Dose-Controlled Script Training for Aphasia: A Quasi-Randomized Pilot Clinical Trial Examining Treatment Frequency

Purpose: The optimal schedule for delivering a given aphasia treatment is unknown. Competing theories suggest that both intensive (i.e., massed) and distributed (i.e., spaced) treatment schedules have benefits. Moreover, studies that have begun to investigate optimal delivery schedules have emphasized treatments for single-word retrieval rather than treatments that target connected language and have greater potential for generalization. The present study compared outcomes from three conditions of a script-training treatment, varying in session frequency. Method: Participants were 45 adults with chronic poststroke aphasia (3 months poststroke). All study activities took place at an urban freestanding rehabilitation center. This dose-controlled quasi-randomized pilot clinical trial involved six 1-hr sessions of computer-based script training in each of three conditions: standard (two sessions per week for 3 weeks; n = 15), intensive (three sessions per week for 2 weeks; n = 15), and attenuating (three sessions in the first week, two sessions in the second week, and one session in the third week; n = 15). Treatment was delivered via the AphasiaScripts software, which involved repeated practice of each sentence from a scripted conversation with a virtual clinician. Outcome measures included the proportion of correct to total possible words in the script (i.e., accuracy), duration of script production (i.e., rate), and the percentage of correct essential information units during a conversation with a novel communication partner (i.e., stimulus generalization). Outcomes were analyzed using Bayesian linear mixed-effects models for accuracy, speech rate, and generalization of essential information units. Results: The main effects of treatment, regardless of condition, were positive between baseline and posttreatment, as well as between baseline and 1, 2, and 4 weeks posttreatment (follow-up). There were no statistical differences in the main effects of accuracy or rate between conditions and no interaction effects. The main effects of generalization, regardless of condition, were not statistically reliable between baseline and posttreatment but were positive between baseline and follow-up 4 weeks after treatment. Conclusions: Results suggested that, given six 1-hr script-training treatment sessions, session frequency did not influence outcomes in terms of accuracy, rate, or generalization. Rather, all three frequency conditions produced statistically reliable positive treatment effects. Supplemental Material: https://doi.org/10.23641/asha.33023660