中文摘要：

目的探索犬伤的瞬时危险因素流行现状,针对主要的瞬时危险因素开展切实可行的干预措施,降低犬伤事件的发生。方法采用病例交叉设计方法,通过文献搜索以及小组访谈而制订问卷对犬伤病人进行面访。应用1∶1匹配的病例对照研究资料处理方法进行单因素分析,用条件logistic回归进行多因素分析。结果单因素及多因素分析的结果表明与狗玩耍（P_单=3.889,95%CI=2.317-6.528;P多=19.810,95%CI=7.528-52.132）、干扰（或阻止）狗吃东西（P_单=18.971,95%CI=2.543-141.510;P多=114.872,95%CI=11.528-1144.686）、由于瞬时性动作而惊吓到狗（P_单=24.333,95%CI=7.669-77.203;P多=100.062,95%CI=24.404-410.281）、你突然出现,狗看见你很兴奋（P_单=7.500,95%CI=2.642-21.289;P多=12.941,95%CI=6.622-148.936）等危险因素都具有统计学意义。结论犬伤发生与多项瞬时危险因素有关系。

英文摘要：

Objective To explore the transient risk factors related to dog injury and develop feasible intervention measures to reduce the occurrence of dog injury. Methods A case- crossover study was performed to determine the transient risk factors related to dog injury. The questionnaires on the subjects were conducted by trained investigators face to face and data were input after verified correctly. Results The findings from univariate analysis and multivariate analysis showed that there were significantly statistical associations between playing with dog（ P_univariate= 3. 889,95% CI = 2. 317- 6. 528; P_multivariate= 19. 810,95% CI = 7. 528- 52. 132）,disturbance of dog while eating（ Punivarite= 18. 971,95% CI = 2. 543- 141. 510; P_multivariate= 114. 872,95% CI = 11. 528-1144. 686）,surprising the dog（ P_univariate= 24. 333,95% CI = 7. 669- 77. 203; P_multivariate= 100. 062,95% CI =24. 404- 410. 281）,and exciting dog when familiar people appeared（ P_univariate= 7. 500,95% CI = 2. 642-21. 289; P_multivariate= 12. 941,95% CI = 6. 622- 148. 936）. Conclusion The multiple transient risk factors were associated with the dog injury.