Risk assessment of chronic obstructive pulmonary disease using a Bayesian network based on a provincial survey

Abstract

Introduction The diagnosis of chronic obstructive pulmonary disease (COPD) is based on spirometry tests, which are difficult to perform in some populations. Objectives We aimed to construct a risk assessment model using a Bayesian network (BN) that would enable screening of high-risk populations. Patients and methods A provincial survey of COPD was performed via face-to-face interviews and spirometry tests among a population aged 40 years and above in Liaoning Province, northeastern China. The potential risk factors were initially identified by multivariable logistic regression, and then a BN was built. To verify its performance, cross-validation and external dataset validation were performed, and the area under the curve (AUC) and accuracy of the BN were calculated. Result s The estimated age-adjusted prevalence of COPD in the entire population was 21.23% (95% CI, 18.35-24.11). Logistic regression revealed that low education level (odds ratio [OR], 2.35; P <0.001), elderly age (OR, 4.19; P <0.001), ever smoking (OR, 1.49; P = 0.03), and lower level of satisfaction with air quality (OR, 1.55; P = 0.03) were associated with COPD. In the BN, frequent cough was the strongest single risk factor for COPD development (risk, 0.374). The risk increased as more factors were specified, and the highest risk (0.738) was noted for the combination of older age, smoking, wheezing during sickness, and frequent cough. The cross-validation indicated that BN performed better than logistic regression, with a mean AUC of 0.85 and an optimum accuracy of 0.87 (cutoff, 0.473). Conclusions The BN based on questionnaires had a more accurate performance in predicting the risk for COPD. The increased risk associated with exposure to a combination of several risk factors should be noted.