The letter from Dr. Salvador Díaz-Lobato and Dr. Sagrario Mayoralas Alises published in Archivos de Bronconeumología1 discuss about the criteria for long-term oxygen therapy (LTOT) and whether these criteria should change depending on the altitude at which the patient lives—an important issue in respiratory medicine, especially because a rather large portion of the population resides at moderate to high altitudes. Let us remember that, in Latin America, at least, many people live at moderate altitude. For example, 1 out of 4 Mexicans lives in metropolitan areas in the Valley of Mexico (2240m), and half of the population (approximately 55 million people) live at more than 1500m above sea level; in Peru, 1 out of 3 inhabitants lives above 2500m, while in the metropolitan area of La Paz, Bolivia, a little more than 2 million people live at an altitude of 3600m.

There is definitely a need for well-designed double-blind randomized clinical trials studies—designed like the Nocturnal Oxygen Therapy Trial Group (NOTT)2 or the Medical Research Council (MRC)3 but conducted at different altitudes—to document which groups could benefit, in terms of survival, from being included in a LTOT program. Lacking such studies, we are marooned in uncertainty and speculation and, as a result, conserving ever-scarce financial resources may be the basis for deciding who should receive oxygen. For example, the deterioration seen with a certain PaO2 level (<55mmHg, for instance) might be similar at any residence altitude, and in that situation—the one underlying most recommendations—prescribing oxygen to improve survival would be the right course of action no matter how many individuals meet the criteria. However, study groups differ in their adaptation to altitude, possibly for genetic reasons. Tibetan populations and the Aymara people of Peru, for example, differ in their response to hypoxemia at altitudes of around 4000m, the latter showing a higher incidence of polycythaemia and pulmonary artery hypertension than the Tibetans, which most likely indicates a less adaptive response.4 The Tibetans not only have a lower PaO2 than the Aymara people but also have increased vasodilator response, which explains their lower incidence of pulmonary artery hypertension.5 Altitude is fertile ground for the expression of natural selection, which has followed at least 2 pathways with different degrees of “efficiency.”

The Díaz-Lobato and Mayoralas letter quotes Dr. Thomas Petty, who said, “If we were to apply the LTOT criteria in Denver, Colorado (1609m above sea level), we would have to put the entire population on O2,” to which there would be financial and logistical rather than physiological or clinical objections. The PLATINO study measured oxygen saturation in a population sample of individuals over 40 years of age in Montevideo (35m above sea level), Caracas (950m), and Mexico City (2240m).6 The main determinants of SpO2 were age, body mass index, FEV1, and—primarily—altitude. The incidence of hypoxemia, defined as an SpO2 of ≤88%, was 6% in Mexico City, which is an enormous proportion, even recognizing the errors in evaluation due to inconsistency between oximetry and gasometry or to variations in its timing.

We believe that oxygen therapy criteria should be adjusted so that limited financial resources are directed to those who would benefit most from the oxygen. This would be based on not only a relatively arbitrary PaO2 or SpO2 level but also demonstration of a consequence of chronic hypoxemia, such as pulmonary artery hypertension or polycythaemia—that is, responses similar to those described for the Aymara people, on average, to higher altitudes, or biomarkers of chronic hypoxia. These are provisional criteria that require scientific validation.

In Mexico, only 8% of individuals with SpO2 ≤88%—an almost universal criterion for LTOT—were receiving supplementary oxygen. In other words, hypoxemia is underdiagnosed in 92%; however, it was also documented that 50% of the individuals on LTOT should not have been receiving it—by the most relevant criterion, at least.6

In summary, we urge studies similar to NOTT or MRC to be conducted at moderate altitudes aiming to demonstrate (1) the benefit of supplementary oxygen and the circumstances under which it is reached and (2) the priority groups to whom limited financial resources for oxygen therapy should be directed.

Under the present circumstances, and in the absence of reliable information, we would consider the priority patient group to be those who have (1) a chronic PaO2 of ≤55mmHg as well as evidence of chronic hypoxemia on non-invasive tests, such as polycythaemia or pulmonary artery hypertension on echocardiogram, or (2) a concomitant disease, such as coronary or cerebral vascular disease that would be exacerbated under conditions of hypoxemia. In the absence of these medical conditions, we run the risk of simply treating the PaO2 and, above all, dramatically expanding the indications for oxygen.