Abstract
BACKGROUND: Dog bite studies are typically based on hospital records and may be biased towards bites requiring significant medical treatment. This study investigated true dog bite prevalence and incidence at a community-level and victim-related risk factors, in order to inform policy and prevention. METHODS: A cross-sectional study of a community of 1280 households in Cheshire, UK, surveyed 694 respondents in 385 households. Data included dog ownership and bite history, demographics, health and personality (Ten Item Personality Inventory (TIPI) brief measure). Multivariable logistic regression modelled risk factors for having ever been bitten by a dog, accounting for clustering of individuals within households. RESULTS: A quarter of participants (24.78%, 95% CI 21.72 to 28.13) reported having ever been bitten by a dog during their lifetime, with only a third of bites described requiring further medical treatment and 0.6% hospital admission. Incidence of dog bites was 18.7 (11.0-31.8) per 1000 population per year. Males were 1.81 times more likely to have been bitten in their lifetime than females (95% CI 1.20 to 2.72, P=0.005). Current owners of multiple dogs were 3.3 times more likely (95% CI 1.13 to 9.69, P=0.03) to report having been bitten than people not currently owning a dog. Regarding all bites described, most commonly people were bitten by a dog that they had never met before the incident (54.7%). Individuals scoring higher in emotional stability had a lower risk of having ever been bitten (OR=0.77 for 1 point change in scale between 1 and 7, 95% CI 0.66 to 0.9, P=0.001). CONCLUSION: This study suggests that the real burden of dog bites is considerably larger than those estimated from hospital records. Further, many bites do not require medical treatment and hospital-based bite data are not representative of bites within the wider population. Victim personality requires further investigation and potential consideration in the design of bite prevention schemes.