Claim state and evidence level are separate. Software can exist without evidence of a human outcome, and a study can exist without proving this exact product or setup.
StateWhat kind of statement is this?
SetupWhere and for whom was it examined?
ResultWhat was observed or implemented?
LimitsWhat remains uncertain, missing, or corrected?
Across trials in young adults, anxiety and physiological stress show small-to-moderate effects while cognitive performance is the most variable outcome measured.
Pooled randomised trials report less anxiety and pain during dental procedures. The reviewers grade that finding low certainty and report a high risk of bias.