Abstract
To the Editor:
Referring to the recently published RADICAL trial, 1 the authors have alluded to the possibility their sample was statistically underpowered. As a result of both study design and increased education, I feel that this is certainly the case.
The mountaineering community should celebrate the increasing difficulty of studying acute mountain sickness (AMS) prophylaxis as a result of increased education and awareness. However, care should be taken when determining sample size from previous studies, largely because of the changed nature of ascent profiles. A greater emphasis is now placed on acclimatization through a slowed rate of ascent above 3000 m. This may differ from the rates of ascent and the final height attained described in previous papers. An increased rate of ascent and greater final height are thought to be associated with an increased risk of developing AMS. Additionally, severe signs and complications of AMS will become more infrequent as education prompts early recognition and intervention.
Although this study was powered to detect noninferiority and a difference in severity of AMS based on a previous study, 2 the authors have identified that such a study may have biased sample size calculation.
In the study by van Patot et al, participants lived between 1400 and 1600 m and were taken from 2000 m to 4300 m by car over a 2-h period. 2 In contrast, McIntosh et al recorded altitudes of residence of 632±804 m and travel altitudes from 745±803 m. 1 The high-risk period will have been the drive or flight, likely to Lukla airport (2860 m) from Kathmandu (1400 m), which represents a significantly different ascent profile when compared with the study by van Patot et al. The exact nature of this risk is not known because participants followed their own itinerary.
This leads me to offer 2 cautions: 1) the recommendation of 62.5 mg twice daily acetazolamide as prophylaxis should be delayed until a consensus has been reached on its efficacy in reducing incidence and severity of AMS based on studies of suitable sample size; and 2) future studies should ensure they are powered from studies engaging in similar ascent profiles where applicable.