Does your COPD patient need in-flight oxygen? New algorithm may help

Who Needs In-Flight Oxygen? New Method May Help

by Brett Ley, MD

COPD patients without a long-term indication for supplemental oxygen may still be at risk for severe hypoxemia during air travel since cabin pressures are generally maintained to simulate altitudes of about 8000 feet. In-flight supplemental oxygen is recommended when the partial pressure of arterial oxygen (PaO2) is expected to fall below 50 to 55 mmHg, but predicting which patients are at risk can be difficult. Current guidelines rely on resting oxygen saturation and assessment of clinical risk factors. Prediction equations also exist but often overestimate risk. The gold standard is the high altitude simulation test (HAST), however this test is not widely available. Therefore, Anne Edvardsen et al developed a clinical algorithm based on resting and exertional pulse oxygen saturation to help determine which COPD patients should be prescribed in-flight supplemental oxygen, potentially limiting the number of patients requiring HAST.

What They Did

This was a prospective cohort study of 100 patients with moderate to very severe COPD (i.e. FEV1 < 80%) referred to a single center in Norway for pre-flight evaluation. A similar cohort of 50 patients was used for validation. All patients underwent HAST (breathing 15.1% FiO2 to simulate 8000ft); and pulmonary function tests, pulse oximetry, arterial blood gas, and a six-minute walk test (6MWT) variables (all done at sea level) were evaluated for their ability to predict a positive HAST (defined as PaO2 < 50 mmHg—BTS threshold for recommending in-flight oxygen). A clinical algorithm was then constructed with a focus on the best performing, non-invasive predictor variables, namely resting room air O2 saturation (SpO2) and lowest 6MWT saturation (6MWT SpO2).

What They Found

Here's the algorithm: If SpO2 < 92%, then prescribe in-flight O2. If SpO2 92-95%, then perform 6-minute walk test (6MWT):

• If 6MWT SpO2 < 84%, then prescribe in-flight O2.

• If 6MWT SpO2 >= 84%, then perform high-altitude simulation test (HAST).

If SpO2 >95%, then perform 6MWT:

• If 6MWT SpO2 < 84%, then perform HAST.

• If 6MWT SpO2 >= 84%, then in-flight O2 not recommended.

Here's how it performed in the derivation and validation cohorts, respectively:

• Sensitivity was 99% and 100%, specificity was 82% and 80%

• 33% and 40% were recommended to undergo HAST; meaning HAST would have been avoided in 67% and 60%

• One patient and zero patients (<1% overall) were misclassified as not needing in-flight O2 by the algorithm, but would have by HAST

• Five patients and four patients (6% overall) were recommended in-flight O2 by the algorithm, but did not need it based on HAST. However, these patients had a mean HAST PaO2 near the cut-off value.

What It Means

This algorithm conveniently categorizes GOLD stage II-IV COPD patients who plan to fly into one of three groups:

1. Prescribe O2 without further testing

   • resting SpO2 of < 92%

   • resting SpO2 92-95% AND 6MWT SpO2 < 84%

2. Perform HAST and prescribe O2 based on result

   • resting SpO2 92-95% AND 6MWT SpO2 >= 84%

   • resting SpO2 > 95% AND 6MWT SpO2 < 84%

3. No O2, no further testing

   • resting SpO2 > 95% AND 6MWT SpO2 >= 84%

The test characteristics for the algorithm (high sensitivity and moderate specificity) seem appropriate for this clinical application, i.e., it is probably more important to not miss patients who need O2 on the plane at the expense of giving it to a few who might not have needed it. Also, the algorithm is largely in keeping with current BTS guidelines, implementing resting SpO2 as a first step, and essentially substituting clinical risk factor assessment with six-minute walk test (though consideration of clinical risk factors should probably remain a part of the evaluation process). It goes further in capturing those with a resting SpO2 > 95% who might need O2. Several limitations and remaining questions should be noted:

1. The algorithm only applies to those tested at sea level.

2. The algorithm only applies to situations where 6MWT is performed according to guidelines, and does not apply to those who cannot perform a 6MWT.

3. It reduces the number of patients who need HAST by nearly two-thirds, but does not eliminate the role of the test entirely.

4. This study does not address how the algorithm performs in other chronic cardiopulmonary diseases.

5. The study was small, and I would guess that the Norwegian population studied here was not terribly ethnically or racially diverse (table 1 did not include these characteristics), so I would hesitate to apply the algorithm to other populations without broader validation.

6. The study uses a surrogate outcome (HAST PaO2) rather than actual complications during air travel; however, severe complications are actually pretty rare and a study based on these outcomes may not be feasible or ethical.

If it does perform similarly in other chronic cardiopulmonary diseases and diverse patient populations, this algorithm could become a valuable risk assessment system for evaluating patients at risk for severe in-flight hypoxemia. I’ll be interested to see if future guidelines pick this up.

Clinical Take-Away: This algorithm for the pre-flight evaluation of COPD patients that utilizes resting room air (sea level) oxygen saturation and six-minute walk test desaturation may effectively discriminate who needs in-flight supplemental oxygen, and reduce the number of patients requiring more sophisticated testing with HAST. Anne Edvardsen et al. Air travel and chronic obstructive pulmonary disease: a new algorithm for pre-flight evaluation. Thorax 2012;67:964-969.