During exhalation air flow occurs because of the pressure difference between the alveoli and the surrounding atmosphere. The increase in alveolar pressure acts to compress the air inside the lung and because of this compression the decrease in lung volume during exhalation is always going to be greater than the volume of exhaled air. This effect is known as Thoracic Gas Compression (TGC).

The flow rate that occurs for a given alveolar pressure depends primarily on airway resistance. When an individual has airway obstruction their airway resistance is increased and they often attempt to increase their expiratory flow by increasing the amount of force they apply during exhalation. This increased force further compresses the air inside the lung and increases TGC. Numerous researchers have shown that there is usually a substantial differences in TGC between subjects with normal lungs and those with airway obstruction.

Routine spirometry and lung volume tests cannot measure thoracic gas compression. It can only be measured in a special kind of plethysmograph. Unfortunately the nomenclature for this type of plethysmograph is a bit muddy and it is variously known as a transmural, constant pressure, volume-displacement or flow plethysmograph (I prefer volume-displacement because I think this sums up its mode of operation most succinctly). In this type of plethysmograph the subject breathes in and out through an opening in the box. The expansion and contraction of their lungs causes air to flow in and out of the box through a separate opening. The pressure inside the plethysmograph is monitored and used to compensate for any delays in box flow.

Because the air in the lung is compressed during a forced exhalation this means that more air flows into the plethysmograph than flows out through the subject’s mouth. These two signals are compared to determine the difference in flow rates.

 

Largely because it is difficult to compare an entire curve, researchers have tended to report the difference in expiratory flow rates at specific lung volumes. One approach as been to use the flow rates at 25%, 50% and 75% of vital capacity. Another has been to correct and re-calculate FEV1 for TGC.

Regardless of how it is measured investigators have shown that the difference between mouth and thoracic flow rates increase with increasing airway obstruction and decrease when it lessens. This has been shown in patients with asthma, both during a methacholine challenge and after bronchodilation. Patients with emphysema tend to have a large difference between mouth and thoracic flow rates, particularly at high lung volumes, and not unsurprisingly, patients with interstitial lung disease tended to have the least discrepancy between the two. Although the difference between mouth and thoracic flow rates is usually highest close to TLC, obese patients tended to have significant differences even at lower lung volumes.

Of note, one group of researchers used the difference between exhaled mouth and box volumes to estimate alveolar pressure and thereby continuously calculate airway resistance during a forced exhalation. This seems to be an elegant answer to the accepted practice of measuring airway resistance but there must be a flaw to this approach that I am not aware of, either in its basic assumptions or to its clinical relevance, since this technique does not appear to have been used by any but the original researchers.

The mechanical difference between the more common constant-volume and the volume-displacement plethysmograph is relatively small and plethysmographs capable of both modes of operation have been available since at least the 1980’s. They are still available from one or more different equipment manufacturers and are equipped with software to measure mouth and thoracic flow-volume loops and to overlay them. So why isn’t this technique more commonly known and used?

Although there is a general correlation between the degree of airway obstruction and the difference in thoracic and mouth flow rates, I think the key factor for its lack of general adoption is that it is not individually predictable. This is due in part to differences in patient effort, static recoil and available muscle strength. It may also be due in part to the site and the nature of the airway resistance as well. For this reason there are no normal values and although there are statistically significant differences following bronchodilation for a study group as a whole that is different from knowing what’s the significant difference for an individual. In addition there has been no standardization in how it is measured (i.e. flow at 50% vs corrected FEV1 as well as the fact that some researchers report the volumes at which flows are measured either relative to TLC or relative to RV which means that the flow rates at 25% from one researcher can be the same those as 75% from another). A final barrier is that measurement of TGC requires specialized equipment that is unlikely to be generally available.

Thoracic gas compression occurs to one extent or another in everyone. There is a general correlation between airway obstruction and the degree of TGC. The amount of an individual’s TGC or changes in TGC following treatment or therapy may well be clinically relevant but the lack of normal values means that its use should be approached with caution.

References:

Coates AL, Peslin R, Rodenstein D, Stocks J. Measurement of lung volumes by plethysmography. Eur Respir J 1997; 10: 1415-1427.

Coates AL, Desmond KJ, Demizio D, Allen P, Beaudry PH. Sources of error in flow-volume curves. Effect of expired volume measured at the mouth vs that measured in a body plethysmograph. Chest 1988; 94: 976-982.

Pellegrino R, Confessore P, Bianco A, Brusasco V. Effects of lung volume and thoracic gas compression on maximal and partial flow-volume curves. Eur Repir J 1996; 9: 2168-2173.

Piirila PL, Hodgson U, Wuorimaa T, Smith HJ, Sovijarvi ARA. Thoracic gas compression during forced expiration in patients with emphysema, interstitial lung disease and obesity. BMC Pulmonary Medicine 2014; 14: 34.

Sharafkhaneh A, Officer TM, Goodnight-White S, Rodarte JR, Boriek AM. Novel method for measuring effects of gas compression on expiratory flows. Am J Physiol Regul Integ Comp Physiol 2004; 287: R479-R484.

Sharafkchaneh A, Babb TG, Officer TM, Hananla NA, Sharafkhaneh H, Boriek AM. Confounding effects of thoracic gas compression ob measurement of acute bronchodilator response. Am J Resp Crit Care Med 2007; 175: 330-335.

Stanescu D, Veriter C, Van Leemputten R, Brasseur L. Constancy of effort and variability of maximal expiratory flow rates. Chest 1979; 76(1): 59-63.

Walamies MA. Thoracic gas compression profie during forced expiration in health and asthmatic schoolchildren. Respir Med 1998; 92: 173-177.

Zamel N, Jones JG, Bach SM, Newburg L. Analog computation of alveolar pressure and airway resistance during maximum expiratory flow. J Appl Physiol 1974; 36(2): 240-245.

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

A number of physicians and researchers I’ve known and respected have said that in spirometry it always comes down to FEV1 since it is the primary indicator for airway obstruction. Certainly FVC and the FEV1/FVC ratio are important but because patients can stop exhaling early for any number of reasons FVC can be underestimated which in turn can cause the FEV1/FVC ratio to be overestimated so they are not quite as reliable as FEV1.

There are, of course, a number of factors that can cause FEV1 to be mis-estimated. It can be underestimated due to cough or glottal closure and it can be overestimated because of excessive back-extrapolation. Nevertheless, I think that overall the FEV1 tends to be the most accurate and reliable number obtained from spirometry.

This spirometry report came across my desk this morning: 

	Observed:	% Predicted:	Predicted:
FVC (L):	5.01	114%	4.39
FEV1 (L):	3.86	117%	3.30
FEV1/FVC:	77	103%	75
PEF (L/sec):	4.91	55%	8.99

Because a reduced FEV1 is a reliable indicator of airway obstruction doesn’t that mean that a normal or as in this particular case, a slightly elevated FEV1 rules it out? Well, actually no, it doesn’t.

At first glance this looked like a very normal spirometry test, although I was a bit curious about the low PEF. For this reason I pulled up the raw test results and although I saw they were very reproducible one look at the flow-volume loop told me what was going on.

 

The Cardio-Thoracic Surgery division of my hospital has an active tracheal reconstruction and stenting program. For this reason we see a fair number of patients with tracheomalacia. Tracheomalacia occurs for a wide variety of reasons and can be congenital or caused by intubation or it can be secondary to cystic fibrosis, relapsing polychondritis, emphysema and a number of other lung diseases. What it basically means is that the trachea has lost some or all of its rigid cartilagenous support and has become floppy and easily collapsible. When a patient with tracheomalacia performs a Forced Vital Capacity, the trachea collapses and limits expiratory flow. This is shown by the plateau on the flow-volume loop.

When you suspect tracheomalacia on the basis of a flow-volume loop one key element that needs to be confirmed is whether the loops are reproducible or not. Performing an FVC maneuver with maximal effort always causes airways to undergo some degree of narrowing and it is this factor that tends to define the maximal flow-volume loop envelope. When an FVC maneuver is performed with a relaxed rather than a maximal effort, airways will not narrow as much and, depending on the underlying condition, the FEV1 can be significantly larger than one from a maximal effort. This is a reason that we use Peak Flow as one of the criteria when selecting which spirometry effort will be reported.

 

Most importantly, however, when a patient performs an FVC sub-maximally it is very difficult to do this reproducibly. Like anybody else, an individual with tracheomalacia will have a certain amount of variability in their spirometry efforts, nevertheless there will be a threshold flow rate that their efforts should not exceed. When efforts are maximal, the maximal flow rates in the flow-volume loops should cluster near the threshold.

 

It also helps if there appears to be a true plateau, as if the effort was cut off at the threshold, but the narrowing that occurs from tracheomalacia can be variable so this isn’t a given. In this case, the patient’s flow-volume loops were somewhat erratic, but did pretty much met the threshold criteria. The only loop that didn’t was from an exceptionally erratic effort where the peak flow, if that is what it was, occurred late in the effort.

This patient had a slightly elevated FEV1 and FEV1/FVC ratio and yet still has significant airway obstruction. The patients that we see with tracheomalacia usually have a reduced FEV1 so there usually isn’t a question that some level of airway obstruction is present. You can’t say that this patient’s FEV1 was a lie because his FEV1 really was 117% of predicted, but this is a case where FEV1 alone is misleading. The reduced PEF is a clue, but in a normal flow-volume loop peak flow occurs only momentarily so although suggestive it is not definitive in any way. In fact, none of the numerical results that come from an FVC maneuver can tell you that an expiratory plateau is present (it might be possible to use the MEF25%, MEF50% and MEF75% for this purpose but these are only single points on the flow-volume curve and I’ve haven’t seen any research that links these values to an expiratory plateau). I doubt that there are any computer algorithms that would have caught this and I also suspect that there are many places where spirometry is regularly performed where it would not be recognized. 

Tracheomalacia is a consequence of many lung diseases. It is often unrecognized because it is either relatively minor or because other forms of airway obstruction predominate. The more advanced cases we see because of our hospital’s Cardio-Thoracic Surgery’s tracheomalacia clinic are less common, but there is a distinct flow-volume loop signature for tracheomalacia that anybody performing or interpreting spirometry should be aware of. 

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

Internet Resources for Pulmonary Patients:

A tri-fold brochure for patients.

PDF Version

MS Word Version

LibreOffice Version

Adult PFT Reference Equations:

Spreadsheets containing the adult reference equations for spirometry, lung volumes, DLCO and 6-minute walk with source bibliographies.

Adult Reference Equation Explorer – MS Excel Version

Adult Reference Equation Explorer – LibreOffice Version

LOINC PFT Codes:

The LOINC coding system is used to identify medical information transmitted between different medical institutions. The following spreadsheets contain the LOINC codes for pulmonary function tests.

Excel Version

LibreOffice Version

---

John B. West Respiratory Physiology Lectures on YouTube

Based on his lectures and his classic textbook, Respiratory Physiology.

1. Structure and Function
2. Ventilation
3. Blood Gas Transport
4. Acid-Base Balance
5. Diffusion
6. Pulmonary Blood Flow
7. Pulmonary Gas Exchange, Part 1
8. Pulmonary Gas Exchange, Part 2
9. Mechanics of Breathing, Part 1
10. Mechanics of Breathing, Part 2
11. Control of Ventilation
12. Defense Systems of the Lung
13. Respiration under Stress
14. Respiration at the Limit

---

Websites:

AARC Office Spirometry Certificate Program

The Office Spirometry Certificate program is intended to provide a way for people outside the traditional pulmonary function lab setting to demonstrate understanding and receive quality feedback on performance.  $$$.

ANZRS

Australia and New Zealand Society of Respiratory Science

ATS Spirometry Waveforms

Two sets of standardized waveforms that are used to test spirometry software. The first set (24 files) was created by John Hankinson at NIOSH in Morgantown, WV. The data is in liters at 10 msec intervals (i.e. 100 hz) and the second set (26 files) contain flows (L/sec), at 0.002 second (500 hz) sampling intervals.  Since the files consist solely of volume data points, flow is synthesized by calculating the difference between adjacent volumes and dividing by the difference in time.

ATS/ERS Pulmonary Function Testing Guidelines

The ATS/ERS standards for spirometry, lung volumes, diffusing capacity, 6-minute walks, and PFT interpretation.

COAL WORKERS’ HEALTH SURVEILLANCE PROGRAM (CWHSP)

A NIOSH program for training and certification of spirometry clinics for regular testing of coal workers.

Global Lung Initiative

Resource for the GLI spirometry reference equations. Downloadable software, spreadsheet and articles.

NIOSH Spirometry

Links to NIOSH spirometry training programs, training guide, and publications.

NIOSH Spirometry Longitudinal Data Analysis (SPIROLA) Software

SPIROLA software is an easy-to-use visual and quantitative tool intended to assist health care providers in monitoring and interpreting computerized longitudinal spirometry data for individuals as well as for a group.  Primary audience is Occupational testing.  No $$$.

PubMed

Searchable database of biomedical journal articles with links.  Maintained by the NIH.

Spirohub

British online community for healthcare professionals with an interest in respiratory care .

Spirometry.  Interpreting the flow-volume loop.

A website with an introduction to spirometry and the flow-volume loop.  In English, French and Dutch.

Spirxpert

On-line training in spirometry concepts, testing issues, problems, and statistics. Content managed by Philip Quanjer, MD.  In English, Dutch, French and Spanish.

---

Open Access Pulmonary Journals:

These journals are either completely open access or journals where articles are available a year or two after the publication date.

BMC Pulmonary Medicine

BMJ Open Respiratory Research

Breathe

Case Reports in Pulmonology

Clinical Medical Insights: Circulatory, Respiratory and Pulmonary Medicine

European Clinical Respiratory Journal

European Respiratory Journal (18 month delay, all issues back to 1988)

Hypoxia

International Journal of Chronic Obstructive Pulmonary Disease

Journal of Applied Physiology (1 year delay, all issues back to 1996)

Journal of Pulmonary & Respiratory Medicine

Journal of Thoracic Disease

Lung India

Multidisciplinary Respiratory Medicine

PLOS One Pulmonology

Respiratory Care (1 year delay, all issues back to 2003)

Respiratory Medicine Journal (2 year delay, all issues back to 2010)

Respiratory Research

Thorax (3 year delay, all issues back to 1946)

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

There have been numerous criticisms of the single-breath DLCO technique, many of them quite valid. In particular, the standard equation for calculating DLCO makes no consideration for the inspiratory and expiratory phases of the maneuver when lung volume and the alveolar capillary surface area are changing. Some investigators have devised ways of correcting for inhalation and exhalation, however other investigators have sidestepped the issue entirely by showing that DLCO can instead be calculated from information acquired only during exhalation.

During exhalation, once the gas that exits the lung leaves the alveoli, diffusion ceases. Gas that exits the lung during the early part of exhalation will have had less time for CO to diffuse from the alveoli into the pulmonary capillaries and will have a higher concentration of carbon monoxide than will gas that exits later. Exhaled gas can therefore be considered to consist of a continuous set of alveolar gas samples, separated by time and the differences between these “samples” can be used to calculate DLCO.

Several techniques have been developed to calculate DLCOexh. What these techniques have in common is that they all use a relatively standard single-breath DLCO gas mixtures in conjunction with fast responding carbon monoxide and helium or methane gas analyzers. They also require the subject exhale slowly (approximately 0.5 L/sec) after inhaling the DLCO gas mixture to TLC. The primary differences between these techniques lies in the way DLCOexh is calculated.

DLCOexh – Point Sample Technique:

In the study of Newth et al [9] the carbon monoxide and helium gas concentrations were determined for the midpoint of each 10% decrement in lung volume from 20% to 80% of the subjects exhaled volume.

 

For each interval DLCOexh was then calculated from:

 

where:

t₁ = time at start of 10% interval

t₂ = time at end of 10% interval

FA_CO(t1) = CO concentration at start of 10% interval

FA_CO(t2) = CO concentration at end of 10% interval 

and:

 

where:

VI = inspired volume, ml

Vd = dead space (machine + anatomic), ml

FI_He = initial concentration of helium

Fexh_He(50%) = concentration of helium at 50% of exhalation

Vol_exh = volume exhaled from peak inhalation at midpoint of interval

Newth’s study showed that in normal subjects DLCOexh did not change when preceded by different inspiratory flow rates and different breath-hold periods. DLCOexh was, however, constant at all lung volumes only during slow exhalations. 

DLCOexh – Discrete Sample Technique:

In an analysis of the point-sample technique, Graham et al [3] showed that a 1 percent error in the measurement of carbon monoxide could lead to an 8 to 14 percent error in DL_COexh. In light of this they took two approaches towards increasing the accuracy of the DL_COexh calculation.

First, gas concentrations were averaged over larger intervals. After deadspace clearance, the subject’s digitally recorded exhaled volume was divided into both two and three equal intervals. This made the filtering of the exhaled gas concentration signals unnecessary. The average carbon monoxide and methane gas concentrations for each of these intervals were then used to calculate a series of DL_COexh values.

 

Second, they used an iterative procedure for calculating DL_COexh that depends on comparing the ratios of exhaled gas concentrations rather than on the absolute values of the exhaled gas concentrations. A complicated and computationally intensive approach highly suited to computerized analysis, this algorithm starts by first predicting the gas concentration ratios of two gas samples based on an assumed DL_COexh value, which during the first iterative cycle is usually the subject’s predicted DL_COSB. This predicted ratio is compared to the actual ratio and a new DL_COexh value is calculated. This process is repeated until the difference between the new and old DL_COexh values is less than an 0.01 percent, which usually takes about 20 iterations to achieve.

Intrabreath DLCO (Log Slope) Technique:

An alternate approach first proposed by Martonen and Wilson [8], the natural logarithm of exhaled carbon monoxide and of alveolar volume are used to calculate DL_COexh. After exhaling to RV, subjects inhale the DL_CO gas mixture to TLC. After breath-holding for approximately 1 to 2 seconds, the subjects exhale slowly at a constant rate (0.3 to 0.6 L/sec). Gas concentrations measured between 20% and 80% of the exhaled volume are used to calculate DL_COexh as follows:

 

Where:

Ve = exhaled flow rate in liters/second

FA_CO = Fractional concentration of CO

FA_CO0 = Fractional concentration of CO at end-inspiration

VA = Alveolar volume in liters

VA₀ = Alveolar volume in liters at end-inspiration

And:

 

Where:

Vi = inspired volume

Fi_CH4 = Fractional concentration of inhaled methane

Fe_CH4 = Fractional concentration of exhaled methane

Vexh = Volume exhaled from end-inspiration

The slope of the relationship of ln FA_CO/FA_CO0 vs ln VA/VA0 is analyzed through least squares regression analysis and uses data from the beginning of dead space washout to closing volume.

 

A possible advantage of this technique compared to the other more discrete techniques is that essentially the entire exhaled gas stream is used to calculate DLCO although most investigators have excluded data after closing volume has been achieved.

Comparing Single-breath DLCO with DLCOexh:

Comparable values between single-breath DLCO and DLCOexh results have been shown in groups of patients with normal lungs, patients with obstructive lung disease and patients with severe congestive heart failure pre- and post-heart transplant. A closer inspection however, of the scattergrams presented in Wilson’s [13], Quantz’s [10] and Huang’s data [4], however, shows that some individuals had significant differences in results between the two techniques and the reasons for this have not been studied.

One of the most serious criticisms of the DLCOexh technique is that it assumes that the lung fills and empties homogeneously and that the rate of diffusion remains constant over a wide range of lung volumes. Both of these assumptions have been shown to be incorrect by numerous investigators. The same criticism can be leveled at the single-breath technique, but the effect that heterogeneous filling and emptying has differs between the two approaches and this has not been studied other than to compare averaged results.

Newth [9] stated that DLCO did not change as lung volume decreased in patients with normal lungs and this finding is seconded to some degree by Cotton [1], however other researchers [3][5][11][7] have shown that DLCOexh does change, with the most common finding to be an elevated value near TLC and reduced value near RV. This leaves open to question about which lung volume DLCOexh is “most” correct and should be the reported value.

Additionally, alveolar volume (VA) remains a significant component of the DLCOexh calculation and how this is derived differs between specific DLCOexh techniques. When the method used to calculate VA was reported investigators measured either the methane or helium gas concentration at 50% of the exhaled vital capacity [11][12], or from the methane or helium gas concentration at the beginning of exhalation [13][7][2] whereas other studies [5][4] only stated that expired methane was used to calculate VA without saying whether this value was obtained at the beginning or middle of exhalation, or was a mixed-expired value.

Wilson [13] noted that VA was smaller and DLCO/VA higher when using the log slope technique when compared to DLCOSB. This difference was attributed to the fact that the longer breath-hold period of the DLCOSB technique allows for a more complete distribution of methane. Other researchers however, have reported higher VA results from the log slope technique than from the DLCOSB technique and noted that these VA measurements were closer to the TLC measured by plethysmography [6] and by nitrogen washout [10] than the DLCOSB VA. The reason for these discrepancies in reported VA has not been determined.

Finally, a steady expiratory flow rate is necessary for the accuracy of the exhaled DLCO calculations. Where reported several studies have used visual feedback and/or a flow restrictor to enable the patient to do this. The expiratory flow rate of 0.5 L/sec appears to have been empirically chosen by Newth [9] because results at this flow rate were reproducible whereas results from a higher flow rate of 2.0 L/sec were not. Subsequent studies have used this target value but where reported subjects have exhaled anywhere between 0.23 and 1.00 L/sec and the effect of these different expiratory flow rates has not been studied.

Proponents of the DLCOexh technique say that it allows DLCO to be measured during exercise but there are only a small number of studies of DLCOexh during exercise and the actual details of how DLCOexh was performed in these circumstances have been limited. This makes it difficult to assess whether DLCOexh is better or even as valid as other exercise DLCO measurement techniques. As importantly, the fact that expiratory flow must be limited to measure DLCOexh limits its usefulness during exercise, particular in comparison with rebreathing DLCO techniques that do not place any significant constraints on the subject’s breathing pattern.

Overall, the exhaled DLCO technique is an interesting approach to measuring DLCO that appears to be able to show results that are comparable to single-breath DLCO. Although it is potentially useful in some patients who have difficulty with the single-breath DLCO maneuver its clinical relevance is hampered by a lack of standardization and the fact that differences in specific individual’s results when compared to the single-breath DLCO have not been explained.

REFERENCES:

[1] Cotton DJ, Newth CJL, Portner PM, Nadel JA. Measurement of single-breath CO diffusing capacity by continuous rapid CO analysis in man. J Appl Physiol 1979; 46:1149-1156

[2] Felton C, Rose GL, Cassidy SS, Johnson RL. Comparison of Lung Diffusing Capacity during Rebreathing and During Slow Exhalation. Resp Physiol 1981; 43:13-22

[3] Graham BL,Mink JT, Cotton DJ. Dynamic measurements of CO diffusing capacity using discrete samples of alveolar gas. J Appl Physiol 1983; 54:73-79

[4] Huang YC, Helms MJ, MacIntyre NR. Normal values for single exhalation diffusing capacity and pulmonary capillary blood flow in the sitting, supine positions and during mild exercise. Chest 1994; 105:501-508

[5] Huang YCT, O’Brien SR, MacIntyre NR. Intrabreath Diffusing Capacity of the Lung in Healthy Individuals at Rest and During Exercise. Chest 2002; 122:177-185

[6] Kiss D, Popp W, Wagner C, Havelec L, Sertl K. Comparison of the single breath with the intrabreath method for the measurement of the carbon monoxide transfer factor in subjects with and without airways obstruction. Thorax 1995; 50:902-905

[7] MacIntyre NR, Nadel JA. Regional Diffusing Capacity in Normal Lungs during a Slow Exhalation. J Appl Physol 1986; 52:1487-1492

[8] Martonen TB, Wilson AF. Theoretical Basis of single-breath gas absorption tests. J Math Biol 1982; 14:203-220

[9] Newth CJL, Cotton DJ, Nadel JA. Pulmonary Diffusing Capacity Measured at Multiple Intervals During a Single Exhalation in Man. J Appl Phys 1977; 43:617-625

[10] Quantz M, Wilson S, Smith C, Stitt L, Novick R, Ahmad D. Advantages of the Intrabreath Technique as a Measure of Lung Function Before and After Heart Transplantation. Chest 2003; 124:1658-1662

[11] Ramage JE, Coleman RE, MacIntyre NR. Rest and exercise cardiac output and diffusing capacity assessed by a single exhalation of methane, acetylene and carbon monoxide. Chest 1987; 92:44-50

[12] Stokes DL, MacIntyre NR, Nadel JA. Nonlinear increases in diffusing capacity during exercise by seated and supine subjects. J Appl Physiol 1981; 51:858-863

[13] Wilson AF, Hearne J, Brenner M, Alfonso R. Measurement of Transfer Factor during Constant Exhalation. Thorax 1994; 49:1121-1126 

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

When DLCO tests are performed my lab’s standard policy to average two or more results that meet the criteria for quality and reproducibility. It is not unusual for us to perform three DLCO tests and have all of them meet quality criteria but to have one test result that is higher than the other two. Unlike spirometry tests, bigger isn’t necessarily better for DLCO, so in a circumstance like this we will average the two closest results rather than choose the highest result. Even though the higher test results can come from a DLCO test with good quality, I think that reproducibility trumps this and that choosing by reproducibility gives us results that are more clinically reliable.

When I review spirometry results and either lung volumes or a DLCO test has also been performed, I will always check the Slow Vital Capacity(SVC) from the lung volumes and the Inspiratory Volume (Vinsp) from the DLCO test to see if they are larger than the reported Forced Vital Capacity. If either of them is I will manually re-calculate the FEV1/VC ratio to see if it indicates the presence of airway obstruction. This is in line with the ATS-ERS recommendations to use the largest Vital Capacity, regardless of the source, for the FEV1/VC ratio.

I have been reviewing the raw test data for all DLCO tests (as well as all the lung volume tests and regular spot checks on spirometry) performed in my lab for at least the last year. Since our software and hardware upgrade a year and a half ago we’ve found a number of problems that have significant effects on the DLCO test results. Depending on the problem they are capable of causing the results to be over- or under-estimated. All of the technicians performing the tests are now well aware of these problems and there haven’t been any problematic DLCO tests selected for a while. Nevertheless, I always check the raw data just to be sure.

Today, I ran across a report that looked quite straightforward. A set of spirometry and DLCO tests had been performed on a frequent-flier patient with pulmonary fibrosis. The patient has restrictive lung disease and lung volumes measured about a year ago were 64% of predicted. Even though the patient’s FVC has decreased since then there is no clinical reason to repeat the lung volume measurements. The results looked like this: 

	Observed:	%Predicted:	Predicted:
FVC (L):	2.28	51%	4.45
FEV1 (L):	1.64	51%	3.21
FEV1/FVC (%):	72	100%	72
DLCO ml/min/mmHg:	12.03	50%	24.23
Vinsp (L):	2.29

When I looked at the raw DLCO data, what I saw was: 

	Test #1	Test #2
DLCO:	11.99	12.05
VA (L):	3.54	3.58
Volume Inspired (L):	2.09	2.49

Both tests met all ATS-ERS quality criteria (including Vinsp > 90% of the FVC) and they were eminently reproducible, but the second test had a significantly larger Inspired Volume. This fact alone is not a reason to reject the first test. Because the VA was essentially the same for both tests this says to me that the DLCO gas mixture was probably equally well distributed through the patient’s lungs for both tests.

However, when I took the Inspired Volume from the second test and recalculated the FEV1/VC ratio it came out as 65.9 which is 91% of predicted. We use an FEV1/VC ratio of less than 95% of predicted as our criteria for airway obstruction. (I’ve discussed this in the past and we do this partly because we think that it is more accurate than using the LLN and partly because of continuity). Once I saw the higher Vinsp, I de-selected the first DLCO test so that the reported Vinsp was no longer averaged between the two tests and that it was instead reported from just the second test.

Normally I would hesitate to do something like this, but in this case the DLCO and VA results were essentially identical between the two tests so excluding the first test made no significant difference whatsoever in the reported DLCO results. More importantly, the higher Vinsp showed that the patient had an obstructive component to their lung disease as well as a restrictive one.

My lab’s software always reports the highest SVC regardless of which test it comes from but it averages the Inspired Volume. Reporting the highest SVC is in line with the ATS-ERS criteria but the ATS-ERS does not speak to how Vinsp should be reported. Since I believe that averaging DLCO results is the proper thing to do I haven’t previously questioned whether or not Vinsp should be averaged along with DLCO and VA but now I think that the highest Vinsp should be reported instead.

I’ve often suspected that many patients with pulmonary fibrosis also have some airway obstruction as well. I base this on the fact that despite having a significantly reduced vital capacity many of these patients can take a lot longer than 6 seconds to completely exhale their FVC (assuming their dyspnea lets them exhale that long) and this happens even when their FEV1/FVC ratio is normal or elevated. I wonder if other criteria like FEV3/FVC ratio might highlight some of these patients or whether performing SVC maneuvers along with FVC maneuvers to try to get a higher VC is a better approach.

Anybody that reviews and interprets PFT reports needs to be aware how critical numbers like SVC and Vinsp end up on the report.  This is a somewhat unusual case where the results from a DLCO test were used to show airway obstruction. Although I knew that Vinsp was averaged, in this case the care I’ve been taking to review the raw data from DLCO and lung volume tests has paid off because if I hadn’t looked at the raw data for the DLCO test this patient’s airway obstruction would have gone un-reported. 

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

The ATS-ERS standards on lung volume measurements says that measured TLC and RV can be calculated either by

RV = FRC – ERV then TLC = RV + SVC

or by

TLC = FRC + IC then RV = TLC – SVC

with the preference going to the first method. Strictly speaking, given the same FRC and SVC measurements either method is going to end up with exactly the same calculated TLC and RV values. Conceptually speaking I believe that TLC = FRC + IC is a more relevant way to think about TLC but this is only because I think that patients find it easier to perform a quality IC maneuver than a quality ERV maneuver.

A while back I found out that the predicted TLC in my lab’s test systems was being derived from the predicted RV from one set of equations and the predicted FVC from another set of equations (i.e. predicted TLC = predicted RV + predicted FVC). This is likely done so that there will be no discrepancy between the predicted FVC and predicted SVC on reports. I am not sure this is the correct decision since SVC does tend to be slightly larger that FVC but the difference is admittedly small (<1%) in healthy subjects so it is not likely to be significant.

Does it matter, however, for predicted TLC and RV which value’s reference equation you start with and which FVC reference equation you use with them? 

There are, of course, many different reference equations for lung volumes and spirometry, but to keep this simple I will choose the ones that I think are the most common and most relevant. For a 50 year old, 175 cm Caucasian male therefore, the predicted lung volumes look like this:

Equation:	TLC	FRC	RV	SVC
Quanjer	6.90	3.42	2.16	4.74
Crapo	6.74	3.60	1.98	4.76

And the predicted FVCs look like this:

Equation:	FVC:
NHANESIII	4.88
GLFI	4.80
Morris 1988	4.71
Knudsen 1983	4.50

For a 50 year-old, 165 cm Caucasian female the predicted lung volumes look like this: 

Equation:	TLC	FRC	RV	SVC
Quanjer	5.76	2.97	1.97	3.76
Crapo	5.79	3.27	2.03	3.79

And the predicted FVCs look like this: 

Equation:	FVC:
NHANESIII	3.66
GLFI	3.57
Morris 1988	3.41
Knudson 1983	3.28

If you start with the predicted RV, then the range of predicted TLCs (and the percent difference from the original predicted TLC) derived using a predicted FVC look like this: 

Male:	Original:	NHANESIII	GLFI	Morris	Knudson
Quanjer	6.90	7.04 (+2.0%)	6.96 (+0.8%)	6.87 (-0.4%)	6.66 (-3.6%)
Crapo	6.74	6.86 (+1.8%)	6.78 (+0.6%)	6.69 (-0.7%	6.48 (-4.0%)
Female:	Original:	NHANESIII	GLFI	Morris	Knudson
Quanjer	5.76	5.63 (-2.3%)	5.54 (-3.9%)	5.38 (-7.1%)	5.25 (-8.8%)
Crapo	5.79	5.69 (-1.7%)	5.60 (-3.3%)	5.44 (-6.0%)	5.31 (-8.2%)

If you start with predicted TLC, then the range of predicted RVs (and the percent difference from the original predicted RV) derived using a predicted FVC look like this 

Male:	Original:	NHANESIII	GLFI	Morris	Knudson
Quanjer	2.16	2.02 (-6.5%)	2.10 (-2.8%)	2.19 (+1.4%)	2.40 (+11.1%)
Crapo	1.98	1.86 (-6.1%)	1.94 (-2.0%)	2.03 (+2.5%)	2.24 (+13.1%)
Female:	Original:	NHANESIII	GLFI	Morris	Knudson
Quanjer	1.97	2.10 (+6.6%)	2.19 (+11.2%)	2.35 (+19.3%)	2.48 (+25.9%)
Crapo	2.03	2.13 (+4.9%)	2.22 (+19.0%)	2.38 (+17.2%)	2.51 (+23.6%)

So for average males, when comparing the original predicted TLC to the predicted TLCs derived using a predicted RV and FVC the differences range from +0.14 L (+2.0%) to -0.26 L (-4.0%). For average females, the differences range from -0.10 L (-1.7%) to -0.51 L (-8.8%).

For average males, when comparing the original predicted RV to the predicted RVs derived using a predicted TLC and FVC the differences range from -0.08 L (-6.5%) to +0.26 L (+13.1%). For average females, the differences range from +0.10 L (+4.9%) to +0.51 L (+25.9%).

Note: For these examples I used average heights from the more common reference equations. I have noted before that when a subject is at or outside the normal range of heights (63” to 76” for Caucasian males, 57” to 70” for Caucasian females) the differences between reference equations become more pronounced so the results presented are probably at the lower end of what is routinely possible.

In both absolute and a percent difference values, there was less difference in the derived TLC and RV predicted values for males than for females. It does depend however, on which reference equations are paired together and which direction the calculations are going in. The NHANESIII and GLFI reference equations come from much larger populations and underwent a far more significant statistical analysis than the Morris or Knudson reference equation and for these reasons I think that they should the ones used by all PFT labs. They are also the equations that showed the least difference between the original predicted TLC and RV, and the derived TLCs and RVs. There are, of course, valid reasons having to do with continuity and ethnicity why a PFT Lab would choose a different reference equations than these.

Does deriving TLC and RV from separate equations matter? And does it matter which direction the calculations move towards?

For the majority of patients the source of the predicted TLC will likely not matter all that much. It will matter, of course, for patients that are at the lower limit of normal because there even a 1% difference in predicted TLC can change how the results are interpreted. From my own experience I’d say the number of patients who will be affected will only be a few percent of all those who have their lung volume measurements (although that will depend on the selected reference equations) and that the “error” in using a derived TLC is probably smaller than the potential errors within the lung volume test itself.  

Interestingly, although the magnitude of the differences between the original and derived TLCs and the original and derived RVs were similar in liters, they were significantly different when looked at as a percent. For this reason I suppose it actually does make sense to start with the predicted RV and to derive TLC rather than the other way around. Having said that, when lung volumes are measured TLC is far and away the most important and critical result. For this reason alone I think that the measured TLC should be compared to a “real” predicted TLC and not one derived from different reference equations.

I think we all need to be more aware where our predicted values come from. It took some significant digging into my lab’s software to find out how the predicted TLC was being derived. I don’t think this was due to any particular subterfuge on the part of our equipment vendor, just that there are so many reference equations and so many ways to use them that it was never explicitly documented.

Reference equations have always seemed to be an arcane part of our field. Their uses and limitations are often marginally understood and for this reason I think that any inconsistencies in their use are often overlooked. I will be the first to admit that my understanding of statistics is only basic at best. Even so, I believe that any individual responsible for interpreting PFT results should take the time to know which reference equations are being used and how the predicted values are being derived simply because of the implications this has on the interpretation. 

References:

Crapo RO, Morris AH, Clayton PD, Nixon CR. Lung volume in healthy nonsmoking adults. Bull Eur Physiopathol Respir 1983; 18:

Hankinson, JL, Odencrantz JR, Fedan KB. Spiromtric reference values from a sample of the general U.S. Population. Amer J Resp Crit Care, 1999; 159: 179-187.

Knudson RJ, Lebowitz MD, Holberg CJ, Burrows B. Am Rev Resp Dis 1983; 127: 725-734

Morris JF, Koski A, Temple WP, Claremont A, Thomas DR. Fifteen-year interval spirometric evaluation of the Oregon Predictive

Quanjer PH, Stanojevic S, Cole TJ, Baur X, Hall GL, Enright PL. Hankinson JL, Ip MS, Zheng J, Stocks J. Multi-ethnic reference values for spirometry for the 3-95 year age range: the global lung function 2012 equations. Eur Respir J 2012; 40: 1324–1343.

Stocks J, Quanjer PH. Reference values for residual volume, function residual capacity and total lung capacity. Eur Respir J 1995; 8: 492-506.

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

The Hospital Information Systems (HIS) at different medical centers have grown up mostly in isolation from each other. Even when an HIS is installed by a national vendor, each individual hospital has tended to make its own customizations and to follow past conventions. This is changing and it is changing because there are a number of issues driving rapid improvements in inter-hospital communication. The Meaningful Use (MU) Act is major factor and one that has been helping to set the pace, but because improved communication lowers costs and improves the quality of care insurers and medical institutions have been moving in this direction for their own reasons as well.

The regulations and standards for Health Information Exchange (HIE) are evolving rapidly. The overall framework for HIE resides in the Consolidated Clinical Data Architecture (C-CDA) and HL7 messaging protocols. This has given hospitals a unified approach towards managing their communication channels between physicians, clinics, other hospitals and insurers but one problem limiting the usefulness of this has been the different nomenclature used by different institutions for the same pieces of information.

When databases are grown in isolation they tend to end up with labels for data elements that are idiosyncratic and unique to each medical center. There needs to be a way to resolve this Tower of Babel and that is what the Logical Observation Identifiers Names and Codes (LOINC) organization is doing.

The basic idea is simple, and that is to provide a single, unique code for all medical terms and test results. Given just how many terms and result names are in common use around the world this is in reality a monumental task and still a work in progress. Depending on how you want to measure it, probably around 90% of the most common PFT test result names and observations have a LOINC code.

If your PFT Lab has an HIS interface, test data is either stored as a complete report (usually in PDF format), or as individual elements (FVC, FEV1, TLC, DLCO etc.) and occasionally as both. Storing and distributing PFT results as a complete report simplifies the PFT Lab’s HIS interface and does meet the current minimum requirements for HIE but this is likely going to change. The reason for this is that once information is enclosed in a PDF file it cannot be searched and can only be read by a human. The need to perform searches on patient test results by clinicians, insurers, researchers and automated decision systems is eventually going to make interfaces composed of individual data elements mandatory, not just for PFT Labs but for all clinical departments.

This means that sooner or later your HIS will have to store PFT data as individual elements and this is when the LOINC codes will become important. This is also when your knowledge of your lab’s PFT results will help to make sure that your test data ends up with the right LOINC code and that they will make sense when somebody at a different institution tries to read the them.

The LOINC naming system is in some ways a simplistic answer to a complicated problem. For example, we are used to thinking about an FEV1 being performed pre-bronchodilator, post-bronchodilator, post-exercise or as a predicted or percent of predicted value. There are no modifiers for LOINC codes, so there has to be a separate LOINC code for each possible FEV1 value. This means that a lot of LOINC codes are needed to describe even simple PFT test results and that there are still gaps.

The LOINC organization cheerfully acknowledges this problem and makes the ability to submit labels for new LOINC codes easy. New code submissions go into a queue and eventually to a relevant committee. When approved by committee the new code will be added to the next LOINC database update, which come out at regular intervals.

The LOINC database and its associated software RELMA (Regenstrief Institute LOINC Mapping Assistant) can be downloaded for free from the LOINC website. Unfortunately, RELMA is only available for Windows PCs but a CSV file of the database can be downloaded for OS X and Linux computers. The current download zip file for Windows PCs which includes the database and RELMA software is about 370 megabytes.

The need for HIE has also exposed some deficiencies in my PFT Lab’s database. There are a number of tests we perform on a more or less regular basis (6-minute walk, walking oximetry and HAST for example) that can only be entered as text values in the patient’s test notes. There is no place for these results in the database we got from the lab’s equipment vendor and I am not sure how widespread this problem is for other vendor’s databases but I suspect it is a common one. There are LOINC codes for some of these values, but at the moment there is no way to associate them with the actual test results. At the moment any solution we can come up with will be very idiosyncratic and unique to my lab. The databases that all PFT labs get from our equipment vendors are probably the single place we use store all our test results, even when they are not performed using the vendor’s equipment. For this reason I would like to see all of the PFT equipment vendors realize the need for a more comprehensive solution to a PFT Lab’s needs and provide more universal ways of entering these orphan test results.

If (when) your hospital is involved in HIE then your IT department will probably be aware of LOINC codes. Applying them correctly may be a separate issue and I would strongly recommend that all PFT Lab managers be proactive about this problem, both for the need to store PFT test results in their HIS as individual data items and in the need for their associated LOINC codes.

I’ve spent some time searching the LOINC database for PFT Lab-related codes and have placed what I’ve found into a spreadsheet. Feel free to download the Excel version or the LibreOffice version but be aware that I selected codes based on the needs of my PFT Lab and there may well be codes your lab needs that mine doesn’t. Also be aware that in many cases there are multiple LOINC codes for what appears to be the same value so in some cases you’ll just have to take your best guess at what the right one is.

Excel spreadsheet of LOINC PFT Codes

LibreOffice Spreadsheet of LOINC PFT Codes

The opportunities for improved medical care that will come from universal adoption of HIE by all medical centers, clinics, doctor’s offices and researchers are quite exciting and long overdue. The technology necessary for HIE has been around for decades but the costs involved in implementing it as well as a healthy dose of institutional inertia and the NIH (not invented here) syndrome have slowed its adoption. The benefits now greatly outweigh the costs and we need to make sure our PFT Labs will be part of this evolutionary step in information management.

Kudos to the people involved in LOINC, C-CDA, HL7, MU2 and their associated committees and organizations. This is difficult work far from the public eye and very much not the “sexy” part of medicine, and yet it is probably going be responsible for some of the biggest improvements in medical care we will see in the coming decade.

Links:

LOINC:

http://loinc.org/

Meaningful Use:

http://www.cms.gov/Regulations-and-Guidance/Legislation

http://www.himss.org/meaningfuluse?navItemNumber=13303

C-CDA/HL7:

http://hl7book.net/Whitepapers

http://wiki.siframework.org/Companion+Guide+to+Consolidated+CDA+for+MU2 

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

The last several decades has seen a complete transition to the use of computers in pulmonary function testing. This has improved Lab efficiency, but it is also the new baseline. Further improvements in technology may improve the reliability and accuracy of test equipment and test results, but it is unlikely to improve PFT Lab efficiency any more than it already has.

Report management, which is really information management, has started but hasn’t yet completed the same technological transition and it is here that significant improvements can still be made. These improvement will not only improve the efficiency of the pulmonary function lab, but also its clinical effectiveness for the physicians and patients that are the lab’s customers.

To one degree or another most pulmonary function labs are still dominated by traditional reporting systems which are labor intensive and slow. Managing paper reports for a patient visit usually consists of:

• Patient reports are kept in folders and either a new folder needs to be created or the patient’s existing folders need to be pulled from file cabinets.
• Printing the test results and then collating the reports with patient’s lab folder.
• Delivering a stack of reports and lab folders to a reviewer who makes penciled notes on the reports.
• The stack of reports and lab folders is transferred to a typist who types the interpretation into the lab database.
• The final reports are printed, collated with the patient lab folders and stack of lab folders and reports are delivered to the physician who then physically signs each report.
• Reports are photocopied and snail-mailed to the ordering physician and medical records.
• The lab folders are re-filed.

Not every pulmonary function lab still uses all of these steps to manage reports of course, but large parts of this overall process are often still major components in report management. So why are we still moving paper around when what we really want to do is to move the information that’s on the paper around?

Most equipment manufacturers now offer some kind of hospital information system (HIS) interface with their software but what this primarily means is that they are able to communicate using HL7, which is currently closest thing to a universal communications protocol in the medical field. Although the two systems have the potential to be able to talk to each other, it doesn’t mean that you can just plug your PFT system into an ethernet jack and walk away.

An HL7 interface is like a Swiss Army knife and is capable of transmitting many different types of information in a structured manner but there is, as yet, no standard, off-the-shelf interface that will work for all PFT Labs and all hospital information systems. For labs whose hospital’s HIS systems come from a major vendor there may well be a previously configured interface that will save planning and design time. Having said that HIS systems are often highly customized to a specific hospital and automating the report process in any way is going to make significant changes in existing work flows. The implications of these changes and in what is needed to make them work has to be planned for in advance and is most definitely not something to be coped with after the fact.

Planning for an HIS interface is going to require active participation of your hospital’s IT department and your pulmonary function equipment vendor. Much if not most of the technical negotiations will be between them. The hard reality of time, budget, policies and technical limitations will act as overall constraints to any interface but it’s important to not just sit back and let others make all the decisions. The PFT lab and the patients and physicians it serves are the primary clients in this process and determining many of the operational details of the interface is going to be important to ensure that the most efficient use will be made of it. For all these reasons the lab needs to be persistent in advocating for its own interests and needs to be aware of the consequences of planning decisions.

The most significant decision that is going to need to be made is in what form the PFT test results will be transferred in: ie, as an entire report (PDF or similar format) or as individual data elements (age, height, barometric pressure, FVC, FEV1, etc.). This may seem like a simple decision but it has very far-ranging implications for the entire report management process and so needs to be made carefully.

Using an entire report as the primary element in the HIS interface is probably the easiest way to create and maintain an interface, and for that reason it is also probably the most common approach. This only works however when all of the reports the Pulmonary Function lab creates can be made using only the one vendor’s software. If there are any tests or reports at all for which this is not true, then they will not be able to be uploaded into the hospital’s information system.

An interface using individual data elements is going to take longer to design, longer to create and will be harder to maintain but it’s advantage is that it can be tailored very closely to a pulmonary function lab with complex reporting needs.

Note: There are a number of issues driving rapid improvements in inter-hospital communication. The regulations and standards for this are evolving just as rapidly. At the present time the ability to transmit PFT results as a single report is possible and meets minimum requirements but this is likely to change. A major reason for this is that once information is enclosed in a PDF or other image file it cannot be searched and can only be read by a human. The need to perform searches on patient test results by clinicians, insurers, researchers and automated decision systems is eventually going to make interfaces composed of individual data elements mandatory, not just for PFT Labs but for all clinical departments.

Regardless of which approach is chosen, the same information that is in the final, signed version of a report must be in the hospital information system for both legal and clinical reasons.

The next most significant set of decisions is going to revolve around on-line physician signing. Electronic signing of reports is probably the area where the next biggest potential gain in efficiency can be found and this will be for both physicians and the PFT Lab. There are going to be a number of technological, procedural and policy options and limitations to electronic signing and everybody who has a stake in this process (and this definitely means the physicians) needs to be part of the decision-making.

A framework for managing the HIS interface has to be determined. Something or somebody has to decide when reports are transferred from the pulmonary lab network into the hospital information system. Automated transfer may be a built-in property of the equipment manufacturer’s interface but it may be simpler and perhaps safer to have it triggered by a human. For this reason it makes sense that the Pulmonary Function lab controls the interface and this is mainly because it is the lab that knows when it is appropriate to transfer results to the HIS.

Because there is always a time lag between the time tests are performed and the time a final, signed report is available there is much to be said for being able to upload a preliminary report that contains just the test results. This is a policy issue as much as it is a technical issue, however. Since reports in the hospital information system are part of a patient’s official medical record then if there are going to be multiple copies of a patient’s report then there will have to be an audit trail for all reports. This issue will also apply to the correction of any reporting errors. Because maintaining an audit trail and multiple copies of the same report adds a layer of complexity, this may limit the interface to only the final copy of the report.

Finally, the interface must be able detect that an error has occurred in the transfer process. Most commonly these are due to a mistake in patient identifiers. The interface should be able to log transfer errors and transfer successes. There should be a protocol for correcting errors and re-transferring reports. There will also need to be a process for cross-correlating patient visits and reports to ensure that all visits have reports and all reports have been transferred.

An HIS interface can be a major project that will take time and resources to complete but is very worthwhile. It will allow the entire report management process for your PFT Lab to be revamped and improved, leading to a number of long-term benefits.

When the costs of performing pulmonary function testing is analyzed, report management is usually an ignored component. This may be because costs are shifted to clerical staff in another cost center or because they are just considered overhead. Ignored or not, the cost in employee hours exists and the last time I checked, wages continue to rise, not fall. Any automation of the report management process will permanently reduce the number of employee hours spent managing reports.

Timeliness is just as important a benefit. There is a great deal of concern in effective patient care in all areas of the health industry. Slow reporting reduces effective clinical care and increases its long term costs. Timely pulmonary function reporting will be a factor in improving outpatient patient care and decreasing hospitalizations. Admittedly the effects of timeliness are difficult to analyze so the benefits will have to be taken on faith.

A final benefit is that it allows the pulmonary function lab to do more with either the same or fewer resources. Although trivial in the larger scheme of things, if nothing else substantially less money will be spent on paper and printer ink cartridges. Far more importantly however is that the staff time spent shepherding reports will be reduced, leaving more time for testing.  

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

One of the hallmarks of chronic asthma is airway inflammation. This frequently causes an increase in the perfusion of the airways which in turn can appear as an increased DLCO in routine PFTs. A number of investigators have noted that this inflammation can also cause an increase in exhaled air temperature. This increase in exhaled air temperature is not due to an increase in body temperature but to increase in the rate of heat exchange between the airways and respired air due to the increased airway perfusion.

Because the increase in exhaled air temperature also correlates reasonably well with exhaled Nitric Oxide (NO) levels it would seem that measuring exhaled air temperature as part of spirometry or other pulmonary function testing could either act as a substitute for exhaled NO measurements or at least indicate which patients would benefit the most from exhaled NO measurements. It turns out however, that making these measurements is a lot more complicated than it would appear at first glance.

The most important factor that makes exhaled temperature difficult to measure is that it varies throughout exhalation. This has lead to two different approaches to measuring exhaled air temperature. First by measuring the rate at which exhaled air temperature changes during a slow exhalation. Second, by measuring the plateau temperature (PLET) which usually occurs near the end of exhalation and also usually from a slow, controlled exhalation.

One of the early group of researchers noted that exhaled air temperature increased exponentially during exhalation. For mathematical reasons, they determined that the temperature slope from the beginning of exhalation to 63% of the plateau temperature best characterized the rate of increase. In this and related studies exhalation was standardized with a mouth pressure of 10 cm H2O and at a flow rate of 10-11 LPM and results were expressed as degrees centigrade per second.

 

Other researchers have noted more significant differences in plateau temperatures than in the temperature slope although the definition of plateau temperature is open to interpretation. One study defined it as a 2 second period where the exhaled temperature did not change more than 0.5 degree centigrade. Another as the average temperature in the last 20% of exhalation and occurring more than 6 seconds after the start of the maneuver. Several others simply stated it was the “end-expiratory plateau” temperature without any further explanation.

Although there is general agreement that asthmatics have higher exhaled air temperatures than patients with normal lungs there are discrepancies when the results from different studies are compared. Part of the reason for this is that expiratory flow has varied between ~5 and ~10 L/min in different studies when the measurements were made from a single expiratory maneuver. Another reason is that there has been no standardization in the equipment used to measure exhaled air temperature. More than one investigator has noted that both the slope and the plateau temperatures decrease the further the measurements are made from the subject’s mouth. Temperature measurement have been made anywhere from ~1 cm to ~15 cm away from the subject’s mouth but in several studies the distance was not noted.

Another factor is that thermistors and thermocouples have their own thermal mass and therefore take a finite time period to respond to changes in temperature. Researchers from different studies have noted response times of between 1 and 64 milliseconds but it some cases is not clear to me what is meant by response time. The speed of a response to a change in temperature is usually characterized by what is called the time constant and it usually takes a period of approximately 5 time constants to reach at least 95% of the final response. Some studies attempted to carefully characterize their thermisters or thermocouples while other studies just reported the device’s specified “response time” which likely  came from a manufacturer’s data sheet.

Finally there is a certain amount of noise in the temperature signal. This noise can be a significant fraction of degree centigrade. Where noted there were a couple of different approaches used towards averaging this noise but in some cases this signal noise was not even acknowledged.

Study results:

Slope: (degrees centigrade per second)

Study	Normal	Asthma	COPD
D	4.12 +/- 0.41	8.17 +/- 0.83
E	4.00 +/- 0.26		1.86 +/- 0.15
F	4.23 +/- 0.41	7.27 +/- 0.6
H	1.93 +/- 0.21	2.43 +/- 0.23
I	120.90 +/- 95.90	116.43 /- 87.05

PLET (degrees centigrade)

Study	Normal	Asthma	BPD	COPD
A	26.97 (26.58-27.38)	29.60 (29.20-30.02	26.72 (25.11-27.57)
D	34.45 +/- 0.8	35.75 +/- 0.6
E	34.45 +/- 0.8			34.55 +/- 0.6
G		31.1 +/- 0.3
H	27.47 +/- 0.24	30.18 +/- 0.14
I	30.27 +/- 1.25	31.15 +/- 1.19
J	34.84 (32.29-35.84)	35.45 (34.12-36.09)

One very interesting question these results bring up is whether PFT results are being properly corrected for temperature. Spirometer and pneumotach results are routinely corrected using the BTPS factor which assumes that exhaled air is at body temperature (nominally 37 degrees). The maximum exhaled temperature that was observed in all of these studies for normal subjects was less than 35 degrees. Research from several decades ago showed temperatures in the middle bronchi was pretty much at body temperature during normal tidal breathing so the question is why is there a discrepancy of at least 2.5 degrees between this and exhaled air temperature. Although this is probably due in part to heat recovery within the airways during exhalation it is also true that during a forced vital capacity maneuver the rate at which inspired air comes to equilibrium with body temperature depends on the inspiratory flow rate, inspiratory volume and how long the breath is held at TLC before exhaling. Expiratory flow rates can affect the rate of heat recovery by the airways as well. It is possible, therefore, that during an FVC maneuver inspired air does not always reach body temperature before being exhaled.

Even if it is assumed that inhaled air does reach body temperature this also brings into question how different systems correct for BTPS. Some test systems measure the temperature of their spirometer or pneumotach and even perform a “dynamic” BTPS correction but fact is that there are many that do not. I’d be curious what assumptions are being made in some of these test systems in order to make the BTPS correction, particularly since it is apparent that exhaled air temperature varies throughout exhalation.

Researchers have shown that exhaled air temperature is higher in asthmatics than in normal subjects. Exhaled air temperature has been shown to correlate highly with airway perfusion and this has been confirmed by measuring airway perfusion more or less directly. It has also been confirmed by seeing a decrease in exhaled air temperature when asthmatics are treated with inhaled steroids and an increase when normal subjects receive an inhaled bronchodilator. One study also showed, not surprisingly, that exhaled air temperatures in patients with COPD were lower than normal subjects. Exhaled air temperature has also been shown to correlate reasonably well with exhaled Nitric Oxide levels.

Exhaled air temperature has the potential to be a simple and inexpensive screening tool, both in the presumptive diagnosis of asthma and in determining which patients could benefit from nitric oxide measurements. Exhaled air temperature measurements however, lack standardization both in how they are analyzed (slope versus plateau) and in the hardware used to measure it. This makes it difficult to determine the range of normal values and means that it is not yet ready for routine clinical use.

Since an elevated DLCO is often seen in asthmatic patients, I would be interested to see if it also correlates with an elevated exhaled air temperature. I would not be surprised to find that exhaled air temperature, exhaled NO and DLCO all correlated with each other since they are all dependent to one extent or another on airway perfusion.

References:

[A] Carraro S, Piacentin S, Lusiani M, Uayan ZS, Filippone M, Schiavon M, Boner AL, Baraldi E. Exhaled air temperature in children with bronchopulmonary dysplasia. Ped Pulmonlology 2010; article ID PPUL-10-007.R1.

[B] Madan I, Bright P, Miller MR. Expired air temperature at the mouth during a maximal forced expiratory maneuver. Eur Respir J 1993; 6: 1556-1562.

[C] McFadden ER, Denison DM, Waller JF, Assoufi B, Peacock A. Direct recordings of the temperatures in the tracheobronchial tree in normal man. J Clin Invest 1982; 69:700-705.

[D] Paredi P, Kharitonov SA, Barnes PJ. Fasther rise of exhaled breath temperature. A novel marker of airway inflammation? Amer J Respir Crit Care Med 2002; 165: 181-184.

[E] Paredi P, Caramori G, Cramer D, Ward S, Ciacci A, Papi A, Kharitonov SA, Barnes PJ. Slower rise of exhaled breath temperature in chronic obstructive pulmonary disease. Eur Respir J 2003; 21: 439-443.

[F] Paredi P, Kharitonov SA, Barnes PJ. Correlation of exhaled breath temperature with bronchial blood flow in asthma. Respiratory Research 2006; 6:15

[G] Piacentini GL, Bodini A, Zerman L, Costella S, Zanolla L, Peroni DG, Boner AL. Relationship between exhaled air temperature and exhaled nitric oxide in childhood asthma. Eur Respir J 2002; 20: 108-111.

[H] Piacentini GL, Peroni D, Crestani E, Zardini F, Bodini A, Costella S, Boner AL. Exhaled air temperature in asthma: methods and relationshipo with markers of disease. Clin Exp Allergy 2007; 37: 415-419.

[I] Pifferi M, Ragazzo V, Previti A, Pioggia G, Ferro M, Macchia P, Piacentini GL, Boner AL. Exhaled air temperature in asthmatic children: a mathematical evaluation. Pediat Allerg Immunol 2008; DOI: 10.1111|j.1399-2027.2008.00742

[J] Popov TA, Dunev S, Kralimarkova TZ, Kraeva S, Dubuske LM. Evaluation of a simple, potentially individual device for exhaled breath temperature measurement. Respiratory Medicine 2007; 101: 2044-2050.

[K] Popov TA. Human exhaled breath analysis. Ann Allergy Asthma Immunol 2011; 106: 451-456.

[L] Zawadski DK, Lenner KA, McFadden ER. Comparison of intraairway temperature in normal and asthmatic subjects after hyperpnea with cold, cold and ambient air. Amer Rev Resp Dis 1988; 138: 1553-1558.

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

The manager of a nearby PFT Lab made an interesting statement recently, and that was that DLCO measurements made with a sample balloon were superior to those made with a real-time gas analyzer. I think that he is biased to some extent by the fact that the manufacturer of his lab’s testing systems only supports the sample balloon approach to DLCO testing and so that is what he is used to. The same can be said of me however, because all of my lab’s equipment performs DLCO tests using real-time gas analysis and that is what I am used to.

I think that each approach has some benefits and some weaknesses but I’ll start first with a bit of background. When the single-breath DLCO test was standardized in more-or-less it’s current form in the mid-1950’s, the only gas analyzers available at that time were slow. They required relatively large samples of gas (>100 ml) and took over 10 seconds to settle to their final reading. The only way to perform the DLCO test was to capture a sample of alveolar gas and then analyze the entire sample.

In a very simplified view, a sample balloon DLCO system works something like this:

Real-time sampling of exhaled DLCO gas was first attempted in the 1960’s using a mass spectrometer. Because carbon monoxide and nitrogen have essentially the same mass (~28), the CO gas component had to use an isotope of Oxygen, O₁₈. This gas mixture was horrendously expensive (in the late 1970’s my lab bought a large cylinder of 0.1% CO₁₈ in air for a research project and it cost $7000 at that time) which made this approach impractical for routine clinical use. In the early 1990’s a rapid-responding infrared analyzer capable of measuring CO and methane (CH4, a relatively insoluble gas that could take the place of helium) was developed and by the mid-1990’s several equipment manufacturers offered real-time DLCO measurements. 

In a very simplified view, a real-time DLCO system works something like this:

For the majority of patients either system is capable of obtaining accurate DLCO test results. It is for the more marginal patients, those with very severe restrictive or obstructive lung disease, that the sample balloon system shows its limitations. The most significant disadvantage of the sample balloon system is that the washout volume must be set in advance. For patients with small vital capacities it can be difficult to get an alveolar sample that does not contain at least some deadspace gas while at the same time getting a sample volume that is large enough to analyze accurately. The ATS-ERS specifications for DLCO testing states that the washout volume should be reduced to 0.50 L (it’s normally 0.75 – 1.00 L) when the patient’s vital capacity is less then 2.00 liters. I am sure that most test systems will automatically default to the appropriate washout volume but that only applies if a vital capacity test has already been performed. In addition the ATS-ERS specifications also states that if the patient’s vital capacity is less than 1.00 L, then the washout volume can be reduced even further but they do not indicate how much further it could or should be reduced. All of this implies that a technician should be able to override the default washout volume settings but whenever they do it is going to be an educated guess at best and there is nothing in the results of the alveolar gas analysis that will indicate how successful the washout setting was.

Because a real time DLCO system allows a technician to inspect the exhaled gas waveforms, the washout and sample volume can be selected after the test has been performed. This, I believe, allows for more accurate testing on patients with marginal vital capacities. Having said that, all equipment manufacturers that I am familiar with display the exhaled DLCO gas waveforms as concentration versus time and I think this is a mistake. Because there is no indication of washout or sample volume when the results are displayed this way it is possible to mistake a point where the patient slowed or stopped their exhalation as the alveolar plateau or to select to a very small sample volume that may not be truly representative. I personally think that washout and sample volume are more important to DLCO test quality than breath-holding time and for this reason I think that the analyzer results should be displayed as concentration versus exhaled volume.

The fact that the gases in an alveolar balloon are usually well mixed is a possible advantage for that type of system. Real-time systems must depend on software to average results from the alveolar sample and this makes them dependent on whether or not a technician has selected the right alveolar sample and that the equipment manufacturer has developed the correct algorithm for this purpose.

Note: This is an aspect to present-day Pulmonary Function testing where I have a number of concerns. I understand that equipment manufacturers have invested significant resources in developing their software and have every reason to want to keep it proprietary. At the same time, the accuracy of the test results we report depend on this proprietary software and its hidden algorithms. Manufacturers can state that their equipment meets all ATS-ERS standards but we have almost no way to verify this and every time I have raised valid concerns about errors I have had a great deal of difficulty getting an answer. This is in part due to the fact that software has become so complex that finding the engineer or programmer responsible for the algorithm in question or even determining which algorithm is the one in question is difficult. I also know of situations where PFT software was developed by contract programmers who were let go when the project was completed and now there is no one left that knows what’s in the software. Although I believe that all manufacturers should make the critical algorithms in their software open to inspection, they are unlikely to do this on their own. I am not a fan of government regulation but like inspecting drugs and milk for purity this is an area where it may be necessary. Alternatively, the ATS-ERS or a manufacturers group could develop a set of open source algorithms for critical testing functions and allow anybody to adopt them. I am open to suggestions.

The fact that the alveolar sample is mixed inside a sample balloon can also be looked at as a potential disadvantage. A sample balloon will always contain a small amount of machine deadspace gas (probably on the order of 10-20 cc) from the test system’s valves (more if the sample balloon was not completely evacuated before the test). Although the amount deadspace gas is small this will affect the final concentration of the alveolar sample and the degree to which this affects the gas concentrations will depend on the actual sample volume which is usually unknown. Because a real time system continuously samples a patient’s exhaled air there is essentially no machine deadspace to contaminate an alveolar sample.

An advantage to sample balloon systems is that they are primarily mechanical in nature and the gas analyzers are relatively simple. This tends to make them less expensive and probably more physically robust than real-time systems. This is not to say they don’t have their share of problems since they are just as prone to leaks, sticking valves and pump failures as real time systems but if I had to place a test system in a remote location where service was difficult, then an alveolar balloon system may be a better choice.

Many factors go into choosing test systems including ease of use, accuracy, cost and durability, and even then some labs are “locked in” to manufacturers that may not offer this choice in the first place. For me it comes down to the fact that real time DLCO systems are more capable of getting accurate test results on marginal patients and that this reason alone makes them a better choice. There is no reason to believe, however, that either approach produces more accurate results for the majority of patients. It could be argued that the results of extremely marginal patients are questionable regardless of which approach is used to measure their DLCO but I think the number of confounding factors are greatly reduced with real time systems and regardless of the questions I have about the software driving them I believe they are inherently superior.

References:

Brusasco V, Crapo R, Viegi G. ATS/ERS Task Force: Standisation of lung function testing. Standardisation of the single-breath determination of carbon monoxide uptake in the lung. Eur Respir J 2005; 26: 720-735.

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