Comparing inpatient stroke rehabilitation care and outcomes for people with and without aphasia in Australia.

This is a retrospective audit, so it gives a snapshot of documented care and cannot prove that aphasia caused the differences in care or outcomes. Aphasia was recorded as yes or no from clinical notes, with no type or severity information, so the study cannot tell whether severe aphasia, mild aphasia or other factors explain the differences. The validity of the aphasia diagnosis was not checked, and documentation practices may differ for patients with and without aphasia. Patients with unknown aphasia status were excluded, and for care-process questions, missing or not documented information was counted as care not provided, which may understate care that happened but was not recorded. Although models adjusted for some patient factors and hospital differences, unmeasured confounders may have contributed, and the authors state the outcome findings should be interpreted cautiously. The data come from Australian rehabilitation hospitals in 2016, 2018 and 2020, so they may not apply to other countries, outpatient services or current practice. The audit did not include processes of care specific to aphasia, so it cannot describe whether communication supports were actually effective.

Authors reported institutional funding from the Stroke Foundation for some authors, and two authors were affiliated with the Stroke Foundation in leadership roles. One author received an Australian Government Research Training Program scholarship. Some authors received NHMRC Research Fellowship support, and one received a Future Leader Fellowship from the National Heart Foundation of Australia. One author reported educational grants from Boehringer Ingelheim and Amazon Web Services paid to the institution.

Do not read these differences as proof that aphasia caused poorer care or worse outcomes. This was an observational audit of records, aphasia type and severity were not recorded, and the authors say unmeasured factors and documentation differences may have contributed to the findings.