Nomograms to predict severe PH and survival in COPD patients using non-invasive parameters

This study aimed to develop simplified nomograms for predicting the likelihood of severe pulmonary hypertension (PH) in chronic obstructive pulmonary disease (COPD) patients and for predicting survival in COPD-associated pulmonary hypertension (COPD-PH) patients.

A total of 179 COPD patients (128 without severe PH, and 51 with severe PH) were analyzed at the Shanghai Pulmonary Hospital Affiliated with Tongji University between 2013 and 2022. Variables, including demographic data and clinical examination findings, were collected. A multivariable logistic regression analysis was used to identify statistically significant PH variables for establishing a nomogram model. The multivariate Cox hazard analysis identified the predictors of death or lung transplantation, which were used to construct a nomogram. The models were evaluated based on discrimination, calibration, and clinical efficacy using the concordance index (C-index), calibration curve, and decision curve analysis.

Peripheral capillary oxygen saturation at peak (peak SpO₂), peak oxygen consumption per kilogram (peak VO₂/kg), peak heart rate (peak HR), and pulmonary arterial systolic pressure (PASP) were associated factors for severe PH based on the multivariate logistic regression analysis and were used to develop a nomogram. The C-index for the training and validation cohorts was 0.906 (95% CI: 0.85-0.96) and 0.93 (95%CI: 0.85-1.00), respectively. The areas under the receiver operating characteristic (ROC) curve were 0.906 and 0.93 for the two cohorts. Predictors included in the survival nomogram model were age, diffusing capacity for carbon monoxide percentage of measured to predicted value (DLCO% predicted), and minute ventilation/carbon dioxide output slope (VE/VCO₂ slope). The model was constructed to predict 1-, 2-, and 3-year survival. The C-index of the nomogram for the training and validation cohorts was 0.80 (95% CI: 0.71-0.89) and 0.69 (95% CI: 0.52-0.86), respectively. The calibration plots were close to the diagonal line in both cohorts. Decision curve analysis (DCA) showed the nomogram model provided a good net benefit.

The nomogram models based on clinical variables from the non-invasive testing offer an individualized tool to predict severe PH in COPD patients and survival in COPD-PH patients.