Dog-assisted interventions (DAIs) to improve health-related outcomes for people with mental health or neurodevelopmental conditions are becoming increasingly popular. However, DAIs are not based on robust scientific evidence.

To determine the effectiveness of DAIs for children and adults with mental health or neurodevelopmental conditions, assess how well randomised controlled trials (RCTs) are reported, and examine the use of terminology to classify DAIs.

A systematic search was conducted in Embase, PsycINFO, PubMed, CINAHL, Web of Science and the Cochrane Library. RCTs were grouped by commonly reported outcomes and described narratively with forest plots reporting standardised mean differences and 95% confidence intervals without a pooled estimate. The quality of reporting of RCTs and DAIs was evaluated by assessing adherence to CONSORT and the Template for Intervention Description and Replication (TIDieR) guidelines. Suitability of use of terminology was assessed by mapping terms to the intervention content described.

Thirty-three papers were included, reporting 29 RCTs (with five assessed as overall high quality); a positive impact of DAIs was found by 57% (8/14) for social skills and/or behaviour, 50% (5/10) for symptom frequency and/or severity, 43% (6/14) for depression and 33% (2/6) for agitation. The mean proportion of adherence to the CONSORT statement was 48.6%. The TIDieR checklist also indicated considerable variability in intervention reporting. Most DAIs were assessed as having clear alignment for terminology, but improvement in reporting information is still required.

DAIs may show promise for improving mental health and behavioural outcomes for those with mental health or neurodevelopmental conditions, particularly for conditions requiring social skill support. However, the quality of reporting requires improvement.