Published in:
Open Access 01-12-2021 | Pulmonary Hypertension | Research article
Risk prediction in medically treated chronic thromboembolic pulmonary hypertension
Published in: BMC Pulmonary Medicine | Issue 1/2021
Login to get accessAbstract
Background
At present, there is no generally accepted comprehensive prognostic risk prediction model for medically treated chronic thromboembolic pulmonary hypertension (CTEPH) patients.
Methods
Consecutive medically treated CTEPH patients were enrolled in a national multicenter prospective registry study from August 2009 to July 2018. A multivariable Cox proportional hazards model was utilized to derive the prognostic model, and a simplified risk score was created thereafter. Model performance was evaluated in terms of discrimination and calibration, and compared to the Swedish/COMPERA risk stratification method. Internal and external validation were conducted to validate the model performance.
Results
A total of 432 patients were enrolled. During a median follow-up time of 38.73 months (IQR: 20.79, 66.10), 94 patients (21.8%) died. The 1-, 3-, and 5-year survival estimates were 95.5%, 83.7%, and 70.9%, respectively. The final model included the following variables: the Swedish/COMPERA risk stratum (low-, intermediate- or high-risk stratum), pulmonary vascular resistance (PVR, ≤ or > 1600 dyn·s/cm5), total bilirubin (TBIL, ≤ or > 38 µmol/L) and chronic kidney disease (CKD, no or yes). Compared with the Swedish/COMPERA risk stratification method alone, both the derived model [C-index: 0.715; net reclassification improvement (NRI): 0.300; integrated discriminatory index (IDI): 0.095] and the risk score (C-index: 0.713; NRI: 0.300; IDI: 0.093) showed improved discriminatory power. The performance was validated in a validation cohort of 84 patients (C-index = 0.707 for the model and 0.721 for the risk score).
Conclusions
A novel risk stratification strategy can serve as a useful tool for determining prognosis and guide management for medically treated CTEPH patients.
Trial registration: ClinicalTrials.gov (Identifier: NCT01417338).