Con: Pulse Oximetry Is Useful in Predicting Acute Mountain Sickness

Equipment

Obtaining an accurate measurement in the high altitude environment is a challenge (Windsor, 2012). Take for instance, a trekker sitting at Barafu Camp (4600 m) on Kilimanjaro. He's been told by his guide that a SpO₂ measurement of less than X% puts him at risk of developing acute mountain sickness (AMS) and therefore means that he'll be prevented from making a summit bid. At first he struggles to get any measurement because he's shivering and his finger tips are too cold. Eventually when he's warm enough, he notices that the measurements are changing on a second by second basis. Unknown to him, the interplay between arterial PO₂ and the sigmoidal shape of the hemoglobin-oxygen dissociation curve means that small differences in ventilation quickly affect SpO₂. Indeed, he finds that by hyperventilating for a few seconds his SpO₂ hits 90% or more. At the same time he also notices that by making small changes in his hand position the SpO₂ measurement alters. The high level of ambient ultraviolet light around him is causing interference. Confused by all this, our trekker decides to consult the instruction manual. He is surprised by what he reads (Luks and Swenson, 2011). First, he discovers that the SpO₂ measurements are not entirely accurate. There is an error range of +/− 3% for every measurement he has seen so far. Second, the operating temperature of the device is between 5 and 40 degrees C. After a freezing night spent in a tent at 4600 m our trekker is worried that the device may be too cold. Finally, and perhaps most surprisingly, the measurements obtained have been close to the lower limits of the “operating range” of the device. For its accuracy, the device relies upon comparisons between co-oximetry and pulse oximetry measurements of healthy human volunteers. Since these individuals were never exposed to potentially hazardous levels of hypoxia, each manufacturer has been forced to extrapolate from data obtained at higher values. In reality this means that the device is unreliable since levels of accuracy, precision and bias are likely to be affected below measurements of 80%.

Science

A single SpO₂ measurement gives a momentary snapshot of the degree an individual's hemoglobin is saturated with oxygen. Does this represent oxygenation of the brain, lungs or other organs that are responsible for AMS, high altitude pulmonary edema (HAPE) or high altitude cerebral edema (HACE)? If we assume that SpO₂ reflects the partial pressure of oxygen in the arterial circulation (PaO₂) this may be true to some extent, however this does little to quantify the individual's arterial oxygen delivery (DaO₂), which is largely responsible for shifting the gas to where the high altitude pathology takes place. An accurate measurement of DaO₂ relies upon knowledge of hemoglobin concentration ([Hb]) and cardiac output (Q) measurements. Two trekkers arriving in Barafu Camp with the same SpO₂ will have divergent DaO₂ measurements if their hematological and cardiovascular systems respond to hypoxia in different ways. Of course, we know that knowledge of DaO₂ is not enough to determine how well organs such as the brain and lungs are oxygenated at altitude. DaO₂ simply reflects what is passing through the major arteries and does not take into account the very extensive changes that occur in the arterioles and capillaries that supply individual organs at altitude (Martin et al, 2009). Even if it were possible to measure the delivery of oxygen to individual organs, predicting the development of high altitude illnesses would still prove difficult. We now know that genetics plays a large part in the development of high altitude illnesses such as AMS. On a cellular level, there are those of us who respond better to hypoxia than others (Cheviron and Brumfield, 2012). Therefore, it seems likely that performance will vary from mountaineer to mountaineer, despite similar measurements of SpO₂, DaO₂ and the oxygenation of vulnerable organs. Evidence to support this lengthy explanation can be found when we seek out the answers to two questions:

• Do SpO₂ measurements vary between healthy individuals and those who go on to develop AMS?

There are a growing number of studies that have sought to address this question. The problem for them all is striking the balance between sensitivity and specificity. Most show excellent levels of sensitivity. Let us consider for instance one of the earliest studies performed by Roach et al on Denali in 1998 (Roach et al, 1998). Here, researchers were able to set a SpO₂ “cut off” of 84% and identified all 21 of those who went on to develop AMS. However specificity suffered. On Denali, a further 56 mountaineers were found to have a SpO₂ of less 84%, however not one developed AMS. Sensitivity? 100%. Specificity? 31%. In subsequent studies, at a number of different altitudes, a wide range of SpO₂ limits have been chosen. However the same pattern emerges—a SpO₂ measurement is chosen and all those who develop AMS are identified (high sensitivity). While meanwhile large numbers of mountaineers and trekkers who obtain a low measurement remain symptom free (low specificity). Clearly this combination casts considerable doubt over whether a significant clinical difference in SpO₂ actually exists between those who go on to develop AMS and those that do not. Indeed two of the most recent studies suggest that any difference is small and within the error range of the devices used (Chen et al, 2012; Wagner et al, 2012).

• Do SpO₂ measurements vary with acclimatization?

We know that rates of AMS fall when individuals spend long enough at altitude for acclimatization to take place. If the pulse oximeter is to prove a helpful guide in identifying AMS, it would seem reasonable to expect higher SpO₂ measurements in acclimatized individuals. A recent systematic review has identified 53 published studies that measured SpO₂ in healthy lowland residents and provided a detailed ascent profile (Khan et al, 2012). Using the latest Wilderness Medical Society altitude guidelines it was then possible to gauge, albeit cautiously, whether the members of each study were acclimatized or not. The SpO₂ measurements of both groups were then compared. While a statistical difference was identified (P=0.004), a considerable overlap in measurements was clearly observed. At altitudes between 3000 m and 5000 m no clear pattern emerges. Since pulse oximetry cannot distinguish between those who are acclimatized and those who are not, the chances of identifying those at risk of AMS seems very unlikely.

Predicting AMS is difficult. For a number of practical and theoretical reasons SpO₂ measurements are unlikely to provide the solution. Our trekker on Kilimanjaro should instead monitor himself for symptoms of AMS and adopt an ascent profile that gives him the best possible chance of a safe and successful summit attempt.