Rebuttal to the Con Statement.

Abstract

There is no large overlap in the variables derived from our exercise testing. In any clinical score, for example, the Framingham score for cardiovascular diseases, there is a large overlap for individual makers. For example, you may have a high value of systemic arterial pressure, you may be at low risk if this is not associated with other risk factors such as high cholesterol or inactivity. The proper use of sophisticated statistical analyses such as the one used in our previous study (Richalet et al., 2012) is that you take into account a combination of risk factors. The usefulness of our test cannot be evaluated only by the 7% increase in AUC. The best tool to evaluate the benefits of a new score is the Net Reclassification Index. We showed that adding the physiological variables to the model of prediction allowed us to correctly re-classify 30% of subjects with previous experience at high altitude and 54% of subjects with no previous experience. This is a very clear gain of efficiency, as mentioned in the recent paper validating the risk prediction score (Canoui-Poitrine et al., 2014). Even if only one third of the participants returned their questionnaire, this cohort is still the largest ever studied. Moreover, there was no difference in the clinical or physiological characteristics between the responders and the nonresponders, so that no argument is available to suppose that this could be a bias in the prediction model. Pre-acclimatization is clearly a protecting factor for SHAI. However, from a practical point of view, this is not relevant since people who come to the consultation before a planned trip to high altitude are not acclimatized, unless they live permanently at high altitude, which is rarely the case, at least in Europe (Richalet et al., 2013). The strength of our study is the large number of subjects, which was obtained by pooling all severe cases (severe AMS, HAPE, HACE). This is a crucial point, which derives from our definition of SHAI. Our approach, here, is not mechanistic, looking at the precise pathophysiological causes of HAPE or HACE, although we address the basic common reason of all severe manifestations, i.e., severe hypoxemia, whatever the following effects on the lungs or on the brain. Our approach is very practical for a physician meeting with a person who wants to know his own level of risk. The common trait of all SHAI manifestations is that the subject has to stop his trip and take urgent measures (such as descent, re-oxygenation, drugs, etc.). With the standards of epidemiological studies (cancer, cardiovascular, etc.) no score has ever been proposed with small groups of patients such as the ones mentioned by Bartsch (2014) with pulmonary vascular response to hypoxia or exercise. Therefore, we can bet that there will never be any study capable to propose a specific risk score for HAPE or HACE, due to the small number of cases, unless we develop a very large multicentric study regrouping all mountain medicine consultations around the world. This is our proposal!