Study On The Prediction Model Of GOLD Grade Level In Patients With Qi Deficiency Constitution In Stable COP

Purpose:To explore the influencing factors of stable progression of chronic obstructive pulmonary disease.To predict the GOLD grade level of patients with stable COPD qi-deficiency constitution,the regression analysis method was used to establish the prediction model of GOLD grade level of patients with stable COPD qi-deficiency constitution,and the verification was carried out.Material and method:This study collected 165 patients with stable COPD qideficiency constitution who met the standards of intake and discharge.Clinical case data of patients were collected to establish a database.The model was established in 80%of patients（132 cases）and validated in 20%（33 cases）.132 patients were divided into two groups （GOLD2 and GOLD3）according to the criteria of severity of airflow restriction.General information,personal history,past medical history,physical examination,TCM syndrome,tongue,pulse and laboratory examination were compared between the two groups.In order to predict and screen the independent variables,single factor analysis and binary logistic regression analysis are used to determine the independent variables into the prediction equation,establish the prediction model,and present the model in the form of formulas.Hosmer-lemeshow goodness of fit test and other methods were used to evaluate the model as a whole.The predictive threshold of the model was calculated by statistical method ROC curve,and the data of the validation group（33 cases）obtained by random assignment were used to verify the model.Results:Of 165 patients,132（80%）were established models and 33（20%）were validated models.Among 132 patients,95 cases were Grade GOLD2 and 37 cases were Grade GOLD3.Through binary logistic regression analysis,the final results screened out smoking history,drinking history,cough and shortness of breath as the influencing factors of Qi-deficiency constitution in COPD stable period.Among the 132 modeling cases,there were 37 cases of GOLD2 and 23 cases of GOLD3 smoking history between the two groups（P<0.05）.There were 25 cases of GOLD2 and 17 cases of GOLD3 in drinking history between the two groups（P<0.05）.Cough and shortness of breath were tested by Fisher’s exact test respectively, indicating significant differences（P<0.05）.Only P<0.05 was found in the tongue between the two groups,indicating that the significant difference was in line with the basic principle of random grouping.The prediction model is established as follows: ln（P/1-P）=-7.770+1.707x₁-1.995x₂+1.341x₃+1.411x₄+2.413x₅+1.193x₆+2.752x₇+3.811x₈x₁=smoking history,x₂=drinking history,x₃=mild cough,x₄=moderate cough,x₅=severe cough,x₆=mild shortness of breath,x₇=moderate shortness of breath,x₈=severe shortness of breath.The results of Hosmer-Lemeshow goodness of fit test of this model were P=0.552, suggesting that the model with this level of grading had better effect.AUC=0.861,P>0.75,indicating that the model’s discrimination is fair.Cox&Snell R square was 0.317, Nagelkerke R square was 0.456,the model accuracy（83.5%）,sensitivity（59.5%）,specificity（92.6%）,Jordan index（0.50）,positive likelihood ratio（8.04）and negative likelihood ratio（0.44）,The positive likelihood ratio is high.All the above results indicate that the model has certain predictive function.According to the ROC curve,calculate the prediction critical value of the prediction model P=0.4354.Internal validation showed that the accuracy, sensitivity and specificity were 81.8%,91.7%and 55.6%respectively,and AUC=0.903.The predicted classification was in good agreement with the actual situation,indicating that the model had good discrimination ability.Conclusion:1.Smoking history,drinking history,cough,shortness of breath are the possible influencing factors of COPD patients with qi deficiency constitution.2.The equation of GOLD grade prediction model for patients with chronic obstructive pulmonary disease at stable stage is as follows: ln（P/1-P）=-7.770+1.707x₁-1.995x₂+1.341x₃+1.411x₄+2.413x₅+1.193x₆+2.752x₇+3.811x₈ x₁=smoking history,x₂=drinking history,x₃=mild cough,x₄=moderate cough,x₅=severe cough,x₆=mild shortness of breath,x₇=moderate shortness of breath,x₈=severe shortness of breath.The prediction critical value of the prediction model was P=0.4354,that is,P<0.4354was equivalent to GOLD gradeⅡof qi-deficiency in copd stable stage,and P≥0.4354 was equivalent to GOLD gradeⅢof QI-deficiency in COPD stable stage.3.The area under the ROC curve of the prediction model is 0.861,and the prediction results are practical to a certain extent.After verification,the area under ROC curve was 0.903.The prediction of classification is basically consistent with the actual classification,indicating that the model has good discrimination ability.