Elsevier

Respiratory Medicine

Volume 109, Issue 8, August 2015, Pages 1019-1025
Respiratory Medicine

Clinical Trial Paper
Chronic obstructive pulmonary disease and coronary disease: COPDCoRi, a simple and effective algorithm for predicting the risk of coronary artery disease in COPD patients

https://doi.org/10.1016/j.rmed.2015.05.021Get rights and content
Under an Elsevier user license
open archive

Highlights

  • Nowadays there are no tools to predict in a quick, simple and non-invasive manner the risk of coronary heart disease in COPD.

  • The aim of this study was to identify a tool for predicting the risk of CAD in COPD patients.

  • COPDCoRi is an algorithm for predicting the risk of CAD in COPD patients via a rapid, inexpensive and non-invasive approach.

  • COPDCoRi has a high diagnostic accuracy and a good profile for both sensitivity and specificity.

Abstract

Background

Chronic obstructive pulmonary disease (COPD) is often associated with cardiovascular artery disease (CAD), representing a potential and independent risk factor for cardiovascular morbidity. Therefore, the aim of this study was to identify an algorithm for predicting the risk of CAD in COPD patients.

Methods

We analyzed data of patients afferent to the Cardiology ward and the Respiratory Diseases outpatient clinic of Tor Vergata University (2010–2012, 1596 records). The study population was clustered as training population (COPD patients undergoing coronary arteriography), control population (non-COPD patients undergoing coronary arteriography), test population (COPD patients whose records reported information on the coronary status). The predicting model was built via causal relationship between variables, stepwise binary logistic regression and Hosmer–Lemeshow analysis. The algorithm was validated via split-sample validation method and receiver operating characteristics (ROC) curve analysis. The diagnostic accuracy was assessed.

Results

In training population the variables gender (men/women OR: 1.7, 95%CI: 1.237–2.5, P < 0.05), dyslipidemia (OR: 1.8, 95%CI: 1.2–2.5, P < 0.01) and smoking habit (OR: 1.5, 95%CI: 1.2–1.9, P < 0.001) were significantly associated with CAD in COPD patients, whereas in control population also age and diabetes were correlated. The stepwise binary logistic regressions permitted to build a well fitting predictive model for training population but not for control population. The predictive algorithm shown a diagnostic accuracy of 81.5% (95%CI: 77.78–84.71) and an AUC of 0.81 (95%CI: 0.78–0.85) for the validation set.

Conclusions

The proposed algorithm is effective for predicting the risk of CAD in COPD patients via a rapid, inexpensive and non-invasive approach.

Keywords

COPD
Coronary artery disease
Comorbidities
Predictive algorithm

Cited by (0)