Objective To provide evidence for timely diagnosis and treatment of Kawasaki disease through assessing the relationship between the duration of fever and the incidence of coronary artery lesion in patients with Kawasaki disease. Methods To retrospectively analyze the clinical information of 352 inpatients with Kawasaki disease (including typical Kawasaki disease, incomplete Kawasaki disease, and non-responsive to intravenous immunoglobulin treatment Kawasaki disease ) from January 1997 to December 2007. The relationship between the duration of fever and the incidence of coronary artery lesion was presented by a linear trend plot, using Cochran-Armitage trend test. A value of P lt; 0.05 was considered statistically significant. Results Among 352 patients with Kawasaki disease, 88 had coronary artery lesions. Sixty-eight out of 294 patients with typical Kawasaki disease, 20 out of 58 patients with incomplete Kawasaki disease, and 18 out of 44 patients with non-responsive Kawasaki disease had coronary artery lesions. Linear trend analysis showed that the duration of fever in all 352 patients with Kawasaki disease and 294 cases with typical Kawasaki disease was positively correlated with the incidence of coronary artery lesion (Plt;0.05). However, in patients with incomplete Kawasaki disease and non-responsive Kawasaki disease, the relationship between the duration of fever and the incidence of coronary artery lesion was not significant (Pgt; 0.05). Conclusion The longer the duration of fever was in patients with Kawasaki disease, higher the risk of coronary artery lesion.
