Long-term monitoring of oxygen saturation at altitude can be useful in predicting the subsequent development of moderate-to-severe acute mountain sickness.

ABSTRACT

OBJECTIVE: The use of pulse oximetry (Spo2) to identify subjects susceptible to acute mountain sickness (AMS) is the subject of debate. To obtain more reliable data, we monitored Spo2 for 24 hours at altitude to investigate the ability to predict impending AMS. METHODS: The study was conducted during the climb from Alagna (1154 m) to Capanna Regina Margherita (4559 m), with an overnight stay in Capanna Gnifetti (3647 m). Sixty subjects (11 women) were recruited. Each subject was fitted with a 24-hour recording finger pulse oximeter. The subjects rode a cable car to 3275 m and climbed to 3647 m, where they spent the night. RESULTS: In the morning, 24 subjects (6 women) had a Lake Louise Questionnaire score (LLS) ≥ 3 (AMS(+)), and 15 subjects (4 women) exhibited moderate-to-severe disease (LLS ≥ 5 = AMS(++)). At Alagna, Spo2 did not differ between the AMS(-) and AMS(+) subjects. At higher stations, all AMS(+) subjects exhibited a significantly lower Spo2 than did the AMS(-) subjects: at 3275 m, 85.4% vs 87.7%; resting at 3647 m, 84.5% vs 86.4%. The receiver operating characteristics curve analysis resulted in a rather poor discrimination between the AMS(-) subjects and all of the AMS(+) subjects. With the cutoff LLS ≥ 5, the sensitivity was 86.67%, the specificity was 82.25%, and the area under the curve was 0.88 (P < .0001) for Spo2 ≤ 84% at 3647 m. CONCLUSIONS: We conclude that AMS(+) subjects exhibit a more severe and prolonged oxygen desaturation than do AMS(-) subjects starting from the beginning of altitude exposure, but the predictive power of Spo2 is accurate only for AMS(++).