When co-design works (sort of): the case of the Australian elder abuse screening instrument.

Applying co-design methodologies is increasingly recommended for engaging diverse end-users and bridging evidence-practice gaps. Yet, one of the ongoing challenges for research using co-design is the lack of evidence as to whether co-design leads to better outcomes than not using co-design. In this article, we outline how, despite adhering to a time and resource intensive co-design process with strong moral and ethical foundations, its implementation by end-users led to mixed outcomes around improved elder abuse screening. We discuss the implications of these ambiguous results, arguing that "noise" in our data might be inevitable due to the inherent sensitivities associated with elder abuse screening and offer a polemical recommendation about why the Australian Elder Abuse Screening Instrument (AuSI) should nevertheless be rolled out.