Eleven studies (n=289) were included in the review: one parallel controlled study (n=27), seven within-subjects designs (n=105), one matched-controls design (n=14), one quasi-experimental study (n=79) and one survey design (n=64).
Validity was not formally assessed, although the authors highlighted several methodological limitations: use of non-blinded raters, rating of behaviours by facility staff, small sample sizes, limited duration of the intervention and limited follow-up, lack of matched control and baseline groups, and poor response rate.
Agitation and aggressive behaviours (six studies).
Five studies found a significant reduction in aggression, agitation or behavioural disturbances in patients exposed to visiting AAT or resident animals (no data reported for one study, p-values ranged from p<0.05 to p=0.001). In one study levels of agitation rose significantly after the intervention ended (p=0.000).
Social behaviour (four studies).
Three studies reported significantly greater frequency of social behaviours in the presence of a therapeutic animal (p=0.009, p<0.001, no data available for one study). The fourth study reported greater frequency of words, meaningful information units and verbal initiations, however, no data were available and it was not reported whether this increase was significant.
Nutrition (one study).
One quasi-experimental study found that the introduction of aquaria into the dining room of special care units was associated with significantly increased food intake (p<0.001) and resident weight (p<0.000) in comparison with units without aquaria.
Pet substitutes (two studies).
One study found significantly decreased physical (p=0.036) and overall agitation (p=0.046) with a plush cat but not with a robotic cat. Residents demonstrated significantly increased pleasure and interest with the robotic cat (p=0.007 and p=0.028, respectively). The other study reported higher levels of responsiveness to a toy dog compared with a robotic dog, although it is unclear whether these differences were significant and no data were provided.
Other outcomes showing significant changes following the introduction of AAT were: reduction in heart rate (one study p=0.021), ward noise (one study, p=0.001), caregiver burden (one study p=0.047) and anxiety (p=0.004). Use of medication and cognitive function did not significantly change. Three studies found that severity of dementia was not associated with responsiveness to AAT. One study of robotic and toy substitutes found decreasing engagement as severity of dementia increased.