Discourse recovery after severe traumatic brain injury: exploring the first year.

The control group was cross-sectional with a single data point, so practice effects could not be fully separated from recovery. The control participants were from the United States while the TBI participants were from Australia, so small language or cultural differences may have influenced scores. Participants were not assessed before 3 months post-injury, so recovery in the earliest period is unknown. The final cohort was 57 participants, and only 10 were recruited at 6 months, which limits subgroup certainty. The study used a single narrative task and a single discourse genre, so it may not capture all communication demands. Therapy data captured amount of speech pathology therapy across time points but not the nature or quality of therapy. Not every participant was assessed at every time point, so prevalence percentages are based on different participant numbers. The reliability analysis showed acceptable agreement, but the authors note results may differ with untrained users or other settings.

The authors reported no conflicts of interest. The study was part of a larger project supported by NHMRC #632681, and the authors said they alone were responsible for content and writing.

Do not conclude that more speech pathology therapy caused the observed recovery. The study measured only the amount of therapy, not its nature or quality, and adding therapy amount did not significantly alter the discourse recovery model.