Many meta-analyses and systematic reviews have tried to assess the efficacy of animal-assisted interventions (AAIs), reaching inconsistent conclusions. The present work posits a critical exploration of the current literature, using some recent meta-analyses to exemplify the presence of unattended threats. The present comment illustrates that the field (1) comprehends inconsistencies regarding the terms and definitions of AAIs; (2) pays more attention to the characteristics of the animals than to the action mechanisms of AAIs; (3) does not provide a clear connection between anthrozoology (how humans and non-human animals interact in communities), benefits of the human-animal interaction (HAI), and the design of AAIs; and (4) implicitly reinforces these phenomena through research designs. Thus, some conclusions extracted from these meta-analyses need further discussion. Increasing the internal validity of AAIs in empirical studies is an urgent task, which can be addressed by (1) developing a better understanding of how anthrozoology, the HAI, and AAIs relate to each other; (2) highlighting the mechanisms that explain the results in an empirical and specific way; and (3) changing the design of interventions, adopting a component-centered approach, and focusing on the incremental efficacy and efficiency of AAI programs.
Keywords: animal-assisted interventions; animal-assisted therapy; human–animal interaction; internal validity.