Application Of Survival Analysis In TB Control In Shanxi Province

Object This study was carried out to find a good management method to treat tuberculosis (TB) patient. World health organization (WHO) recommended DOTS (Directly Observed Treatment Short course) strategy in 1991 for controlling TB. DOTS was recommended in Shanxi province in 1994. There are many TB patients in Shanxi and these patients live in rural area, where there are few health service. It's hard to practice DOTS strategy. It's necessary to find an effective way to take place DOTS. So we carried a special study.Method The material of TB control and its follow-up studies can be sorted as survival data. At present, the multivariate analysis for survival data is Cox's proportional hazards regression model. It was proposed by D.R.Cox in 1972. It can not only use data with censoring and free distribution, but also find risk factors efficiently with eliminating mixed factors. Basic on this point, this paper uses Cox model to screen the risk factors in TB control.First of all, we used Kaplan - Meier method, then Cox model's forward stepwise and backward stepwise methods to screen prognostic factors.Result Cox model show that sex, the classification and the result of sputum smear are principal factors in delaying sputum conversion among initial and re-treatment patients. Female cases are easier to cure than male cases, their median are 24and 32days, respectively. 95% CI are (20.83,27.17) and (15.72,48.28), respectively. The initial treatment cases are faster to cure than the re-treatment cases, their median are 23 and 46 days, respectively. 95% CI are (20.07,25.93) and (35.81,56.19), respectively.Conclusion This study demonstrates that in TB control, sex, the classification of cases and quantity of TB bacilli have significant effect on smear conversion. The more TB bacilli in sputum, the more difficult to cure the patient. The model of management hasn't shown significant difference.