Author: Richard Johnston

While reviewing reports today I ran across a couple of lung volume tests from different patients where the SVC was over a liter less than the FVC. Suboptimal SVC measurement can affect both the TLC and the RV and in one case the TLC was slightly below normal (78% of predicted) and in the other the TLC was within normal limits but the RV was over 150% of predicted. Both patients had had lung volume measurements previously and the current TLC was significantly different than it had been before.

I seem to run across this problem at least once a week so I am reasonably used to making manual corrections. I’ve discussed this previously but basically I use the position of the tidal loop within the maximal flow-volume loop obtained during spirometry to determine IC and ERV and then re-calculate TLC and RV accordingly.

Anyway, for this reason I had tidal loops, and IC and ERV on my mind while I was reviewing other reports. Shortly after this I came across a report that had “fair FVC test quality and reproducibility” in the tech notes so I pulled up the raw spirometry test data and took a closer look.

What I found was that the patient had performed five spirometry efforts and that the FVC and FEV1 was different on each test. All five spirometry efforts met the ATS/ERS criteria for back-extrapolation, expiratory time and end-of-test flow rates. I clicked back and forth between the different spirometry efforts to make sure the right FVC and FEV1 had been selected and when I did I noticed that the position of the tidal loop was shifting left and right and that the closer it was to TLC, the lower the FVC and FEV1 were and vice versa.

We almost never use IC and ERV when talking about spirometry, but using the end-exhalation of the tidal loop as a marker for FRC I noticed that the ERV was pretty much the same for all efforts and that it was the IC that was changing the most. This told me that although the patient was likely exhaling completely with each effort, their inspiration wasn’t as maximal as it should have been every time.

Almost all of the ATS/ERS criteria for assessing the quality of a spirometry effort (i.e. back extrapolation, expiratory time, end of test flow rates) are oriented towards a maximal exhalation. But other than stressing that the FVC maneuver should start from TLC (i.e. a maximal inspiration) there are no criteria whatsoever for assessing the quality of an inspiratory effort. In a sense this is not surprising since there is nothing in particular in a flow-volume loop or volume-time curve that can indicate that an inspiration has truly been maximal.

In this instance however, when the spirometry efforts were compared with each other the change in IC was a strong indicator for the quality of each effort’s inspiration. By itself IC can’t be the sole indication of a maximal inspiration, but I can’t help but think that if the technician performing the test was aware that IC was changing while ERV was remaining fairly constant that they would have spent more time coaching the patient on the inspiratory part of the maneuver. But IC and ERV aren’t measured from an FVC maneuver and as importantly, the conventions that are used when displaying flow-volume loops makes it difficult to see changes in these values.

Specifically, although our lab software (and probably the software from most of the other manufacturers of spirometers and lab systems) can overlay the flow-volume loops from different spirometry efforts the standard convention is to position the flow-volume loops using the maximum inspiration of each effort.

When flow-volume loops are displayed this way it’s hard to compare the IC and ERV from each effort even if you have a really good eye. But if flow-volume loops were positioned according to the end-exhalation of the tidal loops, IC and ERV would become much more apparent.

Using IC and ERV to help assess the quality of spirometry maneuvers would require some changes in the way that we think about the test. Strictly speaking however, even though measuring the numerical values for IC and ERV would be useful, all that’s really necessary is to change the way flow-volume loops are displayed (or at least have the option) since that alone would make the differences more evident.

I fully understand why IC and ERV should be measured from a slow VC when lung volumes are being calculated. Relaxed and stable tidal breathing is needed for a repeatable FRC and the steady effort of an SVC maneuver is needed to obtain accurate and repeatable IC and ERV measurements. Tidal breathing is not included in the ATS/ERS spirometry standards and strictly speaking the quality of any pre-maneuver tidal breathing is not a necessary component towards obtaining a quality FVC.

Having said that, I’ve always felt that starting an FVC maneuver with a couple tidal breaths was one way to to make sure the patient wasn’t leaking around the mouthpiece and it also lets me cue them better for the FVC maneuver. I have almost always started the FVC with at least a couple tidal breaths and this is the approach I have always taught to others. So for me at least, as well as most of the spirometry performed in my lab (and I suspect many other labs as well), tidal loops are usually included with each FVC effort.

Given the circumstances surrounding the FVC maneuver I suspect that the IC’s and ERV’s obtained from an FVC are likely less repeatable and more variable than those obtained from an SVC but interestingly enough however, I’ve never seen any study that looked at this or at the repeatability of the IC and ERV from SVC maneuvers, so this will have to remain speculation on my part.

Looking back at my original problem (the one that got me thinking about tidal loops, IC and ERV in the first place) having a numerical value for IC and ERV from the FVC would make it easier to re-calculate lung volumes. As importantly I’ve seen lab software where the FVC was automatically substituted for the SVC when it was larger, but a distinct problem with this approach is how the FVC is inserted into the lung volume calculations. Since IC and ERV aren’t measured from the FVC maneuver what usually happens is that the RV calculated using the ERV from the original SVC is retained (no matter how bad the SVC was) and a new TLC is calculated from the FVC and RV. Since the SVC is being replaced in these circumstances because it is underestimated to begin with (and at a guess the ERV is more likely to have been underestimated and RV overestimated) then calculating a TLC using a questionable RV means the TLC will also be questionable. This is where having an IC and ERV from the FVC maneuver are more likely to lead to a more accurate RV and TLC than using the IC or ERV from the SVC it’s replacing.

We’re not used to thinking about IC and ERV in conjunction with spirometry but whenever an FVC maneuver is preceded by tidal breathing it’s possible to measure IC and ERV. Making the IC and ERV more visually apparent by positioning flow-volume loops based on the end-exhalation of the tidal loop and not the maximal inspiration gives additional information about the quality of a patient’s inspiratory and expiratory efforts that is not presently available. For this reason I’d like to see this as an option when overlaying spirometry efforts during testing. I’d also suggest that the next time the ATS/ERS revises the standards for spirometry that tidal breathing, at least an option, should be included as part of the FVC maneuver.

PFT Blog by Richard Johnston is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License

I’ve been interested in ventilation inhomogeneity for a while and as ways to measure it I have looked at VA/TLC ratios, the Lung Clearance Index (LCI) and the phase III slope of the single-breath N2 washout (SIII_N2). All of these tests are able to provide some information about ventilation inhomogeneity but each has their own limitations and just as importantly although their results have a relatively clear relationship with ventilation inhomogeneity it’s not quite as clear what exactly it is they are measuring. A friend recently pointed me to an on-line article in Chest that discusses the dual-tracer single-breath washout test in patients with COPD. The apparent advantage of this test is that it is able to provide information about the site of the ventilation inhomogeneity. Although dual tracer gases have been used to study airway function for over 50 years the limitation of this technique has been the need to use a mass spectrometer. Some recent advances in technology have made it possible for this type of testing to be performed with a significantly less expensive gas analyzer and this has revived an interest in the dual-tracer gas single-breath washout (DTG-SBW).

The two tracer gases in question are Helium and Sulfur Hexaflouride (SF6). Helium has a density of 4 gm/mol and the density of SF6 is 146 gm/mol, and it is the difference in densities between these two inert and insoluble gases that make this test useful. In order to understand why we need to revisit to the anatomy of the terminal airways.

As the airways branch from the trachea towards the alveoli up to the fifth generation the cross-sectional area remains about the same (which is also why the majority of airway resistance occurs within the first five generations) but after this each succeeding generation shows an increase in cross-sectional area.

When air flows into the lung during an inhalation, air velocity is highest in the trachea but because of the increase in cross-sectional area of the airways air velocity decreases as it travels deeper into the lung. Somewhere around the respiratory bronchioles and acinus the convective air velocity is so low that gas molecules travel faster by diffusion than they do by convection. The point at which this occurs has been called the diffusion wavefront. Up to the diffusion wavefront convective transport will be equal regardless of gas density but the speed or rate of diffusion in any given gas is dependent on its density. Heavier molecules diffuse more slowly than lighter ones and the diffusion coefficient for helium is approximately 6 times higher than it is for SF6. This means that the diffusion wavefront for helium usually occurs in the terminal or respiratory bronchioles and for SF6 it usually occurs somewhere inside the acinus.

At the end of a tidal breath that contains helium and SF6 the difference in diffusion wavefronts causes the SF6 concentration to elevated in the terminal and respiratory bronchioles and for helium to be more evenly distributed through the bronchioles and the acinus. During exhalation therefore, the ratio between SF6 and helium will be highest at the beginning of exhalation and decrease thereafter. An increase in ventilation inhomogeneity in the terminal bronchioles (as opposed to the acinus) will therefore act to further separate the SF6 and helium and steepen the expiratory slope.

The DTG-SBW is performed during tidal breathing and only over a single tidal breath. The subject breathes on a gas circuit with a flow sensor, O2 analyzer, CO2 analyzer and an ultrasonic mass molar analyzer. After several normal and reproducible tidal breaths the blow-by breathing circuit is switched from air to a mixture containing helium and SF6 and the subject takes a normal tidal breath. At end-inhalation the breathing circuit is switch back to air.

The ultrasonic mass molar analyzer measures gas density, not individual gas concentrations. For this reason the test gas mixture usually contains 5% SF6, 26.3% helium, 21% O2 and balance N2 which has a density that is essentially the same as air. During air breathing the changes in density that occur during exhalation are primarily due to CO2. After the inhalation of the DTG-SBW gas mixture the exhaled CO2 signal is used to subtract the contribution of CO2 to the exhaled gas density. This produces a calculated signal which primarily reflects the changes in the ratio between SF6 and helium.

The measurement that is most commonly made from the more recent DTG-SBW tests is the slope of the SF6-Helium ratio between 60% and 90% or 65% to 95% of the exhaled tidal volume. This is usually termed the phase III slope and is most frequently designated the SIII_DTG. This value is usually normalized by multiplying it by tidal volume in order to make results from different breaths and different individuals comparable and this is designated as the SnIII_DTG. In the past however, the SIII_DTG has been taken from different volume spans such as 35% to 85% and although the overall conclusions are similar, the reported slopes may be different.

SIII_N2 and LCI appear to measure ventilation inhomogeneity at a global level whereas an elevated SIII_DTG is considered to show an increase in ventilation inhomogeneity at the acinar level. One point that is occasionally confusing is that strictly speaking the SIII_DTG is negative (i.e. sloping downwards) and so a lower slope usually means a steeper slope. This is in contrast to the positive phase III slope of the N2 washout where a lower slope usually refers to one that is flatter. Researchers however, often arbitrarily invert the SIII_DTG slope and report it as a positive value. This means its important to read carefully in order to be sure what changes are being discussed.

More than one study has showed that the SIII_DTG and SnIII_DTG results from children show a distinct difference between those with normal lungs and those with cystic fibrosis. One study, was able to show distinct improvements in SIII_DTG in cystic fibrosis patients following bronchodilator and chest physiotherapy and that these improvements were slightly more significant than the changes in FEV1. Another study indicated however, that SIII_DTG has a somewhat lower sensitivity and specificity than the LCI.

Studies in individuals with COPD also showed significant differences in SIII_DTG between between healthy subjects and those with COPD. Interestingly, one study that compared SIII_DTG with SIII_N2 showed that increased slopes from both correlated with decreases in DLCO, but that the SIII_N2 correlated with FEV1 whereas the SIII_DTG did not. In addition the SIII_N2 correlated with an elevated RV/TLC ratio (gas trapping) whereas the SIII_DTG correlated instead with an elevated TLC (hyperinflation).

Studies in individuals with asthma have also shown a correlation between FEV1 and the SIII_DTG. As importantly, they have shown that the terminal airways are a factor during bronchoconstriction in individuals with more severe asthma (or at least more hyperreactive airways).

One study in lung transplant patients where helium and SF6 washout curves and not the SF6-Helium ratio were analyzed showed that the helium phase III slope became steeper well in advance of the onset of bronchiolitis obliterans and better served as an early warning of this disorder than did spirometry.

The use of tidal breath washout in the SIII_DTG as opposed to the vital capacity washout used in the SIII_N2 are largely due to the observation that the phase III slope in a vital capacity maneuver is affected by large-scale and small-scale inhomogeneities that are at least partly dependent on gravity and airway closure. Research has shown that inhalations from FRC are less affected by gravity and other confounding effects, and maximizes the differences between convective and diffusive gas transport. One point that does not appear to have been addressed however, is that the ventilation that occurs during tidal breathing is not homogeneously distributed throughout the lung, even in healthy subjects. Admittedly, the SIII_N2 and the SIII_DTG both appear to partition individuals with airway disorders from healthy subjects equally well, and admittedly the type of ventilation inhomogeneity each measures is expected to be different, but unclear to me how well the SIII_DTG from a tidal breath extrapolates to the lung as a whole.

Although several studies have shown reasonably good inter-test and intra-test reproducibility at least one study showed that SIII_DTG was affected by expiratory flow rates and tidal volume. In particular the SIII_DTG became less steep when tidal volume and expiratory flow rates increased and this occurs even when the SIII_DTG is normalized for tidal volume. This effect has not been completely explained and may place a limit on the clinical significance of the SIII_DTG and the ability to assess changes over time and between individuals.

The ultrasonic mass molar gas analyzer is significantly less expensive than a mass spectrometer but its use still requires technical proficiency. Like the gas analyzers in a CPET system the mass molar gas analyzer, O2 and CO2 analyzers are time delayed relative to the flow signal at the patient’s mouth both due to gas transport and analyzer rise times careful attention must be made to synchronizing signals. One important point is that this type of analyzer does not measure the concentrations of individual gases, only the mass of the entire gas mixture. For this reason it can only measure the ratio of helium to SF6 and then only when careful attention is paid to subtracting the contribution CO2 makes to the mass of exhaled air. This also means that if the phase III slope of both helium and SF6 change by the same amount, then the slope of the SF6-Helium ratio will not change.

The DTG-SBW is an intriguing test that has the potential to provide more information about ventilation inhomgeneity than the LCI and SIII_N2. On the positive side it is easy to perform and only takes a short time to perform the test multiple times, both factors that makes it a test well suited to children. Despite the fact that the SIII_DTG appears to easily differentiate between healthy subjects and those with various airway disorders however, its clinical use is limited by the fact that there are at present no agreed-upon normal values. As importantly, it’s unclear what constitutes a clinically significant change and it’s unclear what changes in the SIII_DTG means in a clinical sense for the various airway disorders. I will also admit to having some reservations about the ultrasonic mass molar analyzer, partly for it’s need for a relatively high level of technical proficiency but also for its limitation in measuring only the SF6-He ratio. For these reasons the DTG-SBW is probably useful in a research setting, but its clinical utility, at least at this time, is probably quite limited.

Decades ago the study of obstructive lung diseases crystalized around the FEV1/FVC ratio and issues that had previously been murky were suddenly seen with a new clarity. Tests of ventilation inhomogeneity have the potential to provide an even more nuanced understanding of airway disorders but at the moment most of these tests measure effects that are at least one step removed from the actual physiological process with results expressed in ways that relate to the test (lung turnovers, phase III slopes) and not the underlying physiology. The use of two tracer gases with markedly different densities appears able to bring some needed clarity to ventilation inhomogeneity measurements but in order for the DTG-SBW to become a routine and useful measurement there needs to more work performed on correlating results with specific airway disorders and with specific changes in airway disorders. Just as importantly, a way will have to be found to express results in a way that makes the underlying physiology clear.

References:

Abbas C, Singer F, Yammine S, Casaulta C, Latzin P. Treatment response of airway clearance assessed by single-breath washout in children with cystic fibrosis. J Cystic Fibrosis 2013; 12: 567-574.

Boeck L, Gensmer A, Nyilas S, Stieltjes B, Re TJ, Tamm M, Latzin P, Stolz D. Single-breath washout tests to assess small airway disease in COPD. CHEST 2016, doi: 10.1016/j.chest.2016.05.019.

Estenne M, Van Muylem A, Knoop C, Antoine M. Detection of obliterative bronchiolitis after lung transplantation by indexes of ventilation distribution. Am J Respir Crit Care Med 2000; 162: 1047-1051.

Gustafsson PM, Ljungberg HK, Kjellman B. Peripheral airway involvement in asthma assessed by single-breath SF6 and He washout. Eur Respir J 2003; 21: 1033-1039.

Huseman K, Berg, Engel J, Pot J, Joppek C, Tao Z, Singer F, Schultz H, Kohlhaufl M. Double tracer single-breath washout: Reproducibility in healthy subjects and COPD. Eur Respir J 204; 44: 1210-1222.

Olfert IM, Prisk GK. Effect of 60º head-down tild on peripheral gas mixing in the human lung. J Appl Physiol 2004; 97: 827-834.

Osborne S. Airway resistance and airflow through the tracheobronchial tree. www.SallyOsborne.com.

Paiva M, Engel LA. Theoretical studies of gas mixing and ventilation distribution in the lung. Physiological Reviews 1987; 67(3): 750-796.

Robinson PD, et al. ERS/ATS Consensus Statement for inert gas washout using multiple- and single-breath tests. Eur Respir J 2013; 41: 507-522.

Schwartzstein RJ, Parker MJ. Respiratory physiology. A clinical approach. Published by Lipincott, Williams and Wilkins, 2006.

Singer F, Stern G, Thamrin C, Fuchs O, Riedel T, Gustafsson P, Frey U, Latzin P. Tidal volume single breath washout of two tracer gases – A practical and promising lung function test. PLOS One 2011; 6(3): e17588.

Singer F, Stern G, Thamrin C, Abbas C, Casualta C, Frey U, Latzin P. A new double-tracer gas single-breath washout to assess early cystic fibrosis lung disease. Eur Respir J 2013; 41: 339-345.

Van Muylem A, De Vuyst P, Yernault JC, Paiva M. Inert gas single-breath washout and structural alteration of respiratory bronchioles. Am Rev Respir Dis 1992; 146: 1167-1172.

Van Muylem A, Paiva M, Estenne M. Involvement of peripheral airways during methacholine-induced bronchoconstriction after lung transplantation. Am J Respir Crit Care Med 2001; 164: 1200-1203.

Weibel ER. “Geometry and dimensions of airways of conductive and transitory zones.” In Morphometry of the human lung. Published by Springer, 1963: 110-135.

West J. Respiratory physiology – The Essentials 7^th edition, published by Lipincott, 2004.

Wilschut FA, Van der Grinten CPM, Lamers RJS, Wouters EFM, Luijendijk SCM. Intrapulmonary gas mixing and the sloping alveolar plateau in COPD patients with macroscopic emphysema. Eur Respir J 1999; 14: 166-171.

PFT Blog by Richard Johnston is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License

When I first started performing CPETs in the 1970’s a patient’s exhaled gas was collected at intervals during the test in Douglas bags and I had a worksheet that I’d use to record the patient’s respiratory rate, heart rate and SaO2. After the test was over I’d analyze the gas concentrations with a mass spectrometer and the gas volumes with a 300 liter Tissot spirometer and then use the results from these to hand calculate VO2, VCO2, Rq, tidal volume and minute volume. These results were then passed on to the lab’s medical director who’d use them when dictating a report.

Around 1990 the PFT lab I was in at the time acquired a metabolic cart for CPET testing. This both decreased the amount of work I had to do to perform a CPET and significantly increased the amount of information we got from a test. The reporting software that came with the metabolic cart however, was very simplistic and neither the lab’s medical director or I felt it met our needs so I created a word processing template, manually transcribed the results from the CPET system printouts and used it to report results.

Twenty five years and 3 metabolic carts later I’m still using a word processing template to report CPET results.

Why?

Well, first the reporting software is still simplistic and using it we still can’t get a report that we think meets our needs (and it’s also not easy to create and modify reports which is a chronic complaint I have about all PFT lab software I’ve ever worked with). Second, there are some values that we think are important that the CPET system’s reporting software does not calculate and there is no easy way to get it on a report as part of the tabular results. Finally, and most importantly, I need to check the results for accuracy.

You’d think that after all these years that you wouldn’t need to check PFT and CPET reports for mathematical errors but unfortunately that’s not true. For example, these results are taken from a recent CPET:

Time:	VO2 (LPM):	VCO2 (LPM):	Reported Rq:	“Real” Rq:
Baseline:	0.296	0.220	0.74	0.74
00:30	0.302	0.214	0.77	0.71
01:00	0.363	0.277	0.77	0.76
01:30	0.395	0.306	0.78	0.77
02:00	0.424	0.353	0.99	0.83
02:30	0.459	0.403	0.92	0.88
03:00	0.675	0.594	0.89	0.88
03:30	0.618	0.584	0.94	0.94
04:00	0.836	0.822	1.00	0.98

Time:	Ve(LPM):	VO2 (LPM):	Ve/VCO2:	“Real” Ve/VO2:
Baseline:	11.7	0.296	39	40
00:30	9.5	0.302	33	31
01:00	10.8	0.363	30	30
01:30	11.7	0.395	30	30
02:00	13.0	0.424	38	31
02:30	14.2	0.459	33	31
03:00	20.4	0.675	31	30
03:30	19.4	0.618	31	31
04:00	26.6	0.836	33	32

Time:	Ti:	Te:	Ti+Te:	RR:	“Real” RR:
Baseline:	2.49	3.01	5.50	11	11
00:30	2.95	3.75	6.70	11	9
01:00	1.33	2.66	3.99	15	15
01:30	1.19	2.29	3.48	18	17
02:00	1.15	1.67	2.82	23	21
02:30	1.14	1.79	2.93	20	20
03:00	1.09	1.64	2.73	23	22
03:30	0.96	1.56	2.52	24	24
04:00	0.85	1.25	2.10	31	29

To be honest, the discrepancies between the results reported by the CPET system software and the re-calculated results are often minor and are possibly explained by hidden digits and rounding errors.

The hidden digit problem occurs when the results are stored with more numbers after the decimal point than are reported. For example, a VO2 value could be reported as 0.296 but could instead be stored as 0.296893. In this case, the software truncated the stored value when reporting it rather than rounding it up to 0.297. When the calculations are performed, they are performed on the entire number rather than the truncated number. Depending on whether the results from these calculations are rounded or truncated there can be a discrepancy when compared to the calculations performed on the reported values.

But this only explains some of the discrepancies. In particular, please note the Rq and Ve/VO2 values at the 2 minute mark. The Rq is reported as 0.99 but calculates from the reported VO2 and VCO2 as 0.83. The Ve/VO2 is reported as 38 but calculates from the reported VO2 and Ve as 31. Unless there is a bug in the software (not impossible) this is probably due to the algorithm that’s used to average results over the 30 second period.

Specifically, over any given time period (20 seconds, 30 seconds, 60 seconds) there usually isn’t an evenly matched number of inspirations and expirations. Ideally a time period would start and end at end-exhalation but since timing is arbitrary it could just as easily start mid-exhalation or mid-inspiration, and the same applies to the end of the timing period. A software algorithm that averages results therefore has to be able to extract and extrapolate data from partial breaths and I don’t know what solution the software engineers have applied to this problem.

In this case the averaging algorithm could have accumulated the calculated Rq and Ve/VO2 from each breath and then averaged them rather than calculating the Rq and Ve/VO2 from the averaged VO2, VCO2 and Ve. But this leads to another conundrum and that is how are the VO2, VCO2 and Ve averaged and are they any more (or any less) accurate than calculated values like the Rq, Ve/VO2 etc?

One point in favor of saying that VO2, VCO2 and Ve are at least somewhat more accurate is that the calculated Rq and Ve/VO2 at the 2 minute mark are outliers and don’t fit with the flow of data. You usually don’t expect that Rq or Ve/VO2 to suddenly change and then revert in the middle of a CPET. It’s not impossible, of course, but you’d also expect corresponding changes in VO2, VCO2 and Ve. This leads me to suspect that the problem lies primarily with the averaging algorithm for calculated values.

None of the companies that manufacture metabolic carts have ever published the algorithms they use to average results, however. These companies often compare their results to those obtained from the “gold standard” (such as it is) of Douglas bags, Tissot spirometers and stand-alone gas analyzers. Even this has its problems however, since the exhaled air collected in Douglas bags are almost always collected from a number of complete exhalations that occur over a period of time that is usually either less or more than that used by a metabolic cart. When this happens, the Douglas bag results that, for example, come from a 28 second or 35 second period need to be extrapolated in order to match a 30 second period from a metabolic cart.

The problems with both rounding and averaging errors is that I use these tabular results both to determine the anaerobic threshold and to calculate the Ve-VO2 and Ve-VCO2 slopes. At the moment I have to assume that the core values such as VO2 and Ve are reasonably correct (particularly since I have no way to verify them) and that derived calculations such as Rq and Ve/VO2 need to be checked and when necessary, re-calculated. Since there is no easy way to download CPET results from our CPET system in a format that can be inserted into a spreadsheet this means I do a lot of manual transcribing (so it’s a good thing that I type fast).

Most labs that I know that are performing CPETs use the reporting software that comes along with their metabolic cart. Aesthetics aside, this can be a problem since some values that I consider critical to interpreting a CPET are often not included. More problematic is that most labs also accept that all of the reported values are accurate and this unfortunately isn’t necessarily true (and I’m not even talking about things like determining the anaerobic threshold). To some extent this is routinely excused or overlooked since cardiopulmonary exercise testing is considered to be an order of magnitude more complicated than regular PFT testing but really, it’s not rocket science. You should never assume that the math on a CPET (or PFT) report is correct until you verify it and all I needed to verify these errors was a pocket calculator.

It would be helpful if the companies that manufacture CPET systems were more forthcoming about their averaging (and other) algorithms but realistically, even with the Douglas bag approach there are inherent limitations to accuracy. The best we can do is to be aware of the more obvious errors and to find work-arounds for them.

And yes, creating a CPET report from scratch takes more time and effort than using a canned report but like baking and cooking from scratch, I think the results are worth it.

PFT Blog by Richard Johnston is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License

Recently a reader asked me a question about the FVC/DLCO ratio. To be honest I’d never heard of this ratio before which got me intrigued so I spent some time researching it. What I found leaves me concerned that a lack of understanding about reference equations may invalidate several dozen interrelated studies published over the last two decades.

Strictly speaking the FVC/DLCO ratio is the %predicted FVC/%predicted DLCO ratio (which is usually abbreviated FVC%/DLCO%) and it appears to be used exclusively by specialists involved in the treatment of systemic sclerosis and related disorders. Specifically, the ratio is used to determine whether or not a patient has pulmonary hypertension. The basic idea is that (quoting from a letter to the editor):

“As we know, in ILD both FVC and DLCO fall and their fall is proportionate, whereas in pulmonary arterial hypertension DLCO falls significantly and disproportionately to FVC.”

A variety of algorithms using the FVC%/DLCO% have been devised. Inclusion factors are usually an FVC < 70% of predicted and a DLCO (corrected for hemoglobin) < 60% of predicted. A number of different threshold values for FVC%/DLCO% have been published ranging from 1.4 to 2.2 with the differences appearing to be dependent on study population characteristics and the type of statistical analysis performed. It is thought that individuals meeting the inclusion factors with an FVC%/DLCO% ratio above the threshold most probably have pulmonary hypertension.

I could quibble about the notion that FVC and DLCO always decrease proportionally in ILD, but my biggest concern with this approach is that after reviewing 19 articles that discussed the FVC%/DLCO% ratio almost none of them indicated which FVC or DLCO reference equations were being used.

The very first study that made the observation about the FVC%/DLCO% ratio (Steen et al, 1992) did indicate that it used Morris’ reference equations for FVC and Burrows’ for DLCO; another article (Zisman et al) indicated that they used Crapo for FVC predicteds and Miller for DLCO predicteds; and one other article (Koh et al) referenced both Morris and Knudsen for FVC predicteds without indicating which was used and without indicating which DLCO reference equation was used.

Reference:	FVC Reference Equation:	DLCO Reference Equation:
[A]	No	No
[B]	No	No
[C]	No	No
[D]	No	No
[E]	No	No
[F]	No	No
[G]	No	No
[H]	No	No
[I]	No	No
[J]	No	No
[K]	Yes (Morris and Knudsen)	No
[L]	No	No
[M]	Yes (Morris)	Yes (Burrows)
[N]	No	No
[O]	No	No
[P]	No	No
[Q]	No	No
[R]	No	No
[S]	Yes (Crapo)	Yes (Miller)

That’s it. None of the other articles indicated which reference equations for FVC and DLCO were used. Unfortunately this leaves me with a strong impression that none of the authors (or journal reviewers for that matter) realize there are number of different (and conflicting) reference equations for both FVC and DLCO, and that this fact has significant implications for interpreting PFT results. One particularly telling comment was that the

“reference, actual, and percent predicted values … were extracted from the [Pulmonary Function] reports…”.

Which means to me that pulmonary function labs were not directly involved in the research and that the FVC and DLCO results came from routine testing.

This apparent ignorance is a concern since a number of the studies performed advanced statistical analyses determining the specificity and sensitivity of a variety of parameters, and in many cases gave very precise thresholds for the FVC%/DLCO% ratio. Any precision, unfortunately, is largely an artifact of which reference FVC and DLCO equations were actually used and I expect that much of the difference in published threshold values for the FVC%/DLCO% ratio are because the different studies used different reference equations (apparently without knowing they did).

As an example, for a 50 year old, 175 cm, Caucasian male, the average predicted FVC is 4.79 L (with a range from the reference equations I have on hand of 4.50 to 5.06) and the average predicted DLCO is 32.2 (ditto, ranging from 24.4 to 36.09). If you took an FVC that was 70% of the average predicted (3.35 L) and a DLCO that was 40% of the average predicted (12.88 ml/min/mmHg), depending on which reference equations you used the FVC%/DLCO% ratio could be anywhere from 1.25 to 2.09.

This is not the first time I’ve run across studies where the results or outcomes were based on the percent predicted of spirometry, lung volume or DLCO results but the reference equations were not identified. These tend to be relatively isolated examples and when I run across them I’ve usually thought the authors (and reviewers) were just being sloppy. Given the effect that reference equations have on results it has also meant I’ve taken any conclusions with a grain of salt.

This is the first time however, that I’ve run across an entire field of study that appears to be completely ignorant of the effect that reference equations have on PFT results. Admittedly, pulmonary function testing is somewhat unique when compared to other areas of medicine in that there are no universal reference values (kudos to the GLI people for trying but let’s be honest, since they’re still foundering on the reefs of ethnicity we’re not anywhere near there yet). And admittedly there is still a certain level of confusion among both pulmonary physicians and technologists on the subject of reference equations (but at least we know the problem exists). Even so, since the whole point of the FVC%/DLCO% ratio is to differentiate between ILD and ILD with pulmonary hypertension, weren’t any of these studies reviewed by a pulmonologist?

The basic concepts behind the FVC%/DLCO% ratio; that lung volumes tend to be reduced when ILD is present, and that DLCO tends to be reduced out of proportion to any decrease in lung volume when pulmonary hypertension is also present; fit well with existing knowledge. In addition percent predicted values are a way to “normalize” PFT results for a study population composed of different genders, ages and heights. The problem is that once you select a specific set of reference equations, any threshold values or other conclusions that you develop are entirely dependent on that selection and cannot be universally applied.

I try to keep up with the field of pulmonary function testing and at least skim the table of contents of a dozen or so medical journals every month looking for articles but I had never heard of the FVC%/DLCO% ratio before last week. So, an interesting thought is to wonder who else is using PFT results without understanding the problems with reference equations?

References.

[A] Beall AD, Nietert PJ, Taylor MH, Mitchell HC, Shaftman SR, Silver RM, Smith EA, Bolster MB. Ethnic disparities among patients with pulmonary hypertension associated with systemic sclerosis. J Rheumatology 2007; 34: 1277-1282.

[B] Coghlan JG et al. Evidence-based detection of pulmonary arterial hypertension in systemic sclerosis: The DETECT Study. Ann Rheum Dis 2014; 73: 1340-1349.

[C] Corte TJ, Wort SJ, Wells AU. Pulmonary hypertension in idiopathic pulmonary fibrosis: a review. Sarcoidosis Vasculitis and Diffuse Lung Diseases 2009; 26: 7-19.

[D] Gladue H et al. Combination of echocardiographic and pulmonary function test measures improves sensitivity for diagnosis of systemic sclerosis-associated pulmonary arterial hypertension: analysis of 2 cohorts. J Rheumatology 2013; 40: 1706-1711.

[E] Goldberg A. Pulmonary arterial hypertension in connective tissue disorders. Cardiology In Review 2010; 18: 85-88.

Gupta R. Letter to the editor: Pulmonary function test as screening test for pulmonary artery hypertension in scleroderma patients. Rheum 2004; 43: 1315

[F] Hao Y et al. Comparison of the predictive accuracy of three screening models for pulmonary arterial hypertension in systemic sclerosis. Arthritis Research & Therapy 2015; 17: 7

[G] Hervier B et al. Pulmonary hypertension in antisythetase syndrome: prevalence, aetiology and survival. Eur Respir Dis 2013; 42: 1271-1782

[H] Hsu VM, Moreyra AE, Wilson AC, Shinnar M, Shindler DM, Wilson JE, Desai A, Seibold JR. Assessment of pulmonary arterial hypertension in patients with systemic sclerosis: comparison of noninvasive tests with results of right-heart catheterization. J Rheumatology 2008; 35: 458-465.

[I] Hudson M et al. Comparison of different measures of diffusing capacity for carbon monoxide (DLCO) in systemic sclerosis. Clin Rheumatology 2013; 32: 1467-1474.

[J] Khanna D, Denton CP. Evidence-based management of rapidly progressing systemic sclerosis. Best Pract Res Clin Rheumatol 2010; 24(3): 387-400.

[K] Koh ET, Gladman DD, Abu-Shakra M. Pulmonary hypertension in systemic sclerosis: an analysis of 17 patients. Br J Rheum 1996; 35: 989-993.

[L] Risbano MG et al. Altered immune phenotype in peripheral blood cells of patients with scleroderma-associated pulmonary hypertension. Clin Trans Sci 2010; 3: 210-218.

[M] Steen VD, Graham G, Conte C, Owns G, Medsger TA. Isolated diffusing capacity reduction in systemic sclerosis. Arthritis Rheum 1992; 35(7): 765-770.

[N] Steen VD. Autoantibodies in systemic sclerosis. Semin Arthritis Rheum 2005; 35: 35-42.

[O] Steen VD, Lucas M, Fertig N, Medsger TA. Pulmonary arterial hypertension and severe pulmonary fibrosis is systemic sclerosis with a nuclear antibody. J Rheumatology. 2007; 34: 2230-2235.

[P] Sivia N et al. Relevance of partitioning DLCO to detect pulmonary hypertension in systemic sclerosis. Plos One 2013; 8(10): e78001.

[Q] Thakkar V et al. N-terminal pro-brain natriuretic peptide in a novel screening algorithm for pulmonary arterial hypertension in systemic sclerosis: a case control study. Arthritis Research & Therapy 2012; 14: R143.

[R] Walkey AJ, Ieong M, Alikhan M, Farber HW. Cardiopulmonary exercise testing with right-heart catheterization in patients with systemic sclerosis. J Rheumatology 2010; 37: 1871-1877.

[S] Zisman DA, Ross DJ, Belperio JA, Saggar R, Lynch JP, Ardehali A, Karlamangla AS. Prediction of pulmonary hypertension in pulmonary fibrosis. Respiratory Medicine 2007; 10: 2153-2159.

PFT Blog by Richard Johnston is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License

I was reviewing a spirometry report last week and when I went to compare the results with the patient’s last visit I noticed that the FVC and FEV1 hadn’t changed significantly. However, the previous results were from 2009 and when the percent predicted is considered there may have been a significant improvement.

2009	Observed:	%Predicted:
FVC:	2.58	87%
FEV1:	1.60	72%
FEV1/FVC:	62	82%

2016	Observed:	%Predicted:
FVC:	2.82	104%
FEV1:	1.65	82%
FEV1/FVC:	59	79%

The answer to whether or not there was an improvement would appear to depend on what changes you’d normally expect to see in the FVC and FEV1 over a time span of 7 years. The FVC and FEV1 normally peaks around age 20 to 25 and then declines thereafter.

[more]

The rate at which FVC and FEV1 declines shown by reference equations comes from cross-sectional studies.

[

Longitudinal studies have often indicated that the rates of decline are different, and often lower than, those from cross-sectional studies. The reasons for these differences has been variously attributed to the learning effect, cohort effects (i.e. dependence on the year of birth rather than aging per se), and differences in study populations. In addition more than one longitudinal study has noted that returning participants are usually better educated with better health. Perhaps just as importantly returning participants are alive and sufficiently capable, mentally and physically, of performing acceptable quality spirometry.

One relatively early finding from longitudinal studies was that the rate of decline accelerates as age increases and this is not evident in most cross-sectional studies. This is at least partially due to the type of statistical analysis that is performed. Linear regression is the most commonly used analytic tool but when the same population is studied with multiple step-wise regression analysis this pattern becomes more evident.

This also means that the rate of decline in FVC and FEV1 from any given longitudinal study are dependent on the demographics of the study population and what type of analysis is performed. For this reason, the reported rates of decline in FEV1 in non-smokers ranges from -10 ml/year to -56 ml/year. This also means that the rates of decline between cross-sectional and longitudinal studies although differing when compared individually are overall fairly similar.

One important point, and this applies to both cross-sectional and longitudinal studies, is that as FEV1 declines normally, the annual decrease when measured as a percent change increases.

Multiple longitudinal studies have shown (not surprisingly) that the rate of decline in both FVC and FEV1 is higher (usually at least double the rate) for smokers than non-smokers or ex-smokers. They have also shown that the rate of decline is also higher for individuals with emphysema, COPD, asthma, cardiovascular disease, cardiac disease or with an increasing BMI. At least one study showed that the rate of decline was greater for Caucasians than for blacks and another showed that rate of decline was greater for taller individuals than for shorter (which is perhaps why a number of studies report the rate of decline as ml/m³). Interestingly, although the use of bronchodilators (either short-acting or long-acting) does not appear to affect the rate of decline at least one study showed that the use of statins decreased the rate of decline among non-smokers, smokers and quitters.

Since there the rate of decline in FVC and FEV1 would appear to be an important factor in a given individual’s clinical management, why isn’t it routinely measured? The primary reason is that more than one study has shown that the normal level of year-to-year variability in FEV1, even in healthy subjects, is many times higher than the rate of decline. For this reason testing must be performed regularly over long periods of time before a rate of decline can be determined with any reliability. In addition, patients that return to the PFT lab are often seen intermittently and often only during periods of ill health. This adds a level of variability to spirometry results that make it far more difficult to determine a meaningful rate of decline.

Getting back to the original question, is the lack of change actually an improvement? In one sense, the patient has probably had some clinical improvement, but when considering the amount of routine visit-to-visit variability in test results, probably not enough to be remarked about, at least not over a 7 year period. If it was considered over a longer time period however, there has to be a point at which no change is both statistically and clinically significant. At the present time both the ATS/ERS and the ACOEM state that when considering changes in FEV1 over the course of 1 year, that a 15% decrease is clinically important. As a starting point therefore, 15% seems far enough above the “noise” level that when comparing results separated by enough time that the FVC or FEV1 should have decreased by 15% but instead remained the same, then there has probably been significant improvement.

References:

Alexeeff SE, Litonjua AA, Sparrow D, Vokonas PS, Schwartz. Statin use reduced decline in lung function. VA Normative Aging Study. Amer J Respir Crit Care Med 2007; 176: 742-747.

Anthonisen NR, Connett JE, Murray RP. Smoking and lung function of Lung Health Study participants after 11 years. Amer J Respir Crit Care Med 2002; 166(5): 675-679.

Burrows B, Lebowitz MD, Camilli AE, Knudsen RJ. Longitudinal changes in forced expiratory volume in one second in adults. Methodological considerations and findings. Amer Rev Respir Dis 1986; 133: 974-980.

Carey IM, Cook DG, Strachan DP. The effects of adiposity and weight change on forced expiratory volume decline in a longitudinal study of adults. Int J Obesity 1999; 23: 979-985.

Cibella F, Cuttitta G, Bellia V, Bucchieri S, D’Anna S, Guerrera D, Bonsignore G. Lung function decline in bronchial asthma. Chest 2002; 122(6): 1944-1948.

Griffith KA, Sherrill DL, Siegel EM, Manolio TA, Bonekat HA, Enright PL. Predictors of loss of lung function in the elderly. The cardiovascular health study. Amer J Respir Crit Care Med 2001; 163(1): 61-68.

Hnizdo E, Yu L, Freyder L, Attfield M, Lefante J, Glindmeyer HW. The precision of longitudinal lung function measurements: monitoring and interpretation. Occup Environ Med 2005; 62: 695-701.

Hnizdo E, Sircar K, Yan T, Harber P, Fleming J, Glindermeyer HW. Limits of longitudinal decline for the interpretation of annual changes in FEV1 in individuals. Occup Environ Med 2007; 64: 701-707.

Kerstjens HAM, Rijcken B, Schouten JP, Postma DS. Decline of FEV1 by age and smoking status: facts, figures and fallacies. Thorax 1997; 52: 820-827.

Peat JK, Woolcock AJ, Cullen K. Decline of lung function and development of chronic airflow obstruction: a longitudinal study of non-smokers and smokers in Busselton, Western Australia. Thorax 1990; 45: 32-37

Ryan G, Knuiman MW, Divitini ML, James A, Musk AW, Bartholomew HC. Decline in lung function and mortality: The Busselton Health Study. J Epidemiol Community Health 1999; 53: 230-234

PFT Blog by Richard Johnston is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License

A couple of days ago my medical director and I had a short discussion about teaching pulmonary fellows to read PFTs and agreed that in order to be good at interpreting PFTs it isn’t the basic algorithms that are hard, it’s gaining an understanding of test quality and testing problems. My medical director then suggested this topic. At first I wasn’t sure I could find 10 errors but after spending a couple hours digging through my teaching files I managed to come up with just a few more than that. So strictly speaking it’s not a top 10 list but I kept the title because I liked it.

Spirometry errors and mistakes seem to fall into four categories: demographics, reference equations, testing and interpretation.

Demographics:

Normal values are based on an individual’s age, height and gender. When this information is entered incorrectly the normal reference values will also be incorrect. These errors often go uncaught because whoever reviews and interprets reports usually isn’t the same person who sees the patient and performs the tests. This type of error often doesn’t get corrected until the results are uploaded into a hospital information system or the patient returns for a second (or third or fourth) visit.

1. Wrong gender.

Pulmonary function reference equations are gender specific and for individuals with the same age and height, men will have a larger FVC and FEV1 than women do. When a patient’s demographics information is manually entered into a PFT system it’s always possible for somebody to enter the wrong gender. When this happens the predicted values will be either over- or under-estimated. This happens in my lab at least a half a dozen times a year and it’s why when I review reports I try to check the patient’s gender right after reading their name.

This is also a problem area for individuals who have gone through gender reassignment (transsexuals). An individual’s physiologic/developmental gender needs to be used to generate predicted values but this may be at odds with their gender recorded in a hospital’s information system. Some PFT lab systems populate their demographics information from their hospital’s information system when an order is received and it may or may not be possible to alter gender once this has happened. In other cases, an individual’s demographics may be cross-referenced when PFT results are uploaded into hospital information system and may throw an error if the wrong gender is present.

2. Wrong height

All lung volumes and capacities scale with height. Like any other manual entry height can be mis-entered and the most common error I’ve seen is for somebody to enter 60 inches when they meant 6 feet 0 inches.

Height can also be mis-measured if the patient isn’t asked to remove their shoes or to stand straight, or if the patient is asked for their height and it isn’t even measured. An error of an inch or two probably won’t make a big difference in a patient’s predicted values (particularly given the discrepancies between different reference equations) but for somebody who’s on the edge of normal and abnormal it can make a significant difference in how a report is interpreted.

3. Wrong age

Volumes and flow rates tend to increase until about age 20 or so and then decline thereafter. A patient’s date of birth is usually used to determine their current age and this, of course, is another opportunity for a manual entry error. The most common errors I’ve seen are for today’s date to be entered instead of the patient’s DOB or for the wrong decade to be entered.

This type of error used to be caught when we uploaded patient results into our hospital’s information system but for both good and bad reasons this no longer happens. Patients are usually asked to confirm their date of birth when they check in but this is confirmed through from the hospital’s information system, not the lab’s software, and there is no process available to cross-check these other than manually. This means that unless staff are paying attention or unless there is a significant error in the patient’s age, this type of error will not be caught.

Reference equations:

The ATS/ERS guidelines say that every pulmonary function lab is supposed to select the appropriate reference equations for the population it serves. This makes eminent sense but unfortunately there aren’t any recommendations on how this is actually supposed to be done. This is also complicated by the fact that reference equations themselves are limited by the size and limited range of ethnicities, ages and heights in the population they study, and the statistics used to analyze them.

4. Limits of reference equations for ethnicity

When a population shares an environment, diet and a large number of genes it is likely that their lung function will be similar as well. This is the basis for ethnicity-based reference equations but environment, diet and genes are rapidly changing and it is not as clear as it once was what ethnicity means. This can make selecting the appropriate ethnicity-based reference equations difficult. Even when an individual’s ethnicity is relatively evident, a PFT lab’s software may not have an appropriate reference equation available or it may make selecting the appropriate reference equation difficult.

As importantly, many test systems just subtract a specific percentage for Blacks or Asians from a reference equation for Caucasians but this makes their results dependent on a reference equation intended for a population they are not part of. This was a common practice up until about a dozen years ago when the ATS/ERS standards recommended the use of ethnicity-based reference equations instead.

As an example of this problem, the following individual was born and raised in India and referred for screening spirometry.

Asian Indian	Observed:	%Predicted:	Predicted:
FVC:	3.40	73%	4.67
FEV1:	2.78	77%	3.58
FEV1/FVC:	82	107%	77

My lab’s software doesn’t allow us to enter Asian Indian as an ethnicity, nor does it have any Asian Indian reference equations. Depending on which reference equations you compare, Asian Indian FVC’s and FEV1’s are approximately 20% less than those for Caucasians. That means these results are probably not abnormal, particularly since the individual was asymptomatic, and continued to be asymptomatic even when followed up over several years.

5. Limitations of reference equations for height

The limits of the statistics used to generate reference equations often becomes quite clear when an individual is very short or very tall. Unfortunately there is no clear definition of what the normal limits for height are and most of the studies that the reference equations come from do not provide any statistics on the height range of their study population. As importantly there are almost no studies whatsoever on individuals at the extreme ends of the human height range.

The following individual is 48” tall and the first set of results were generated by the Morris reference equations:

48” Morris	Observed:	%Predicted:	Predicted:
FVC:	1.73	102%	1.69
FEV1:	1.36	82%	1.65
FEV1/FVC:	79	81	97%

Note that the predicted FVC and FEV1 are almost identical. When the predicted are re-calculated using the NHANESIII reference equations the results look more normal:

48” NHANESIII	Observed:	%Predicted:	Predicted:
FVC:	1.73	86%	2.01
FEV1:	1.36	82%	1.66
FEV1/FVC:	79	101%	78

But realistically, there’s no way to determine whether they are normal or not since the this person is over a foot shorter than anybody tested in the NHANESIII population. To some extent there is a similar problem with the very tall as well. The following individual was 84” high which is a half a foot higher than anybody in NHANESIII study:

84” NHANESIII	Observed:	%Predicted:	Predicted:
FVC:	8.25	103%	8.04
FEV1:	5.40	85%	6.38
FEV1/FVC:	65	80%	81

The percent predicted results indicate this individual likely has airway obstruction but the NHANESIII FEV1/FVC ratio is calculated without the use of height. Almost all spirometry reference equations (and this includes the NHANESIII FVC and FEV1) show that the FEV1/FVC ratio normally decreases with increasing height. So is this really airway obstruction or an artifact of the reference equations?

6. Limitations of reference equations for age

Reference equations are limited at the extremes of age. For the very young there can be different reference equations for infants, children and adolescents. The dividing line between these categories can be unclear particularly since developmental age is not the same thing as chronological age.

For the elderly, most study populations usually have a limited number of subjects over the age of 70, rarely over the age of 80 and no study has ever had subjects over the age of 90. These results came from an individual that was 97 years old:

97 y/o	Observed:	%Predicted:	Predicted:
FVC:	1.17	108%	1.08
FEV1:	0.95	126%	0.75
FEV1/FVC:	81	116%	70

This looks normal, but there is no way to be sure since the slope at which FVC and FEV1 decline with age is determined primarily by a population that is at least 20 years younger. This is complicated by the fact that the NHANESIII reference equations show that the decline with age accelerates with increasing age whereas many other reference equations show a linear decline with age and it’s not clear which of these observations is correct.

Testing:

7. Back extrapolation

Nobody is able to start exhaling and to reach their maximum expiratory flow instantaneously. Back extrapolation is a process that uses the slope of the highest expiratory flow to determine a standardized beginning of a forced vital capacity effort. When the back extrapolated volume is high the beginning of the spirometry effort becomes more indeterminate and the FEV1 is more likely to be overestimated.

	Red:	Blue:
FVC:	2.74	2.67
FEV1:	0.61	0.95
FEV1/FVC:	22	36

The ATS/ERS standard for spirometry states the extrapolated volume “must be <5% or the FVC or 0.150 L, whichever is greater.” Testing software doesn’t usually accept results with too large of a back extrapolation but in this case the technician overrode the computer because the effort had a “better” FEV1. Even so, some patients are unable to perform spirometry without a slow start and a large amount of back extrapolation on every effort and when this happens the computer will likely report the highest FEV1 regardless of whether it had the smallest or largest amount of back extrapolation.

8. Short effort

The FVC is supposed to be the maximal amount of air an individual can exhale after a maximal inhalation but some patients stop early because of glottal closure or because they feel they’ve exhaled enough. When this happens the FVC will be underestimated and the FEV1/FVC ratio overestimated. This may not be evident when the numerical results are reviewed:

	Observed:	%Predicted:	Predicted:
FVC:	2.99	79%	3.79
FEV1:	2.51	88%	2.85
FEV1/FVC:	84	111%	76
Time:	6.9 sec

But becomes more noticeable when the volume-time curve is viewed:

The reported expiratory time is often incorrect and this is because computer software usually doesn’t use an expiratory flow of zero as an indication that exhalation has ended but instead uses an inspiratory effort (or a manual override from the technician) as the marker for the end of exhalation.

9. FEV1 underestimated due to an expiratory pause

The ATS/ERS standards state that an FVC effort should be free of coughs and artifacts that will affect the measurement of FEV1. An early expiratory pause of any kind most often causes the FEV1 and the FEV1/FVC ratio to be underestimated. Software is generally poor at recognizing artifacts of this kind however, and it is only by inspecting the graphics from a spirometry effort that they may be noticeable at all.

When expiratory flow stops, a flow-volume loop will show a notch:

But a flow-volume loop does not contain time, and a volume-time curve actually gives a better estimate of the effect an expiratory pause has on FEV1:

10. FVC underestimated from a mid-expiratory inspiration

A cough during exhalation doesn’t necessarily just pause expiratory flow, it may also cause a small inhalation to occur. Although the ATS/ERS standards give criteria for determining an adequate expiratory flow rate at the end of a test, they don’t address what level of inspiratory flow should terminate an expiratory effort and this issue is left to each manufacturer to decide for themselves. This example actually reported an expiratory time of greater than 6 seconds even though it stopped measuring the exhaled volume at the first inhalation:

And although it is noticeable on the flow-volume loop, once again it is more evident on the volume-time curve

11. Inadequate inspiration

Although there are a number of criteria for judging the end of exhalation it is actually very difficult to determine whether or not a patient has taken a maximal inhalation. The ATS/ERS standards recommend that a forced expiratory effort be followed by a maximal inhalation. This is so that the maximum inspiratory flows can be measured but occasionally the FIVC is greater than the FVC.

When this happens the FVC is certainly underestimated and since the expiratory effort didn’t start from a maximal inhalation, the FEV1 is likely underestimated as well. Sometimes however, the indications that the patient didn’t take a full inhalation are more subtle.

In this instance the inspiratory capacity was almost nonexistent. Inspiratory capacity can be reduced in severe COPD due to expiratory flow limitation and gas trapping but in this case the tidal loop doesn’t show this and the small IC is more likely due to a submaximal effort.

12. FVC underestimated from a Patient leak

The mouthpieces that are used for spirometry are frequently circular and some patients may have difficulty maintaining a seal around them. When they leak, they may leak more during exhalation than during inhalation, or vice versa. When this happens, the tidal loop prior to the start of the FVC maneuver may drift:

A patient may be able to maintain a tight seal at the beginning of the test, but may get distracted and loosen their lips during the test itself. When this happens they leak during exhalation and their inspired volume will be greater than their exhaled volume:

In either case the FVC will usually be underestimated and the FEV1/FVC ratio will likely be overestimated.

13. Zero offset error

This is a hardware error that isn’t terribly common any more, but it can still happen. Spirometry is often performed using flow sensors and the signal from them is analog. The zero level from analog circuitry tends to drift but most test systems usually re-zero this before each test. Sometimes this process goes wrong. When it does, it can add (or subtract) an extra amount to the flow signal throughout exhalation:

Occasionally, it may just add an offset to be beginning of the exhalation:

In these examples the FVC would be overestimated and the FEV1/FVC ratio underestimated but the zero offset can also work in the opposite direction.

Interpretation:

The ATS/ERS guidelines for interpreting pulmonary function results are fairly straightforward, but they are also necessarily simplistic and for this reason there are some deficiencies. In addition, which values should be used to interpret spirometry have changed over time and not everybody agrees with these changes.

14. Not using peak flow when selecting results

The ATS/ERS standards state that the largest FEV1 and largest FVC regardless of which effort they came from should be reported. This is somewhat at odds with another ATS/ERS part of the standards that state that an FVC effort should be performed with maximal effort. Peak flow occurs during the effort-dependent part of the FVC maneuver and numerous investigators have shown that efforts with the largest peak flow often do not have the largest FEV1. This means that if an FEV1 is reported based solely on being the largest value from a group of efforts, it may also be from a submaximal effort.

	Blue:	Red:
FVC (L):	2.72	3.06
FEV1 (L):	1.73	1.99
PEF (L/sec):	6.28	3.82

Although not currently part of the ATS/ERS guidelines it is widely agreed that FEV1 should be selected from an FVC effort with the highest peak flow or at least near the highest peak flow.

15. Not using SVC and IVC when they are available

The ATS/ERS standards state that the largest FVC, regardless of its source, should be reported and used to calculate the FEV1/VC ratio. When lung volumes or a diffusing capacity are measured along with spirometry, there are potentially two additional vital capacity measurements, the slow vital capacity (SVC) from lung volumes and inspired volume (IVC) from the DLCO. If either of these are larger than the FVC, then they should be substituted for the FVC and the FEV1/FVC ratio re-calculated accordingly. Doing this often reveals airway obstruction that was not evident using the FVC from the spirometry effort alone.

16. Using FEF25-75 (aka MMEF)

FEF25-75 is the measurement that will not die. When it was originally defined in the late 1960’s it was touted as a way to measure the flow through smaller airways and to be able to diagnose “small airways disease”. It has since been shown by numerous investigators that it is a poorly reproducible measurement that is highly dependent on the FVC, that it actually says very little about the actual flows in the middle of an exhalation and that it’s usually reduced only when the FEV1 is reduced. As importantly the ATS/ERS criteria used to select the FEF25-75 from different efforts (i.e., the effort with the largest combined FVC and FEV1) is way to standardize selection and in not a way to determine the best FEF25-75 (if there is such a thing).

The ATS/ERS guidelines state that the sole indicator of airway obstruction is a reduced FEV1/FVC ratio and discourages the use of the FEF25-75.

17. Using a change in the FEV1/FVC ratio to indicate a positive bronchodilator response

The FEV1/FVC ratio is dependent on both the FVC and the FEV1. An increase in the FEV1/FVC ratio can just as easily be due to a decrease in FVC as it is to an increase in FEV1. For this reason the FEV1/FVC ratio should not be used to assess the response to bronchodilator. In addition the ATS/ERS guidelines state that a positive response to a bronchodilator is a 12% (and 0.20 L) increase in either FEV1 or FVC which means that both the FEV1 and FVC can increase significantly without a change in the FEV1/FVC ratio.

18. Not comparing current values to trends

Although a patient’s initial spirometry can be useful in diagnosing an underlying condition given the range of possible normal reference values this means there is always some uncertainty involved in this process. Spirometry however, is just as useful, if not more so, when monitoring the progress of a patient’s disorder or their improvement from therapy. In order to get the maximum amount of information out of a spirometry test current results must always be compared to prior results.

* * *

If I was asked to take these and make a real top ten list, in order of importance or how frequently I think they occur (from least to most), they would be:

Importance:	Error / Mistake:
10	#18 Not comparing trends
9	#16 Using FEF25-75
8	#15 Not using SVC and IVC
7	#14 Not using peak flow when selecting results
6	#4 Limits of ethnicity based reference equations
5	#11 Inadequate inspiration
4	#2 Wrong height
3	#7 Back extrapolation
2	#12 Patient leak
1	#8 Short Effort

One reason that spirometry isn’t relied on as much as it should be is that it has a relatively high rate of false positives and false negatives. To (badly) paraphrase Clauswitz, spirometry is simple but when testing people even the simple is very difficult. Spirometry looks simple but the number of possible ways in which it can be performed incorrectly is immense. It’s up to those of us that perform spirometry and those of us that interpret spirometry to be aware of the most common failure modes so that, as best as possible, we can reduce the false positives and false negatives.

This list is an attempt to categorize the most common mistakes and errors seen in spirometry. I suspect that there will be at least some disagreement about what is – and is not – included in this list. These are however, what (beyond the basic algorithms) I would want a pulmonary fellow being taught PFT interpretation to remember and to use.

References:

Brusasco V, Crapo R, Viegi G. ATS/ERS task force: Standardisation of lung function testing. Standardisation of spirometry. Eur Respir J 2005; 26: 319-338.

Brusasco V, Crapo R, Viegi G. ATS/ERS task force: Standardisation of lung function testing. Interpretive strategies for lung function tests. Eur Respir J 2005; 26: 948-968.

PFT Blog by Richard Johnston is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License

I like to think my lab is better than most but every so often something comes along that makes me realize I’m probably only fooling my self.

Earlier this week I was reviewing the DLCO test data for a patient with interstitial lung disease. At first glance the spirometry and DLCO results pretty much matched the diagnosis and I had already seen they weren’t significantly different from the last visit. The technician had written “fair DLCO reproducibility” which was reason enough to review the test data but I’ve actually been making a point of taking a careful look at all DLCO tests, not just the questionable ones, for the last couple of weeks. I took one look at the test data, put my head in my hands, and counted to ten before continuing.

	Reported:	%Predicted:	Test #1:	Test #2:	Test #3:
DLCO:	13.22	66%	10.08	92.17	16.36
Vinsp (L):			2.17	2.20	2.15
VA (L):	3.45	66%	2.89	2.93	4.02
DL/VA:	3.78	91%	3.49	31.5	4.07
CH4:			60.84	60.94	43.15
CO:			34.46	0.51	23.13

Even though the averaged DLCO results were similar to the last visit, the two tests they were averaged from were quite different. Reproducibility was not fair, it was poor. But far more than that, something was seriously wrong with the second test and the technician hadn’t told anybody that they’d had problems with the test system. {SIGH}. It’s awful hard to fix a problem when you don’t even know there is one in the first place.

I usually review reports in the morning the day after the tests have been performed, so the patient was long gone by the time I saw the results. This left me with a problem that I’m sure we’ve all had at one time or another and that was whether any of the DLCO results were reportable.

Towards this end the first thing I looked at was how well the patient had performed the tests. What I saw was that the inspired volumes ranged from 96% to 98% of the FVC and the BHT time ranged from 10.5 to 11.4 seconds, both of which are well within the ATS/ERS criteria. A look at the volume-time graphs showed a rapid inspiration, rapid expiration, no glitches during the breath-holding period and the expiratory sample was positioned within the alveolar plateau correctly. So mechanically at least, all of the DLCO maneuvers had been performed with good quality.

The real differences between the tests were in the baseline and exhaled gas concentrations. For this reason test #2 can pretty much be discounted immediately (and not just because the measured DLCO was 697% of predicted!). Our lab software reports the exhaled CO and CH4 as a percent of the inhaled concentration. An exhaled CO concentration of 0.51% is more than somewhat unlikely physiologically speaking (unless the breath hold period had been a minute or so but not for one that was 10 seconds long) and exhaled CO is usually somewhere between 20% and 60%. The reason for the low CO reading became more apparent when I looked at the baseline CH4 and CO.

I need to digress for a moment here and explain that our lab software does not report the baseline CH4 and CO readings, other than as the first second of the DLCO graph. If the CH4 and CO analyzer readings have a positive baseline (> 0.0) then their traces will be slightly elevated on the graph.

If the baseline CH4 or CO is negative however, they don’t appear in the graph at all and there is no way to tell how negative they are.

It’s possible to download the raw data from the DLCO test into a spreadsheet but since none of the computers attached to our test systems have spreadsheet software it can only be done from one of our desktop workstations. In addition the CO and CH4 are reported on the spreadsheet as fractional concentrations, and not as a percent of the inspired concentrations. It’s not real hard to convert back and forth, but is another step and all this tends to means that the baseline CH4 and CO aren’t easily available.

Anyway, the baseline CH4 and CO values should be close to zero and for the CH4 reading of 0.001 (+0.35%) this was true. The baseline CO reading however, was -0.194 (-63.9%) which means that for some reason the zero offset of the analyzer was way off and this in turn was causing the exhaled CO reading to be reduced as well.

Since our test systems go through what looks like a calibration before each DLCO test how could this happen? This is not documented (or at least not at all clearly) in the equipment’s manual (and I found this out only by talking to the software engineers several years ago) but the pre-test “calibration” only checks the gain of the CH4/CO analyzer and that is only so it can set a gain factor used when calculating the output from the analyzer. Even though the system samples both room air and the inhaled gas mixture during this step, as far as I can tell neither the gain or the zero of the analyzer are checked against a “real” calibration (which is usually performed first thing each morning).

So the DLCO from test #2 isn’t any good but how about the other two tests? For both test #1 and test #3 the baseline CH4 and CO were within normal limits. The difference between the two tests is entirely in the exhaled CH4 and CO. The lower CH4 in test #3 leads to a higher calculated VA and to some extent this is hard to explain physiologically because the inspired volumes from test #1 and test #3 were essentially identical. Interestingly, the exhaled CH4 and VA from test #2 match those of test #1. Admittedly the CO readings from test #2 were off but the baseline CH4 was normal and there is no reason otherwise to believe that the CH4 readings from test #2 are inaccurate.

The problem with using VA this way is that the patient has never has lung volume measurements before, only spirometry and DLCO, so there is no TLC for comparison. The VA from the last two visits however, was 2.72 L and 2.91 L and this looks a lot more like the VA from test #1 and test #2 than it does from test #3 and this makes test #3 questionable as well.

Based on all of the above this leaves test #1 with the least number of reasons to believe it is inaccurate. Except that when compared to the last two visits it is lower by approximately 3 ml/min/mmHg (-23%) and there has been essentially no change in spirometry in the meantime. Given the nature of interstitial disease this certainly isn’t all that unlikely but it at least leaves some doubt about test #1 as well, but it’s the least amount of doubt for any of the tests.

At this point I felt I had no choice but to select test #1 to be reported. I did equivocate though by also noting that the DLCO reproducibility was poor and test quality was questionable.

This situation brings a number of problems to light.

• Given the DLCO of 697% of predicted the tech should have (if possible) moved the patient to a different test system following test #2. At the very least, the DLCO gas analyzer should have been re-calibrated before test #3 was performed.
• The tech should have reported the problems with the test system.
• The tech should have written “poor reproducibility”, not “fair reproducibility” (which is admittedly a minor point but demonstrates a lack of understanding about what constitutes reproducibility).
• The test system should go through a real calibration before each DLCO test, and not just check the gain of the analyzers. At the very least it should compare the zero and gain of the gas analyzer against the last calibration and flag any discrepancies (which is also probably a good idea for the real calibrations).
• The test system software should report the baseline zero for CH4 and CO as part of the test results.

For the first three items I will need to check the lab’s written DLCO procedure and make sure it addresses these issues and update it if necessary. Regardless of what is or isn’t in the written procedure I will also need to put together an email for all the techs that will either remind or notify them of these problems (and hope they remember it in the future). I could wish that the lab staff were more technically inclined and better able to detect these kind of problems with testing but this doesn’t seem to be a common aptitude any more (if it ever was).

As far as the last two items go, I notified the manufacturer of our equipment about these issues and other similar ones over three years ago. We went through a software update three months ago and these problems are still there so I guess it wasn’t a priority for them.

To be honest, problems like this a fairly rare. I’ve been reviewing every DLCO test for the last three weeks (I’ve always reviewed the questionable ones and I get called in to assess results while the patient is still there reasonably often) and the majority of them have results that are either highly repeatable or where the differences can be explained by problems with the test maneuver. When DLCO tests go wrong however, it can be difficult to determine what, if anything, can be salvaged, particularly when you see the results the day after tests were performed.

One thing that continues to bother me is that we are kept at a distance from the hardware in our test systems by the computer software. I strongly suspect that the baseline error in the CO analyzer for test #2 is due to a computer error of some kind. Gas analyzers contain analog components and their zeros and gains do drift over time, but not that much and not in such a short time. The tests were performed at 3:09 PM, 3:16 PM and 3:21 PM, and it is unlikely that the zero baseline (and not the gain as well) for CO would start out normal, change drastically and then return to normal (!!) over the span of 12 minutes. But the gas analyzer is a sealed box buried inside the test system and there aren’t any readouts. There is no way to verify the gas analyzer’s operation other than through our test system software and even though I can download the raw data the zero and gain have already been corrected from the original analog signals by the software.

The ATS/ERS wrote technical standards for the accuracy and precision of test systems over 10 years ago but they are mostly based on technology that is 10 to 20 years older than that. What I’d like to see is that the next set of ATS/ERS standards not only address current technology and practices but include a mandate that all PFT test systems include a real-time “diagnostic” mode that allows direct access to the “real” raw data (voltages and the states of valves and switches) from all the system components (similar to the OBD2 interface in all cars and trucks). I’ll be the first to admit that most labs wouldn’t make much use of this (and even then probably mostly by a hospital’s clinical engineering department or the manufacturer’s service techs) but it would make it possible for technologists and researchers to verify that a test system is operating correctly and accurately.

PFT Blog by Richard Johnston is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License

Most of the COPD patients that are seen in my lab tend to have little change in their FEV1 from visit to visit but their FVC often changes significantly. A change in FVC is usually related to how long a patient is able to exhale and this in turn is usually related to how well they are feeling at the time. This would seem to imply that a significant change in FVC, particularly for a patient with COPD, is, if not clinically significant, at least clinically important even when the FEV1 hasn’t changed.

The problem with this is that expiratory time can be affected by things other than how the patient is feeling. Dyspnea and fatigue, of course. As importantly some technicians are better at motivating patients than other technicians so it can also be related to which technician is performing their tests. Even when the same technician is involved however, there is no guarantee that the level of motivation or a patient’s response to that motivation will be the same.

So how do you know if a change in FVC clinically significant or not?

Recently a spirometry report from a patient with very severe COPD came across my desk. When comparing the results to those of the last visit I could see that there had been a small (but not significant) increase in FEV1 but at the same time there had been a large (and significant) increase in FVC.

Visit 1:	Observed:	%Predicted:
FVC (L):	1.28	36%
FEV1(L):	0.53	19%
FEV1/FVC:	41	53%

Visit 2:	Observed:	%Predicted:
FVC (L):	1.93	55%
FEV1(L):	0.60	22%
FEV1/FVC:	31	40%

Expiratory time is not on our trend report but this got me curious so I pulled up the test data from the patient’s previous visit. From this I could see that the expiratory time had also improved.

	Expiratory Time:
Visit 1:	11.1
Visit 2:	15.0

But it also seemed that the improvement in FVC was greater than the increase in expiratory time could account for. Even with all of the test data in front of me however, there is no easy way to be sure this was the case and for this reason I downloaded the raw spirometry test data for both visits into a spreadsheet. After aligning the volume-time curves (based on their peak flow and FEV1), I truncated the longer FVC so it matched the shorter one, calculated the difference between the two curves and then graphed the results:

It was immediately evident that there was an improvement in exhaled volume throughout exhalation. It was also evident that there was a steady increase in volume occurring throughout the exhalation. To some extent this can be also seen when the flow-volume loops from the two visits are overlayed.

But I’d have to say it’s a lot more apparent on the volume-time curve than it is on the flow-volume loop.

Comparing the two volume-time curves seems to be a really good way to determine whether a change in FVC is clinically significant because it is easy to see whether the improvement is occurring simply because of an increase in expiratory time or because expiratory flow rates have improved. The problem is that this kind of comparison is not available in our lab software and in fact I don’t know of any lab software where it is possible. Unfortunately, because it takes a while to create the volume-time graphs it’s really not practical to do this on any kind of a routine basis.

However, the point of making the comparison was to see whether or not the change in FVC was only because there was a change in expiratory time. One measurement that provides a volume that is standardized by time and is routinely available in most test systems is the FEV6.

Note: The FEV6 and the FEV1/FEV6 ratio have been put forward as a replacement for the FVC and FEV1/FVC ratio a number of times but this has never seemed to gain any real traction. There are a number of reasons for this, partly because the FEV1/FEV6 ratio isn’t quite as sensitive as the FEV1/FVC ratio in assessing airway obstruction but I think this is more because the FEV6 is at its core an arbitrary value and does not reside in the same conceptual space as the FVC.

The FEV6 is measured long enough after the start of exhalation that even a relatively small change in expiratory flow rates over this interval can accumulate into a significant change in FEV6. When I looked at the FEV6 for the two visits, this was the case:

	FEV6:
Visit 1:	1.08 L
Visit 2:	1.33 L

Changes in FVC, particularly for patients with COPD, are often clinically significant. There are a number of confounding factors for FVC however, and for this reason a significant change in FVC is not necessarily clinically significant. This also means that the absence of a change in FVC does not mean that a significant change in expiratory flow rates hasn’t occurred. Since the FEV6 is not dependent on expiratory time in the same way that the FVC is, a change in FEV6 is going to be a more reliable indicator of changes in expiratory flow than the FVC.

This brings up an interesting question and that is occasionally when doing pre- and post-bronchodilator spirometry only the FVC increases significantly. I’ve discussed this previously and at the time I considered this to be an acceptable sign of a positive bronchodilator response even when it was largely due to an increase in expiratory time. I reviewed the example I had used at that time and found that the FEV6 had increased by 0.23 L (+14%) so it was still likely a positive bronchodilator response but at the moment I’m a lot more doubtful that an increase in expiratory time by itself, particularly if unaccompanied by a change in flow rates, should be considered a positive bronchodilator response.

The FEV1 and FEV1/FVC ratio are the primary indicators of airway obstruction and it is the lack of a significant change in FEV1 following an inhaled bronchodilator that is one of the definitions of COPD. A number of studies have shown however, that individuals with COPD often have a decrease in hyperinflation (increase in IC) and a decrease in dyspnea when using a bronchodilator. I am not aware of any study that has looked at the post-bronchodilator changes in FEV6 by itself (all the literature I’ve found has looked at the FEV1/FEV6 ratio and not the FEV6) and the FEV6 may well not be an indicator in the same way that IC can be. FVC does change over time in patients with COPD however, and the FEV6 is an easy way to determine whether or not these changes reflect a real change in expiratory airflow or are instead due to a confounding factor.

I would also note that there have frequently been attempts to measure the “terminal flows” (flow rates near RV) during a forced exhalation. Terminal flows are usually regarded as a reflection of flow through the smaller airways of the lung. Towards this end measurements such as the FEF75-85 and the FEV10% have been proposed but these values are highly dependent on the measured FVC and for this reason have poor repeatability and sensitivity. I’m not going to suggest that the FEV6 could be used as a measure of terminal flows, but it does seem that changes in the FEV6 would likely correlate with changes in terminal flows (at least in patients with COPD).

For all these reasons I’d recommend adding FEV6 to trend reports as a “truth check” on changes in FVC and as a starting point I don’t see any reason not to use what my lab considers to be a significant change in FEV1 and FVC (+/- 10% and +/- 0.20 L) when assessing changes in FEV6.

The FVC is defined as the maximal amount of air that can be exhaled after a maximal inhalation. The FVC is, of course, a function of an individual’s lung volume but for patients with lung disease, and in particular those with COPD, the FVC is also a function of expiratory time, dyspnea, fatigue and motivation as well as the amount of airway obstruction. For these reasons a change in FVC from one visit to another may or may not be clinically significant. A visual comparison of volume-time curves is able to show changes in flow rates during exhalation and for this reason can be used to determine if a change in FVC is due to a change in expiratory time or a change in expiratory flow rates (or both). Conversely, even when the FVC doesn’t change that doesn’t mean that expiratory flow rates haven’t changed and this is something that a comparison of the volume-time curves can also show. At the present time the option to compare volume-time curves from different patient visits isn’t available in any lab software. FEV6 however, is available and because it is measured at a specific time interval after the start of exhalation, changes in FEV6 will reflect changes in expiratory flow rates over that interval. Comparing the FEV6 from visit to visit can therefore help determine the cause of any change (or lack of change) in FVC.

PFT Blog by Richard Johnston is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License

We recently performed a 6-minute walk test with helium-oxygen (heliox) for a patient of one of the physicians that specializes in airway stenting. His reasons for the test weren’t particularly clear (and he hasn’t bothered to try to clarify them with me) but most probably it has to do with differentiating between central and peripheral airway obstruction. Interestingly, he predicted the patient would have a significant improvement in 6-minute walk distance and instead there was little difference between the heliox 6MWT and one performed with 3 LPM supplemental O2.

6MWT:	SaO2:	Distance:
80% Helium – 20% O2, by mask	95%	440 meters
3 LPM O2, by nasal cannula	98%	457 meters

Helium is an inert, insoluble, low mass gas and both its therapeutic use and its use in physiological measurements has to do with it’s low density (and the fact that it’s highly insoluble, but that’s for purposes different than those discussed here).

	Density (g/m3)
He	0.179
N2	1.251
O2	1.429
Air (78% N2, 21% O2)	1.293
Heliox (80% He, 20% O2)	0.429

A typical way to assess its effect is by comparing air and heliox flow-volume loops:

Interestingly, despite an apparent increase in flow rates there is usually no significant difference in FEV1 (one study showed a range of +2% to +7% in a group of over 1500 subjects). The most common heliox FVL measurements are the change in expiratory flow at 50% of the FVC (ΔMEF@50%) and the Volume of Isoflow (which is the point at which the air and heliox expiratory flows become equivalent). Many of the earlier studies with heliox also measured ∆MEF@75% and ∆FEF25-75, and a tiny handful of studies (particularly given the technical difficulties) have measured ∆RAW and ∆sGAW.

The increase in expiratory flows while breathing heliox and the lung volume at which they occur has been traditionally explained in terms of turbulent flow, laminar flow and the equal pressure points (EPP) of the airways. Specifically, the maximal expiratory flow at any given lung volume varies directly with the transpulmonary pressure (from elastic recoil and contraction of the respiratory muscles) and inversely with airway resistance between the alveoli and the EPP. The EPP is where the lateral airway wall pressure is equal to the pressure in the surrounding lung tissue. During a forced exhalation in normal subjects the EPP remains in the large airways until about 75% of the FVC has been exhaled.

The resistance to airflow in the large airways is primarily due to convective acceleration (the increase in gas velocity due to decreasing cross-sectional area as exhaled air moves from smaller to larger airways) and turbulence, both of which are dependent on gas density. Flow changes from laminar to turbulent as its velocity increases and this is described the Reynolds number:

Flows are generally laminar at a Reynolds number less than 2000, fully turbulent above 2800 and transitional inbetween. The Reynolds formula shows that because of heliox’s lower density at the same velocity turbulence will be reduced and a recent study estimated that it is reduced up to 50%. Once flow is turbulent however, density continues to be a primary factor and velocity is determined by:

Laminar flow predominates in the smaller airways (<2.0 mm in diameter) and laminar flow is dependent on gas viscosity and not gas density:

	Viscosity (10-5 Pa s)
He	1.87
N2	1.66
O2	1.95
Air (78% N2, 21% O2)	1.73
Heliox (80% He, 20% O2)	1.89

The important point is that turbulent flow is density-dependent and viscosity-independent, and laminar flow is viscosity-dependent and density-independent.

Because disorders like asthma, emphysema and chronic bronchitis narrow the smaller airways (although of course from different processes) and for these disorders this is where the primary resistance to exhalation resides.  This means that the difference between heliox-breathing and air-breathing flow-volume loops tends to be small for these disorders.

One study showed that in a group of subjects with chronic asthma, those that responded to a bronchodilator showed an increase in flow rates on their flow-volume loops similar to normal subjects whereas those subjects who did not have a bronchodilator response showed no significant change in their flow-volume loops. Moreover, the BD-responders showed a decrease in airway resistance (RAW) while on heliox and non-responders did not.

This very neat explanation has been called into question by a couple of recent papers that showed (at least mathematically) that despite the fact that laminar flow is primarily a visosity-related function that gases undergo a lateral acceleration (i.e. the direction of airflow changes) as they pass airway bifurcations and that the lower density of heliox improves airflow under these circumstances. If true this would imply that there should be improvements in expiratory flow rates regardless of whether the flow is laminar or whether it is turbulent.

So why didn’t our patient’s 6-minute walk distance improve? Their recent PFT results were low normal but still within normal limits:

	Observed:	%Predicted:
FVC:	2.78	82%
FEV1:	2.12	82%
FEV1/FVC:	76	99%
TLC:	4.45	83%
DLCO:	15.78	83%

And given their age (66) there’s nothing particularly remarkable about their flow-volume loop:

Interestingly, there is very little literature on exercise and heliox in normal subjects. One article performed constant-workload testing of a small group of young subjects and showed no significant change in VO2 or minute ventilation although it did show that the work of breathing decreased with heliox. Another study showed an increase in minute ventilation with heliox in an older cohort but unfortunately did not measure gas exchange.

The majority of research on heliox and exercise has been on subjects with pulmonary disease.  One relatively pertinent study of 6MWT testing on subjects with COPD showed that the heliox 6-minute walk distance was significantly higher than when performed on room air, nasal O2 or O2 by mask. This is generally seconded by a review article that closely analyzed eight studies of COPD and exercise and concluded that heliox improved exercise capacity and decreased dyspnea.

There is however, equivocal and sometimes contradictory evidence about why exercise capacity improves in COPD. Although one study on COPD and exercise with heliox showed no change in IC and hyperinflation others had exactly the opposite results. It’s been suggested that heliox improves CO2 elimination (CO2 diffuses many times faster in heliox than in air) and a number of studies have shown small but statistically significant decrease in PaCO2 while breathing heliox and one study showed a small improvement in Ve/VCO2. Regardless of the cause however, most studies showed a decrease in dyspnea and an increase in endurance time when patients with COPD exercised while breathing heliox.

None of this gives any physiological reason why our patient did not improve their 6-minute walk distance and I’m left with other factors as being more likely. I’m particularly curious that the patient said they had much more significant dyspnea and fatigue while breathing heliox than while on supplemental O2. This is somewhat contradictory to the literature that tends to indicate that subjects have a decrease in dyspnea with heliox (I’m also curious why the physician ordered the comparison 6MWT be performed with supplemental O2, particularly given the patient’s normal DLCO).

During the 6-minute walk the heliox and oxygen tanks were wheeled by a staff member and I have to wonder if this somehow set the pace for the patient. This is hard to verify since I did not witness the test being performed but the staff that were there said they stayed behind and the patient set the pace. This means it could be a musculo-skeletal or motivation issue but at the moment that has to remain speculative.

Given that heliox testing can be at least somewhat physiologically informative, why don’t we perform this kind of testing more often?

One reason is that performing spirometry with heliox is technically difficult. Pneumotachs are sensitive to gas density and must be calibrated using a heliox mixture. Given the difference in thermal conductivity and the speed of sound in heliox this limitation may also apply to mass flow sensors (hot wire anemometers) and ultrasonic spirometers but I’ve been unable to find any study that either verified this or attempted to use either of these devices. Although volume displacement spirometers are insensitive to gas composition (and this is one area where they have a significant advantage over flow sensors) they have become relative rare.

There is also a lack of standardization about how to prepare for the test. Some researchers had their subjects wash out the resident nitrogen in their lung by breathing heliox for 2 to 10 minutes before performing spirometry. Other (most) researchers have used 3 vital capacity breaths of heliox instead and at least one study showed no difference in ΔMEF@50% and the volume of isoflow between the two approaches in normal subjects. For this reason the 3-breath approach is probably reasonably acceptable for subjects with normal lungs but in subjects with airway obstruction and ventilation inhomogeneities it remains questionable.

Finally though, the real reason is that heliox testing lacks clinical relevance and this is largely due both to a lack of sensitivity and to testing variability. When it was first performed in the 1970’s the volume of isoflow in particular was thought to be able to detect obstruction in the small airways of smokers and individuals exposed to pollutants (aka small airways disease) but within a decade several studies showed little correlation between ΔMEF@50%, the volume of isoflow, the FEV1, and symptoms or exposure history. At the same time several studies have shown that the test-to-test variability for these values within the same individual is more or less the same magnitude as it is between normal subjects and those with “small airways disease”.

One reason for this is that the measurements depend on the FVC volume and one study of over 1500 subjects found that only about half were able to perform a heliox FVC that was within 5% of an air FVC. Interestingly and somewhat counter-intuitively, FVC’s performed with room air were far more often to be larger than those performed with helox. The same studies also showed that the test-to-test variability for the volume of isoflow is exceptionally high even when FVC reproducibility was controlled and that this variance did not improve with experience.

Heliox has been shown to be therapeutically useful in asthma, COPD, upper airway disorders, airway tumors, vocal cord dysfunction and a variety of other problems. Although testing with heliox can give some indication about whether the majority of an individual’s airway resistance resides in the larger or smaller airways, the results from this testing are often unrepeatable and for these reasons the usefulness of heliox will likely remain on the therapeutic side of things.

References:

Ashutosh K, Mead G, Dickey JC, Berman P, Kuppinger M. Density dependence of expiratory flow and bronchodilator response in asthma. Chest 1980; 77(1): 68-75

Babb TG, DeLorey DS, Wyrick BL. Ventilatory response to exercise in aged runners breathing He-O2 or inspired CO2. J Appl Physiol 2003; 94: 685-693.

Berend N, Nelson NA, Rutland J, Marlin GE, Woolcock AJ. The maximum expiratory flow-volume curve with air and a low-density gas mixture. An analysis of subject and observer variability. Chest 1981; 80(1): 23-30.

Despas PJ, Leroux M, Macklem, PT. Site of airway obstruction in asthma as determined by measuring maximal expiratory flow breathing air and a helium-oxygen mixture. J Clin Invest 1972; 51: 3235-3242.

Dominelli PB, Foster GE, Dominelli GS, Henderson WR, Koehle MS, McKenzie DC, Sheel AW. Exercise-induced arterial hypoxemia and the mechanics of breathing in healthy young women. J Physiol 2013: 591(12): 3017-3034.

Dosman JA, Chong P, Cotton DJ. Detection of peripheral airways obstruction in smokers using air vs helium spirometry. Bull Eur Physiopath Resp 1978; 14: 137-143.

Fairshter RD, Novey HS, Wilson AF. Site and duration of bronchodilation in asthmatic patients after oral administration of terbutaline. Chest 1981; 79(1); 50-57.

Gelb AF, Klein E. The volume of isoflow and increase in maximal flow at 50 percent of forced vital capacity during helium-oxygen breathing as tests of small airway dysfunction. Chest 1977; 71(3): 396-399.

Hess DR, Fink JB, Venkataraman ST, Kim IK, Meyers TR, Tano BD. The history and physics of heliox. Respir Care 2006; 51(6): 608-612.

Hunt T, Williams MT, Frith P, Schembri D. Heliox, dyspnoea and exercise in COPD> Eur Respir Rev 2010; 19: 115, 30-38.

Katz IM, Martin AR, Muller P-A, Terzibachi K, Feng C-H, Caillibotte G, Sandeau J, Texereau J. The ventilation distribution of helium-oxygen mixture and the role of inertial losses in the presence of heterogeneous airway obstructions. J Biomechanics 2011; 44: 1137-1143.

Knudson RJ, Bloo, JW, Kaltenborn WT, Burrows B, Lebowitz MD. Assessment of air vs helium-oxygen flow-volume curves as an epidemiological screeing test. Chest 1984; 86(3): 419-423.

Lam S, Abboud RT, Chan-Yeung M, Tan F. Use of maximal expiratory flow-volume curves in air and helium-oxygen in the detection of ventilatory abnormalities in population surveys. Am Rev Respir Dis 1981; 123: 234-237.

MacDonald JB, Cole TJ. The flow-volume loop: reproducibility of air and helium-based tests in normal subjects. Thorax 1980; 35: 64-69.

Marciniuk DD, Butcher SJ, Reid JK, MacDonald GF, Eves ND, Clemens R, Jones RL. The effects of helium-hyperoxia on 6-minute walking distance in COPD: A randomized controlled trial. Chest 2007; 131(6): 1659-1665.

Palange P, Valli G, Onorati P, Anonucci R, Paolette P, Rosato A, Manfredi F, Serra P. Effect of heliox on lung dynamic hyperinflation, dypnea and exercise endurance capacity in COPD patients. J Appl Physiol 2004; 97: 1637-1642.

Pecchiari M, Pelucchi A, D’Angelo E, Foresi A, Milic-Emili J, D’Angelo E. Effect of heliox breathing on dynamic hyperinflation in COPD patients. Chest 2004; 125: 2075-2082.

Rudnow M, Hill AB. Helium-Oxygen mixtures in airway obstruction due to thyroid carcinoma. Can Anaesth Soc J 1986; 33: 498-501.

Skrinkas, GJ, Hyland RH, Hutcheon MA. Using helium-oxygen mixtures in the management of acute upper airway obstruction. Can Med Assoc J 1983; 128: 555-558.

Swidwa DM, Montenegro HD, Goldman MD, Lutchen KR, Saidel GM. Helium-Oxygen breathing in severe chronic obstructive pulmonaory disease. Chest 1985; 87(6): 790-795.

Zeck RT, Solliday NH, Celic L, Cugell DW. Variability of the volume of isoflow. Chest 1981; 79(3): 269-272.

PFT Blog by Richard Johnston is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License

While reading a recently published article I found they had performed response to hypoxia and hyperoxia testing as part of the study. At one time or another in the past I’ve read about response to hypoxia testing but I’d never heard about hyperoxia testing before. I had some difficulty understanding their interpretation of the study’s results and for this reason I’ve spent some time reading up on the subject. I’m not sure this helped because there appears to be a lack of consensus not in only how to perform these tests but also in how they are interpreted, except perhaps in the most simplistic sense. Hypoxia and hyperoxia testing has been performed primarily to gain a deeper understanding of the way in which the peripheral (carotid) and central chemoreceptors function. There are a variety of sensor-feedback network models and results are often presented in terms of one model or another and this makes comparing results from different studies difficult. Interpretation and comparison is further complicated by the fact that results depend not only on the length of time that hypoxia or hyperoxia is maintained but whether the subject was exposed to hypoxia, hyperoxia or hypercapnia previously.

The ventilatory response to hypoxia tends to have three phases. First, once a subject begins breathing a hypoxic gas mixture within several seconds there is a rapid increase in minute ventilation known as the Acute Hypoxic Ventilatory Response (AHVR). Second, after several minutes there is a decrease in ventilation and this is usually called the Hypoxic Ventilatory Depression (HVD). Third, there is a progressive rise in ventilation after several hours which is related to acclimatization to altitude. It is the first phase, AHVR, that is most commonly measured during a hypoxic ventilatory response test. The actual length of time that is spent in any of these phases is widely variable between individuals and there is also a relatively large day-to-day variability within the same individual.

Although the majority of researchers have measured the response to hypoxia using changes in a subject’s minute ventilation the inspiratory pressure 100 milliseconds following occlusion of the mouthpiece (P₁₀₀ or P0.1) has also been used.

One of the rare points of consensus is that almost all researchers agree that end-tidal CO2 (PetCO2) needs to be rigorously controlled and this is because increases or decreases in CO2 can affect the ventilatory response to hypoxia. For this reason PETCO2 needs to be monitored and it needs to be possible to add (or remove) CO2 to the breathing circuit and most commonly PetCO2 is maintained at 1-2 mm Hg above resting levels.

There is also general agreement that the changes in ventilation that occur with hypoxia are linearly related to SaO2 and for this reason, an oximeter is most commonly used to measure a subject’s level of hypoxia. One conceptual problem with this is that chemoreceptors are sensitive to the partial pressure of O2, not to oxygen saturation (this has been tested with controlled levels of carboxyhemoglobin), but since their output is non-linear (more probably hyperbolic) the ventilatory response better matches SaO2 than it does PaO2. Having said this, many researchers have measured PetO2 instead of SaO2 (PetO2 results are instantaneous while SaO2 lags behind) and then either plotted PetO2 logarithmically or made the assumption that PetO2 was equivalent to PAO2 and then calculated SaO2 using the Severinghaus formula.

Ignoring the more esoteric forms of testing there are generally three approaches to hypoxia testing and researchers have devised both open-circuit and closed-circuit systems for this. A steady-state approach is where a subject breaths gases with specific concentrations of O2 and CO2 and measurements are not taken until ventilation stabilizes. A transient approach is where specific concentrations of O2 and CO2 are applied only over a small number of breaths. A progressive approach is where FiO2 decreases throughout the test. Steady-state and transient approaches are usually performed with open-circuit systems whereas a progressive test tends to require a closed-circuit system.

There are, of course, both proponents and opponents to each approach. Although the steady-state approach allows for data to be obtained and averaged over a reasonable period of time it also risks crossing into the HVD phase of hypoxic response. The transient approach definitely limits measurements to the AHVR phase but the small number breaths that measurements are taken from can easily be affected by small variations in tidal volume and respiratory rate and the length of the test is too short for any significant change in SaO2 to occur and be measured. Because FiO2 decreases steadily during a progressive test, it is thought that the hypoxic response remains in the AHVR phase but it is not clear whether this has been verified or not.

Given the above, unless a specific research protocol is being followed, either a steady-state test limited to two or three minutes or a progressive test appear to be the most likely approaches toward obtaining meaningful and reproducible results.

During hypoxia testing with an open-circuit system the FiO2 is controlled by the flow rate of air and nitrogen into the system and PetCO2 is adjusted by an inflow of CO2. Measurements are usually made at one, two or three oxygen levels. Some researchers have adjusted the FiO2 in order to attain a specific SpO2 (often 90%, 85% or 80%), a specific PetO2 (such as 40, 60 or 80 mm Hg) or one or more preset FiO2’s (such as 5%, 8% or 12% O2).

During hypoxia testing with a closed-circuit system the FiO2 progressively decreases as it is consumed by the subject and testing usually proceeds until a target SpO2 (often 80% or 85%) is attained. PetCO2 is adjusted by changing the flow rate of air sent through a CO2 scrubber.

For either type of test most commonly SpO2 is plotted against minute ventilation and the change in minute ventilation versus the change in SpO2 (ΔVe/ΔSpO2) is calculated. As mentioned SpO2 can also be plotted against inspiratory occlusion pressure P0.1 and the ΔP0.1/ΔSpO2 calculated.

So, what’s a normal hypoxic response?

There is no clear answer to this. I have over two dozen pulmonary function textbooks and ventilatory hypoxic response testing was mentioned in only three of them (Cotes, Ruppel, Wilson) and they declined to publish normal values. I’ve found mean ΔVe/ΔSpO2 values in three studies (Rebeck et al, Stickland et al, Zhang et al) and they ranged across a magnitude of difference (1.65 L/%SaO2, 0.11 L/%SaO2, 0.93 L/%SaO2) despite somewhat similar methodologies. Dean et al noted that a 5 minute exposure to isocapnic hypoxia (FiO2 between 8-10%) increased Ve on average by 37%. Finally, Garcia-Rio et al reported a P0.1 of 0.032 Kpa/%SaO2 (10.4 cmH2O/%SaO2) for normal subjects.

Variability between even normal subjects is high. Rebeck et al reported a range of 0.26 to 4.12 L/%SaO2 in a group of 9 normal subjects and Stickland et al reported a range of 0.05 to 0.67 L/%SaO2 in a group of 30 normal subjects.

As a complication the AHVR tends to be elevated at altitude, in sedentary individuals, in COPD, hyperthermia, thyrotoxicosis, hypercapnia, following a meal and if there has been prior exposure to hyperoxia. AHVR tends to be reduced in athletes, during sleep, during starvation, with hypocapnia and with myxedema.

Hypoxic ventilatory response is often measured and compared between groups or between interventions or over time, and frequently either the results are reported in non-standard ways or only the statistical significance of the differences is reported. In addition, because hypoxic response testing is intended as research on the peripheral and central chemoreceptors, mixtures of gases containing different levels of O2 and CO2 are often used and interpreting the results from these studies in terms of a simple ΔVe/ΔSpO2 response is particularly difficult.

Hyperoxic ventilatory response testing appears to be performed even less frequently than hypoxic ventilatory response testing. The response to hyperoxia is biphasic. First, when a subject initially starts breathing a hyperoxic gas mixture there tends to be a decrease in ventilation. Within approximately two minutes however, ventilation tends to increase and this second phase is called hyperoxic hyperventilation.

Both the hypoventilation and the hyperventilation phases of hyperoxia have been studied by various researchers using both open circuit and closed circuit systems. As with response to hypoxia there is a general consensus that PetCO2 needs to be controlled and again this is because changes in CO2 affect the ventilatory response to hyperoxia.

Notably, there appears to be a dose relationship during the hyperventilation phase in that higher FiO2’s are associated with a higher minute ventilation. For prolonged hyperoxic studies understanding the dose relationship is important since investigators have used hyperoxic FiO2’s anywhere from 24% to 100%. For studies of the initial phase of hyperoxic hypoventilation phase it is unclear if there is a dose relationship since most studies appear to have been performed with FiO2’s near 100%.

During hyperoxia, the initial phase of hypoventilation has been attributed to the inhibition or denervation of the peripheral chemoreceptors. Research has shown that the hyperventilatory phase is due to the action of the central chemoreceptors and in particular it has been suggested that hyperoxia and the reactive oxygen species stimulates the central CO2 receptors.

One interesting point is that research on the hyperoxic ventilatory response calls into question the results from response to CO2 testing. The gas mixture that is most commonly used in response to CO2 tests is 7% CO2, 93% O2 and the elevated FiO2 has been shown to be able act either as a respiratory depressant or as a respiratory stimulant depending on the length of the test and the FiCO2. This is a confounding factor that is rarely appreciated.

So what’s a normal hyperoxic ventilatory response?

Again, there is no clear answer to this. Hyperoxic ventilatory response testing is not mentioned in any of my pulmonary function textbooks. Stickland et al reported that the initial hypoxic hypoventilatory response for normal subjects was a decrease in Ve of 0.73 L and for those with COPD the decrease in Ve was 2.62 L. As mentioned, there is a dose relationship for the hyperventilatory phase. Becker et al reported that in normal subjects Ve increased by 21% at 30% O2, an increase of 61.3% at 50% O2 and 114.7% at 70% O2. As with response to hypoxia testing the range of responses is also quite wide since Becker et al noted that increases in ventilation ranged from +43% to +287% at an FiO2 of 70%.

Although the response the hypoxia and hyperoxia are physiologically important, the clinical utility of hypoxic or hyperoxic ventilatory response testing appears to be exceptionally limited (if it exists at all). This partly due to the lack of clear normal values and standardized approaches to testing but even more so because trying to understand the ventilatory response to hypoxia or hyperoxia solely in terms of a change in ventilation is a misleading oversimplification. There is also a wide range of responses even among normal subjects so it is unclear what would be considered abnormal.

The interaction between peripheral and central chemoreceptors, and particularly how the different ventilatory phases arise, is complex and a number of competing models have been proposed. How an individual’s response to hypoxia, hyperoxia and hypercapnia is interpreted depends greatly on which of these models is selected and to be honest, despite careful reading many of the fine points of these models eludes me. On the plus side though, I have a much better understanding the paper that started the question of hypoxic and hyperoxic testing in the first place, so for me at least it wasn’t a waste of time.

References:

Becker HF, Polo O, McNamara SG, Berthon-Jones M, Sullivan CE. Effect of different levels of hypoxia on breathing in health subjects. J Appl Physiol 1996; 81(4): 1683-1690.

Cotes JE, Chinn DJ, Miller MR. Lung Function. Physiology, Measurement and Application in Medicine. 6^th Edition. Published by Blackwell, 2006.

Dahan A, DeGoede J, Berkenbosch A, Olievier ICW. The influence of oxygen on the ventilatory response to carbon dioxide in man. J Physiol 1990; 428: 485-499.

Dean JB, Mulkey DK, Henderson RA, Potter SJ, Putnam RW. Hyperoxia, reactive oxygen species and hyperventilation: oxygen sensitivity of brain stem neurons. J Appl Physiol 2004; 96: 784-791.

Duffin J. Measuring the ventilatory response to hypoxia. J Physiol 2007; 584(1): 285-293.

Garcia-Rio F, Pino-Garcia JM, Racionero MA, Terreros-Caro JG, Gomezx-Mendieta MA, Prados C, Villasante C. Long-term within-subject variability of inspiratory neural drive response to hypoxia. Chest 1998; 114: 521-525.

Honda Y, Tani H, Masuda A, Kobayashi T, Nishino T, Kimura J, Masuyama S, Kuriyama T. Effect of prior O2 breathing on ventilatory response to sustained isocapnic hypoxia in adult humans. J Appl Physiol 1996; 81(4): 1627-1632.

Jensen D, Mask G, Tschakovsky ME. Variability of the ventilatory response to Duffin’s modified hyperoxic and hypoxic rebreathing procedure in health awake humans. Respiratory Physiology and Neurobiology 2010; 170: 185-197

Liang PJ, Bascom DA, Robbins PA. Extended models of the ventilatory response to sustained isocapnic hypoxia in humans. J Appl Physiol 1997; 82(2): 667-677.

MacFarlane DJ, Cunningham DJC. Dynamics of the ventilatory response in man to step changes in end-tidal carbon dioxide and of hypoxia during exercise. J Physiology 1992; 457: 539-557.

Rebeck AS, Campbell EJM. A clinical method for assessing the ventilatory response to hypoxia. Am Rev Resp Dis 1974; 109: 345-350

Ruppel, G. Manual of pulmonary function testing. 8^th Edition. Published by Mosby, 2003.

Stickland MK, Fuhr DP, Edgell H, Byers BW, Bhutani M, Wong EYL, Steinback CD. Chemosensitivity, cardiovascular risk and the ventilatory response to exercise in COPD. PloS One 2016; 11(6): e0158341.

Weil JV, Zwillich CW. Assessment of ventilatory response to hypoxia. Methods and interpretation. Chest 1976; 70: S124-S128.

Wilson AF, editor. Pulmonary function testing indications and interpretations. A project of the California Thoracic Society. Published by Grune and Stratton, 1985.

Zhang S, Robbins PA. Methodological and physiological variability within the ventilatory response to hypoxia in humans. J Appl Physiol 2000; 88: 1924-1932.

PFT Blog by Richard Johnston is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License