Abstract: Objective  To explore the relationship between the extent of PPD reaction and risk of tuberculosis (TB) among senior high school students,  and to provide scientific evidence for prophylactic treatment object selection.  Methods  1664 students from 2 senior high schools were the target of prospective cohort study. At the beginning of cohort study, the intracutaneous test of PPD and chest fluoroscopy were applied. Chest X-ray was conducted for those with abnormal chest fluoroscopy. People with PPD negative reaction were control group and with PPD positive reaction (diameter of induration in PPD skin test≥5 mm) were exposure group. The exposure group were divided into 5 subgroups according to the extent of PPD positive reaction (5-, 10-, 15-, ≥20 mm and with blister). The Mantel-Haenszel chi-square statistics and Cox regression were applied to analyze the difference of incidence rates between exposure group and control group.  Results  Coxregression model multivariate analysis showed the incidence rates of three subgroups (15-, ≥20 mm and with blister) were 10.14％(14/138), 21.21％(7/33) and 52.00％(26/50),so the diameter of induration ≥15 mm and <20 mm, ≥20 mm and with blaster were risk factors of tuberculosis（Wald χ²=53.366，54.851，124.829，P values were 0.000）, RR(relative risk)  values were 35.55, 63.43 and 161.84, respectively.  Conclusion  The senior high school students of whom diameter of induration in PPD skin test ≥15 mm and with blister were high risk population of tuberculosis and should be the target of prophylactic treatment.

Key words: Tuberculosis,pulmonary, Tuberculin test, Risk factor, Cohort study