Prediction of pulmonary hypertension related to systemic sclerosis by an index based on simple clinical observations

Objective To develop a score to estimate the risk of developing pulmonary hypertension (PH) in patients with systemic sclerosis (SSc). Methods We first examined the prevalence and characteristics of precapillary PH confirmed by right-heart catheterization in a cross-sectional (derivation) sample of 1,165 SSc patients, and we developed a risk prediction score (RPS) based on simple clinical observations associated with PH. We next prospectively tested the 3-year predictive power of the "Cochin RPS" in a separate (validation) sample of 443 patients presenting with PH-free SSc at baseline. Results In the derivation sample, age, forced vital capacity, and diffusing capacity for carbon monoxide/alveolar volume were independently associated with the presence of PH and were used to create the Cochin RPS. PH developed during followup in 20 patients in the validation sample. The area under the receiver operating characteristic curve of the Cochin RPS was 0.87 (95% confidence interval 0.79-0.95). With a cutoff value of 2.73, patients at risk of PH during followup could be identified with 89.5% sensitivity and 74.1% specificity. PH occurred in 0.6% of patients in the lowest 2 quintiles of the Cochin RPS, in 1.7% of patients in the third and fourth quintiles, and in 17.1% of patients in the highest quintile (P <0.0001 by log rank test). Patients in the highest quintile incurred a >35-fold higher risk of developing PH compared with patients in the 2 lowest quintiles (P = 0.001). Conclusion Using routine clinical observations, we developed a simple score that accurately predicted the risk of PH in SSc.