Flow-volume loops are timeless

Recently I’ve been trying to help somebody whose spirometry results changed drastically depending on where their tests were performed. When their spirometry was performed on an office spirometer their FVC was less than 60% of predicted and when they were performed in a PFT lab on a multi-purpose test system their FVC was closer to 90% of predicted. Part of the reason for this was that different predicted equations are being used in each location but even so there was about a 1.5 liter difference in FVC.

One important clue is that the reports from the office spirometer showed an expiratory time of around 2 to 2-1/2 seconds while the reports from the PFT lab showed expiratory times from 9 to 12 seconds. The reports from both locations however, only had flow-volume loops and reported expiratory time numerically. There were no volume-time curves so it isn’t possible to verify that the spirometry being performed at either location was measuring time correctly or to say much about test quality.

The shape of a flow-volume loop is often quite diagnostic and many lung disorders are associated with very distinct and specific contours. Volume-time curves, on the other hand, are very old-school and are the original way that spirometry was recorded. The contours of volume-time curves are not terribly diagnostic or distinctive and I suspect they are often included as a report option more because of tradition than any thing else. But volume-time curves are actually a critically important tool for assessing the quality of spirometry and one of the most important reasons for this is because there is no time in a flow-volume loop.

With this in mind, the following flow-volume loop came across my desk yesterday. The FVC, FEV1 and FEV1/FVC ratio were all normal and it was the best of the patient’s efforts.

fvl_timeless

The contour of this flow-volume loop is actually reasonably normal, except possibly for the little blip at the end.

v-t_timeless

That “little blip” however, was actually two expiratory pauses, one about a second long and one about 4 seconds long. Although the fact that expiratory flow was zero during these pauses does show up on the flow-volume loop, there is absolutely nothing in the flow-volume loop that indicates how long they occurred.

When the pauses are subtracted the total expiratory time in this spirometry effort was only about 2-1/2 seconds and it was performed by an individual that was about 60 years old. When somebody is in the 20’s and I see a short expiratory time I’m usually not terribly concerned that the FVC is being underestimated (unless of course the FVC is below normal and the expiratory time is really, really short) and that is because it is fairly normal for somebody in that age range to be able to completely exhale their FVC in a short time. Somebody in their 60’s? Not so much. So even though the reported expiratory time was about 7-1/2 seconds there’s a good chance that the FVC is actually underestimated but you wouldn’t know it if all you had was the flow-volume loop.

Volume-time curves have have not changed significantly since they were first introduced about 65 years ago but there’s hardly anybody left that even knows how to get spirometry results off of a volume-time curve. Keeping a volume-time curve in a report just for backwards compatibility doesn’t seem to make a lot of sense but if you view it as an aid to assessing test quality it actually makes very good sense. If we’re going to embrace volume-time curves for this purpose then it’s also probably time to update them as well.

In the first place I think that all of the test data, from the very beginning of the test to the very end of the test, should be shown.

vt-curve

But I also think that showing flow as well as volume may turn out to be useful (although I’ll be honest and say that’s not as clear as I’d like it to be). Flow-time curves were seriously considered at one time, but like volume-time curves their contour is not particularly diagnostic. When used to assess test quality however, it’s possible they may be able to add something. A Volume-Flow-Time (VFT) curve would look like this:

vft-curve

When all of the test information is displayed this way it’s notable that the tidal breath baseline at the beginning of the VT curve appears to be drifting and this is an indication that the patient or the test system was leaking air. The prolonged pauses during exhalation may therefore be at least partly due to a leak and for this reason the FVC is quite likely underestimated. And if the FVC is underestimated then the FEV1/FVC ratio is also probably overestimated and maybe this spirometry effort isn’t as normal as it appears to be.

It wasn’t possible to see this happening in the original volume-time curve however because our test system’s software always truncates everything up until 1 second before the beginning of the FVC exhalation (and some systems truncate everything up until the start of exhalation). This is unfortunately true both when reviewing a spirometry effort and when it is reported (all of the volume-time data is on the computer display during the test but as soon as the test ends the display goes into review mode and all of the extra data disappears). If you go back to the original flow-volume loop the tidal flow-volume loops appear to be drifting, but this could be from other causes and isn’t as clear as what the full VT curve shows.

Note: BTW, none of the flow-volume loops from the person I’ve been trying to help from either the office spirometry system or PFT lab system showed tidal flow-volume loops. Without tidal loops it isn’t possible for a flow-volume loop to show any drift (nor is it possible to show negative effort dependence or to give any idea what the patient’s IC and ERV were).

When I started doing pulmonary function testing in the early 1970’s the only way to record spirometry was with a volume-time curve of some kind (kymograph paper and a pen of one kind or another usually). At that time it was also the only way to measure FVC and FEV1 (manually with a ruler and a calculator). But keeping volume-time curves on a report just so that in a pinch we could still make measurements from them doesn’t make any sense. The contour of a volume-time curve is also not particularly diagnostic in any way so if these were the only reasons that a volume-time curve needed to be on a report I would agree that the space it’s taking up could be better used by something else. But a volume-time curve actually contains a wealth of information about test performance that is not present in a flow-volume loop or in the numerical results. Any spirometry report that does not include a volume-time curve is therefore missing critical information about test quality.

Since volume-time curves are no longer needed for their original purpose that also means the should be optimized for quality assessment. For this reason I’d like to see the next ATS/ERS spirometry standards make it a requirement for complete volume-time (or volume-flow-time?) curves to be included on spirometry reports.

As a final note, when I questioned the person I’ve been trying to help about the short expiratory times that always showed up when their spirometry was performed on an office spirometer I was told that their exhalation was always stopped early by the staff member performing the test. The person said that they told the staff member that they were still exhaling but were always told in return that the spirometer said that no air was coming out, so the exhalation didn’t need to continue. This says to me there is either a problem with the spirometer or with the training of the staff member performing the test and that this is most probably the reason for the different in FVC volumes. What was also interesting about this is that even though the expiratory time was reported, and was always between 2 and 2.5 seconds, the physician interpreting the test always wrote “moderate restriction” (from a spirometry result?) and never once commented on the short expiratory time. So it’s not only the staff member that probably needs better training, it’s the physician (a pulmonologist!) who needs it too.

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

2 thoughts on “Flow-volume loops are timeless

  1. Brings back all sorts of memories: switching to the red pen on the kymo for the post-bd effort, etc. Peak flow was a guess, being the fef200-1200, and back at the first Snowbird meeting of the ITS the 200ml was the basis of the talk about extrapolated volumes.
    Long ago, in a galaxy far, far away.

    cheers,
    Mike

    • You’re forgetting to mention the mechanical MVV accumulator on the pulley for the spirometer/counterweight chain with the separate pen. Also watching the chain go slack when somebody gave a real good FVC effort and the spirometer bell rose faster than the counterweight/pen could drop. Fortunately, I had a Wright peak flow meter and didn’t have to try to get peak flow off the paper. On the one hand we had all the test data on the kymograph paper in front of us, but the fact that the equipment itself biased the results leaves me less than nostalgic for the “good old days”.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.