Abstract: Objective To validate the indicators used in evaluation of delay of TB patients before initiation of standard TB treatment and related data collection methods and their feasibility in order to provide evidence base for the development of relevant TB control and prevention strategies. Methods Indicators including total delay, health seeking delay, diagnostic delay, treatment delay were defined according to the critical nodes (onset of symptoms, first visit to doctor, TB diagnosis and treatment initiation) during the health seeking process. The typical survey methods were adopted and three counties were selected from the eastern, middle and western provinces each. New smear positive TB patients registered during May 2010 to September 2010 were administered face-to-face self-designed questionnaire survey. A total of 329 smear positive TB patients were enrolled in this study. Results  In this study, male accounted for 81.5%（268/329）, the average age was 51.8 (95% CI:49.9－53.8) years old, farmers accounted for 71.1%(234/329), junior high school education or less were 86.6% (285/329), an annual income of less than 5000 yuan accounted for 81.2% (267/329); The total delay was 36d(0－597d), health seeking delay was 15d(0－594d) and diagnostic delay was 10d(0－429d), between the two there is no significant difference (t=1.596, P=0.111); According to causes of delay, health system induced diagnostic delay was 7d (0－428d), patient induced diagnostic delay was 1d (0－45d), the difference between the two is statistically significant (t=8.609, P=0.000); According to the key nodes in health seeking process, delay in detection of TB suspects was 0d(0－428d), in referral of TB suspects was 4d (0－267d), the difference between the two is statistically significant (t=3.235, P=0.001). Conclusion The indicators designed in this study could give a more in-depth understanding of the causes of the delay.

Key words: Tuberculosis, pulmonary/diagnosis, Delay, Questionnairs