Effect of Feedback-based Interventions and Daily Practice Schedule on Sentence Production in Persons with Chronic Nonfluent Aphasia

Research Objectives: To examine the effects of feedback-based interventions and daily practice schedules on sentence production in persons with nonfluent aphasia (PWNA).

Design: The study used a crossover design, and it included pre-treatment, treatment and post-treatment phases. Previous findings examined pre- and post-treatments improvements in the outcome measures (Imaezue, 2024; Imaezue et al., in press). Here, we establish improvement in sentence production over time during feedback-based interventions and the effect of daily practice schedule (continuous vs. spaced) on treatment gains. Interventions included recursive self-feedback (RSF) and external feedback (EF) treatments. Treatment order was counterbalanced across participants. Automated speech transcription was completed using BatchAlign2 (Liu et al., 2023) followed by linguistic analysis using CLAN (MacWhinney, 2000) yielding 22,518 speech samples. We used linear mixed-effects model, via RStudio, to analyze the effects of treatments, session and daily practice schedule on sentence production.

Setting: This study was conducted within participants' homes.

Participants: We recruited four right-handed, English-speaking adults with mild-moderate chronic nonfluent aphasia without significant cognitive impairment. Participants included AE2 (6 years post-stroke), AE3 (12 years post-stroke), DP1 (7 years post-stroke) and DP5 (24 years post-stroke). We established aphasia severity and type through the Western Aphasia Battery-Revised (Kertesz, 2007) and cognitive status through the Cognitive Linguistic Quick Test plus (Helm-Estabrooks, 2017).

Interventions: We administered treatments via a mobile application. During RSF, participants read and listened to a sentence from a personalized script, produced the sentence, then revised their production, repeating this process eight times per sentence across twelve sentences. The same protocol was used in EF, however, the participants received corrective speech model of the target sentence after each production attempt. Participants practiced for two hours daily either continuously or spaced throughout the day, totaling ~14 sessions over 2-3 weeks.

Main Outcome Measures: Speaking rate (words per minute) was the primary outcome measure.

Results: Speaking rate improved significantly with sessions (Estimate = 2.67, SE = 0.12, t = 22.44, p = 0.023). Improvement was greater with spaced schedules compared to continuous ones (Estimate = -0.35, SE = 0.07, t = -5.37, p < 0.001). The interaction between schedule and treatment was significant (Estimate = -3.10, SE = 0.52, t = -6.00, p < 0.001), with smaller treatment differences under continuous schedules. RSF yielded a larger improvement in speaking rate compared to EF (Estimate = 0.24, SE = 0.07, t = 3.65, p < 0.001).

Conclusions: RSF and spaced schedules are more beneficial for improving speaking rate in PWNA than EF and continuous schedules.

Author(s) Disclosures: None