Caring for those who care: a scoping review of existing welfare assessment tools for assistance dogs

Assistance dogs (ADs) perform trained tasks that improve the independence of people with disabilities, yet their own welfare has received comparatively little scientific attention. These dogs live and work across highly variable environments and face demands that companion dogs do not, making their welfare strongly context-dependent and potentially subject to change throughout different stages of training and working life. Reliable, context-sensitive assessment tools are therefore needed, but the extent to which such tools exist is unclear. Here we review how AD well-being has been measured in the sparse existing literature, with a view to developing a dedicated welfare assessment tool grounded in the Five Domains Model. Following the PRISMA extension for Scoping Reviews, we screened Scopus and Web of Science. We only considered peer-reviewed studies, published in English, that assessed welfare through directly observable animal-based or environment-based indicators, but excluded studies that only had physiological measures. Two authors independently screened records at title and full-text level, and data were charted narratively. Of 396 records identified, 249 were screened after removal of duplicates, and only eight met our inclusion criteria. Most were published recently (after 2021) and mainly concerned guide dogs. None addressed puppies before the start of training or retired dogs. Welfare was assessed in two ways, either through proxy-report instruments administered to beneficiaries, their families, or organizations, or through direct behavioral observation coded by researchers. The choice of method closely followed the dogs’ working stage, whether in training or actively working. Behavior was the most frequently used indicator, mainly in its stress-related forms, whereas behaviors expressing positive affective states were only rarely observed and analyzed. Only one study presented a tool developed specifically for operative use of AD welfare assessment. Consequently, there are still only a few directly applicable tools for assessing well-being, and the available literature is fragmented, whilst valuable practical knowledge appears to remain confined to individual organizations. The identified papers support a shift toward an animal-centered understanding of welfare and show how necessary it is to develop multidimensional, context-sensitive tools that span the whole working life of the dog and bridge scientific research and applied practice.