To the Editors:
We read with great interest the paper by Coker et al. 1 in an issue of the European Respiratory Journal.
The authors conducted the largest prospective study into the outcomes of flights in patients with various respiratory diseases, showing that 18% suffered respiratory symptoms. They concluded that, although air travel seems generally safe in this population of patients under specialist respiratory care, more detailed studies with oximetry monitoring during flight should be performed to determine which patients are most at risk.
The evaluation of the risk of in-flight hypoxaemia has been determined by studies performed in altitude chambers or by hypoxic gas mixture challenge 2. However, extrapolating the findings to real flights may be misleading, due to the higher altitudes attained by newer aircraft, longer duration of altitude exposure and passengers’ activity inside the aircraft 3.
We have evaluated transcutaneous arterial oxygen saturation (Sa,O2) using Pulsox 3iA (Minolta, Tokyo, Japan), a finger pulse oximeter with a 24-h internal memory and dedicated software (Pulsox DS-3; Minolta) for viewing and analysis of the recordings. Data was acquired from 10 healthy adults (medical staff from the Dept of Pulmonology, Faculty of Medicine, University of Porto, Hospital de São João, Porto, Portugal) with a mean±sd age of 38.9±13.1 yrs, of whom four were female and six were male. Measurements were started just before take-off and ended after landing, for 20 commercial long-distance flights (LFs; ≥2 h) and 18 short-distance flights (SFs; <2 h). Each subject wore the oximeter in a wristband and was able to mark an event by pressing a small button on it each time they were involved in an activity, such as eating, walking and going to the lavatories. After the return flight, all stored data were downloaded on to a computer, producing a cumulative distribution graph (Sa,O2 and pulse rate, with events marked) and saturation parameters analysis (mean, minimum, Sa,O2 dips, time spent with Sa,O2 <90%).
The investigation was carried out in different aircraft models of European and American airline companies in round-trip flights: 10 flights in Airbus A320; nine flights in Airbus A310; one flight in Airbus A319; eight flights in Fokker 100; three flights in Boeing 737; two flights in Boeing 757; one flight in Boeing 777; two flights in McDonnell Douglas; and two flights in BA200. Examples of LFs were Lisbon–New York, Los Angeles–Paris, New York–San Francisco and Porto–Berlin; SFs were Lisbon–Rome, Porto–Paris, Porto–Lisbon and New York–Toronto.
The mean±sd (range) monitoring duration for LFs was 5.1±2.2 (2.2–9.2) h and for SFs was 1.0±0.6 (0.5–2.0) h. On average, the maximal desaturation (difference between baseline and minimum Sa,O2) achieved was 12.8±6.3% for LFs and 4.2±2.6% for SFs (p = 0.001). While in all LFs, subjects reached a minimum Sa,O2 ≤90% (mean±sd 85.5±5.3%), in only four (22.2%) SFs did the subjects develop a minimum Sa,O2 <90% (mean±sd 93.0±2.7%). Although the proportion of time spent with Sa,O2 <90% was small (0.7±0.9% in LFs), in some flights (two trans-Atlantic and one Porto–Berlin) this level was attained for >5 min. In some cases, we also found repetitive oscillations in Sa,O2 signal (fig. 1⇓); in fact, considering 4% dips, LFs had significantly more dips per hour than SFs (3.1±3.0 versus 0.4±0.8; p<0.001). On three occasions (on a New York–Lisbon flight and in two subjects on a Porto–Berlin flight) it reached >5 dips·h−1. By visual inspection of the Sa,O2 trends during LFs, we could not see a stabilisation of Sa,O2 at maximum altitude in all cases: in four cases, Sa,O2 progressively declined until the flight descent began.
Changes in transcutaneous arterial oxygen saturation (Sa,O2) of one subject during a flight from Lisbon to New York. Arrows indicate events as follows: take-off, lunch, lavatory visit and landing. The baseline Sa,O2 was 98%; after take-off it dropped to 95% (#). In the lavatories there was an intense desaturation to 91%. When descent began, Sa,O2 increased, attaining 98% at landing. Apart from progressive decline of Sa,O2, repetitive 4% dips were also seen (1.5 dips·h−1). ○: movement artefacts. ¶: the last movement artefact coincided with an abrupt fall in Sa,O2. Measurement started at 14:00 h.
We also observed that Sa,O2 dropped during meals, walking along the aisle and especially in the lavatories (fig. 1⇑). In fact, in seven LFs it was possible to mark an event while the subject was in the lavatories, and Sa,O2 on that occasion was always close to the minimum (range 83.3–92.8%).
Oxygen desaturation values observed in our study seem variable between subjects, flights and planes. The greater desaturation during LFs in healthy subjects is consistent with the data of Coker et al. 1, who reported worsening symptoms mainly in patients in whom average flight duration was 7.6 h. As intermittent hypoxia is an effective stimulus for evoking cardiorespiratory responses, repetitive 4% desaturations (an original finding) may also contribute to the detrimental effects of air travel 4.
In conclusion, we believe that oxygen saturation levels obtained in real life may be very useful for monitoring the health impact of flying and an important measurement to better determine patients at risk.
Statement of interest
None declared.
- © ERS Journals Ltd
