An IC shows it’s probably not restriction

For the last couple of years it seems that I’ve had more problems than usual with lung volume tests. Even though this seems to date from the time that my lab went through its hardware and software upgrade and we started performing N2 washouts I suspect that these problems have been around for a long time and these events just heightened my awareness of lung volume testing problems.

My lab performs helium dilution, N2 washout and plethysmographic lung volume tests. When you are assessing the quality of lung volume tests the first problem for the helium dilution and plethysmographic techniques is whether or not the Functional Residual Capacity (FRC) was accurately measured and for N2 washout, it’s whether or not the Residual Volume (RV) was accurately measured. Leaks are always an issue for any of these measurement techniques and for helium dilution and N2 washout leaks will almost always cause the Total Lung Capacity (TLC) to be overestimated. For plethysmography the picture is less clear since leaks can cause TLC to be either over- or under-estimated.

Once you accept that the initial measurement of FRC or RV is accurate, however, the next question is whether the SVC is accurate or not. Since SVC is a more relaxed test than a forced vital capacity the SVC volume should be at least the same as the FVC volume and it is often larger. When I see an SVC that is smaller than the FVC I tend to think that the calculated TLC is probably okay and it’s the RV that is more likely to be overestimated. This is because the Inspiratory Capacity (IC) part of the SVC maneuver (“take as deep breath in as you can!”) is the easiest part and when the SVC is low, it is usually because the Expiratory Reserve Volume (ERV, “blow everything out that you can!”) is underestimated.

This report came across my desk a couple of days ago. The lung volumes were performed by helium dilution.

Not_RVD_Results 

At first glance the reduced TLC would seem to indicate the patient has restrictive lung disease but there were a number of things that didn’t quite add up. First, the FVC was reasonably normal. As restrictive diseases progress it’s not uncommon for the TLC to decrease more than the FVC but the difference in percent predicted between the two tests was more than I usually see.

Second, the SVC was significantly smaller than the FVC. If the SVC was reduced because the ERV portion was underestimated, then the RV should be elevated, but in this case the percent predicted RV was reduced even more than the TLC.

When I reviewed the raw helium dilution test data, two tests had been performed and the FRC was almost identical for both tests. The SVC was quite different for both tests however, and the reported SVC was the larger of the two and had the largest IC and ERV as well.

Next I reviewed the patient’s FVC tests and found that there was a lot of variability in FVC, FEV1 and the flow-volume loops. The reported test however, was the best of all the tests the patient had performed both in terms of volume and quality. The FVC test module does not calculate IC or ERV (and in fact I don’t know of any FVC test module that does) but when I looked at the position of the tidal loop within the overall flow-volume loop, it looked like the IC was larger than the one reported from the SVC. In order to confirm this, I transferred the flow-volume loop to a graphics program that has a ruler function and I was able to measure the IC from the FVC as being 2.46 liters.

Not_RVD_w_IC 

Assuming that the IC from the flow-volume loop was correct, the TLC was actually 4.97 liters and 84% of predicted, and that puts it within the normal range.

A simple solution would have been to manually enter a new SVC using the FVC volume and if this was done it would have bring the calculated TLC to 4.87 liters and 82% of predicted (the reason it’s not 4.97 liters has to do with the difference in the assumed ERV). The problem with doing this is that ERV and IC do matter and that this would mix the ERV from one test with the VC from another. This is the same reason I believe that using linked maneuvers (the FRC and SVC from the same test) is much more accurate than measuring FRC and SVC separately and combining the “best” results.

The other problem is that the FVC test is not designed to measure IC and ERV and I used the second tidal loop from the last to measure the IC. It looked like the “best” tidal loop to me but there were other tidal loops and I would have gotten a different IC depending on which I selected. There may also be reasons why the tidal loops from this FVC don’t accurately reflect the patient’s normal end-exhalation FRC and it’s even remotely possible that the IC that was markedly overestimated and the ERV was markedly underestimated.

Although manually entering the FVC volume would probably provide more accurate results than those calculated using the SVC, for all these reasons I couldn’t guarantee that this was actually the case. Rightly or wrongly the best I felt I could do was to add a note that the lung volume test quality was suboptimal and the TLC was likely underestimated, and otherwise left the results alone (although I may re-think this the next time I see this kind of situation). I also emailed the ordering physician (one of the Pulmonary physicians I work with) and alerted them to this.

Errors with helium dilution lung tests usually cause the TLC to be overestimated so when a TLC is reduced I tend to think the test is going to be relatively accurate. Likewise, when a SVC is reduced I tend to think it is the IC that is going to be more accurate than the ERV. In this case the evidence points in the other direction and that makes this is a useful reminder that you always need to be skeptical when you review lung volume test results.

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

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.