Our study has limitations that deserve consideration. First, LVET may be influenced by numerous physiologic and pathologic factors, and, thus, we cannot exclude the possibility that the difference observed in the LVET could be an epiphenomenon reflecting these factors. The LVET varies inversely with heart rate. In our study, the heart rate was similar in survivors and nonsurvivors, but this may be due to type 2 error (n = 9 nonsurvivors). The LVET is also affected by LV loading conditions, age, sex, diurnal and day-to-day variations, and by the type of food. The LVET may be also influenced by associated diseases, including diabetes mellitus, valvular dysfunction, or thyroid abnormalities. Second, it was a single-center study. However, our National Referral Centre for Pulmonary Hypertension is one of the largest pulmonary vascular centers worldwide, recruiting > 50% of French patients treated for PAH or CTEPH. In addition, it is one of the few centers with an ICU devoted to the care of patients with PH. Therefore, this single-center characteristic may also have the advantage of allowing recruitment of consecutive patients with PH with a similar management approach performed by the same team. On the other hand, most of the patients were admitted after a stay in another ICU or pneumology ward across France, and the delay between the start of RV failure management and admission could have influenced the results. Third, at first glance, the size of the study population may appear moderate (n = 53), with a limited number of events (n = 9), and only unpaired, nonparametric comparisons were performed. Importantly, it must be noted that severe precapillary PH is a rare disease. For instance PAH has a low prevalence ranging from 5 to 25 cases per million in France,33 highlighting the major difficulty in obtaining large numbers of patients for clinical studies. Further studies involving a larger population (and, thus, more events) will be needed to help define the most relevant prognosis factors by using multivariate analysis. Fourth, we acknowledge that neither right-sided heart catheterization nor MRI or echocardiography had been performed in these unstable patients with severe PH. Our main pathophysiological hypothesis concerning LVET is, therefore, not documented. We cannot exclude the possibility that there might be an influence of renal failure on BNP evaluation. Fifth, serial measurement of LVET were not performed, such that the precise relationship between LVET, therapies, and outcomes could not be studied, a point that deserves further studies. Sixth, indications for pressors and doses of pressors follow our previously described protocol,2,4 as no evidence-based guidelines are yet available in that setting. The timing of the tonometric evaluation during the management of the patients and the heterogeneity of precise drug administration prevented evaluating the direct effects of the aforementioned therapies on the LVET in every individual.