Inclusion of People With Aphasia in Stroke Trials: A Systematic Search and Review

Abstract

Background

Although people with aphasia (PwA) represent 30% of stroke survivors, they are frequently excluded from stroke research, or their inclusion is unclear. Such practice significantly limits the generalisability of stroke research, increases the need to duplicate research in aphasia-specific populations, and raises important ethical and human rights issues.

Objective

To detail the extent and nature of inclusion of PwA in contemporary stroke randomised controlled trials (RCTs).

Methods

We conducted a systematic search to identify completed stroke RCTs and RCT protocols published in 2019. Web of Science was searched using terms ‘stroke' and ‘randomised controlled trial'. These articles were reviewed by extracting rates of PwA inclusion/exclusion, whether ‘aphasia' or related terms were referred to in the article or supplementary files, eligibility criteria, consent procedures, adaptations made to support the inclusion of PwA, and attrition rates of PwA. Data were summarised, and descriptive statistics applied when appropriate.

Results

271 studies comprising 215 completed RCTs and 56 protocols were included. 36.2% of included studies referred to aphasia/dysphasia. Of completed RCTs, only 6.5% explicitly included PwA, 4.7% explicitly excluded PwA, and inclusion was unclear in the remaining 88.8%. Among RCT protocols, 28.6% of studies intended inclusion, 10.7% intended excluding PwA, and in 60.7%, inclusion was unclear. In 45.8% of included studies, sub-groups of PwA were excluded, either explicitly (i.e., particular types/severities of aphasia, e.g., global aphasia) or implicitly, by way of ambiguous eligibility criteria which could potentially relate to a sub-group of PwA. Little rationale for exclusion was provided. 71.2% of completed RCTs did not report any adaptations that could support the inclusion of PwA, and minimal information was provided about consent procedures. Where it could be determined, attrition of PwA averaged 10% (range 0-20%).

Conclusion

This paper details the extent of inclusion of PwA in stroke research and highlights opportunities for improvement.