Endothelin-1 (ET-1) concentration levels have been extensively used for detecting key factors leading to pulmonary vascular disease in patients with congenital heart disease (CHD) and left to right cardiac shunts. In the present study, we consider patients with the aforementioned characteristics, who have normal pulmonary vascular resistances. It has been shown in the literature that a prevalent hemodynamic factor for triggering an increase in the ET-1 levels of these patients is pulmonary artery blood pressure. High values of pulmonary artery blood pressure are defined as pulmonary arterial hypertension (PAH). These could eventually lead to the most advanced PAH form, known as the Eisenmenger syndrome, which is a most devastating condition for patients. Thus, there is a high risk category where patients could eventually develop the syndrome, and a low risk one. This implies that the true distribution of ET-1 levels can be viewed as a two-component mixture. Based on that, performances of various information criteria for estimating numbers of components are evaluated. On the other hand, the likelihood ratio test is shown to apply and its performance is assessed as well. A more homogeneous subpopulation than the initial one is also included in the analysis since it is expected to provide more reliable results. Our study suggests that all statistical methods under evaluation show a nice performance despite the small sizes of the samples in hand.

endothelin-1, pulmonary arterial hypertension, mixture model, information criterion, likelihood ratio test.