, and mean Aphasia Quotient (AQ) on the Western Aphasia Battery-Revised (WAB-R) was 69.0. Four participants were categorized as having anomic aphasia, one – Broca’s aphasia, and one – transcortical motor aphasia. Subjects were assessed for outcomes using the WAB-R, the Boston Naming Test (BNT), and the Cognitive-Linguistic Quick Test (CLQT). Subjects were trained on the iPad-based Constant Therapy application to complete tasks in the category matching, feature matching, rhyming, and syllable identification exercises. In the background, the app platform captures and reports – among other things – stimuli presented, response latency, and response accuracy. The intervention lasted 10 weeks with both in-clinic and at-home use, following an initial 2 hours or instructions and practice in operating the technology.
The data captured during use permit – for the two groups – calculation of behavioral means and standard errors of participants’ response latencies, proportions of cue use, trials per log-in, accuracies, and more. Group analyses of these data reveal the following: the trained group generally showed longer response latencies on tasks that the untrained group. The untrained group displayed a higher frequency of cue use that did the trained group. The trained group completed significantly more trials per log-in at home than did the untrained group. Overall performance accuracy tended to sag over time, the effects most noticeable in the trained group during in-clinic performance and the untrained group during home performance. In outcomes, 2 of the trained group and 1 of the untrained group improved significantly compared to benchmark on WAB-R’s AQ, while all 3 trained participants and 1 untrained subject showed such improvements on the BNT.
For further reading: J. Gallée, R. Pittmann, S. Pennington et al. (2020) The application of lexical retrieval training in tablet-based speech-language intervention. Frontiers of Neurology, 11:583246, 14 pp.