Nomogram as a predictive model for depression risk in chronic obstructive pulmonary disease

Abstract

Background Depression after chronic obstructive pulmonary disease（COPD）is associated with mortality rates and poor prognosis. This study aimed to develop a nomogram to identify the risk of depression in patients with COPD based on predictors. Methods The Cross sectional study included 494 COPD aged >20 years who were come from the 2005–2008 National Health and Nutrition Examination Survey database. The 345 subjects from the 2005–2008 survey comprised the development group, and the remaining 149 subjects comprised the validation group. The least absolute shrinkage and selection operator (LASSO) binomial regression model was used to select the best predictive variables before further screening of multivariate regression model.The performance of the nomogram was evaluated on the basis of receiver operating characteristic curve(ROC), calibration curve, and clinical decision curve analysis (DCA). Results We reach a decision that there are 10 item,including BMI,Race,Sex,Age,Education,marriage,hypertension,diabetes,CRP,MONO by LASSO regression model.Multivariate regression had selected 4 statistically significant variables for inclusion.as follow:Hypertension,MONO,CRP,Age.hypertension(Odds Ratio[OR],0.836;95%confidence interval [CI],0.206-0.914; P = 0.028),MONO (OR, -2.652; 95% CI, 0.011 to 0.437; P=0.004), CRP (OR,0.238; 95% CI, 1.047 to 1.538; P=0.015) and Age (OR,0.031; 95% CI, 0.947 to 0.992; P=0.009).The AUC area under the curve for the training group was 0.774 whereas the validation group was 0.713, The predictive model was calibrated, and the DCA showed that the proposed nomogram had strong clinical applicability. Conclusion We have developed a simple nomogram to predict depression in COPD individuals based on Nomogram. External validation is needed to further demonstrate its predictive ability in primary care settings.