NARNIA: a new twist to an old tale. A pilot RCT to evaluate a multilevel approach to improving discourse in aphasia

Background: Developing effective interventions for people with aphasia, which both ameliorate impaired language and directly impact on real-life communication, is a key focus of aphasia research. While single-word and sentence-level models of language processing have informed effective interventions, there is limited evidence of intervention approaches that extend the principles from these models to discourse. Aims: This pilot randomised controlled trial (RCT) aimed to compare a novel multilevel intervention, a Novel Approach to Real-life communication: Narrative Intervention in Aphasia (NARNIA), with usual care (UC), with a view to assess the feasibility and acceptability of the approach and estimate effect sizes to design and power a definitive trial. Methods & Procedures: In a prospective, single-blind, RCT, 14 people with mild to moderate aphasia, between 2 and 165 months following stroke, were randomised to the two interventions. Both therapy conditions were delivered four times weekly over 5 weeks. The NARNIA intervention specifically combined word retrieval, sentence production, and discourse macrostructure across a range of everyday discourse (ED) genres. UC comprised any speech-language therapy routinely used in clinical practice, individually tailored to meet the assessed needs of the participant. The Curtin University Discourse Protocol (CUDP) was used to measure verb access, sentence production, and discourse structure. Outcomes & Results: Eight participants were assigned to the NARNIA intervention and six participants to UC. No significant differences were present in baseline prognostic factors between the two groups prior to intervention. Following intervention, orientation aspects of macrostructure were significantly greater in the NARNIA group on ED measures than in the UC group, with no other between-group differences found. Significant changes were, however, seen within groups across ED genres. The NARNIA group made significant gains across all language levels, while the UC made isolated gains in sentence production. Few changes were seen in narrative discourse for either group. While single-word processing was not significantly different between groups, significant within-group differences were seen. While both groups significantly improved in retrieving nouns in isolation, only the NARNIA group made significant improvement in verb processing. No change was seen in constrained sentence production for either group. Conclusions: These findings are highly promising in demonstrating the use of macrostructure to scaffold production of words and sentences and improve discourse organisation. The significantly greater within-group gains for the NARNIA participants will provide a platform to power a larger trial to evaluate the effectiveness of this integrated multilevel intervention for aphasia