About the project

Canine aggression is often considered the most important behavioural problem in dogs because of its frequency and the consequences for people.

Dog bites to humans are a worldwide problem. Many non-fatal and fatal dog attacks may be preventable. However, in order to provide suitable and effective preventive measures these must be based on solid scientific evidence.

This study aimed to review all available, published and unpublished, literature relating to aggressive dog-human interactions. We utilised the systematic review approach to rigorously and systematically identify, select and critically appraise all relevant studies.

The specific aims of this project were to:

  • Conduct a systematic review to identify relevant literature relating to risk factors for aggressive dog-human interaction.
  • Further investigate risk factors for human-directed dog aggression using meta-analysis.
Thereby enabling us to:

  • Identify knowledge gaps and highlight areas where evidence is weak or contradictory, thus indicating areas where further research is needed.
  • Identify areas where there is strong evidence for risk factors upon which preventive measures can be developed.

This work will form the basis of a comprehensive report which will outline the robust evidence currently available concerning risk factors for human-directed dog aggression.

Systematic review

The systematic review is widely used as a tool in human medical research, and increasingly within veterinary medicine.

Although in some respects similar to the more common literature review, the systematic review process attempts to bring the same level of rigour to reviewing the research evidence as should be used in producing that research.  The design of a systematic review focuses on objectivity and the minimisation of bias. 

The systematic review process:


Meta analysis

Meta-analysis is the quantitative pooling of data extracted from studies identified during a review.  Many published studies may fail to identify the effect of potentially important risk factors because of small sample sizes, and consequently insufficient power to detect an effect.

Meta-analysis can combine multiple studies into the equivalent of a single larger study. However, this is only possible where studies test the same hypothesis via comparable study design and methodology.