The Study On Elementary Diagnosis Method Of Chronic Obstructive Pulmonary Disease For The Community Clinic Unit In China

Objective: Chronic obstructive pulmonary disease (COPD) is characterized by an incompletely reversible limitation in airflow. It is a major cause of mobidity and mortality over the world. However, it is often underdiagnosed and undertreated, because the spirometric is not popular in some hospitals, especially in community hospitals and clinics. This study would establish a model for initial prognosis analysis of COPD, which woud be suitable to apply in the community and basical hospital in China.Methods: We first recruited the patients of COPD according to GOLD diagnostic criteria, and the normal people as control by spirometric test and questionnaires. All recruited patients were invited to participate in the study completing the questionnaire and pre- and post-bronchodilator spirometry testing. According the information of these patients, we establish a discriminant function by discriminate analysis based on Bayes' Rule for initial prognosis analysis of COPD. Further we developed the software by JSP + Servlet and Tomcat 6.0 for initialprognosis analysis of COPD. Results:1.A total of 243 patients of COPD were recruited in our study, 216 male and 29 female, ages from 40 to 86 yrs, I, II grade 118, III, IV grade 125. There were total 112 people in control group, 52 female and 61 male, and age from 40 to 79 yrs.2. We determined the following variables by stepwise discriminate analysis: age; gender; index of smoking; body mass index; occupational hazard; living environment; gasping; cough; dyspnea scale as the discriminatory factors. 3. According the discriminatory factors, we establish a discriminant function that could be used to predict the situation of COPD by using the SAS soft ware. For control group:Y0= -72.84619 + 10.42377 X1+ 9.69731 x2+ 0.58220 x3 -0.68222 x5+ 0.00540x7 -0.00684 x8+ 3.30434x11 -0.13336 x12+ 2.60295 x14. For grade I and II COPD patients:Y1= -72.00038 + 12.33394 x1+ 7.73661 x2+ 0.72004 x3 -0.52833 x5+ 0.00786x7 -0.00281 x8+ 2.83796x11+ 0.33442 x12+ 2.02699 x14. For grade III and IV COPD patients : Y2=-72.17307+12.44860x1+ 7.52607x2+0.74809x3-0.11923x5+0.00955x7+0.02361x8+2.64365x11+0.38577x12+3.38533x144. Internal validation: The results showed that the observed agreement of back substitution for there group were 82.29%, 62.7%, and 71.2%, respectively. Prospective validation: 75.3% training sample and 67.0% validation sample were predicted correctly.5. We used 150 patients to test the model; seventy percent of these patients were predicted correctly.6. We developed the software which can be used to predict COPD.7. We used 60 patients to test the software, seventy-eight percent of these patients were predicted correctly.Conclusions: We determined the risk factor of COPD by stepwise screening. According this factor, we establish a discriminant function by discriminate analysis based on Bayes' Rule for initial prognosis analysis of COPD. Further we developed software for initial prognosis analysis of COPD. That software could be used as a simple tool for elementary analysis of state of COPD. It is suitable to apply in the community and basical hospital in China.