When TLC, RV and VC don’t add up

I thought I was done with lung volume issues for at least a little while but a short time ago I was reviewing a report from another PFT lab and I ran across something that didn’t seem to make sense. What the report showed was a normal TLC (99% of predicted) with a normal VC (101% of predicted) but the RV was 70% of predicted.

When I took a closer look, it was evident that the predicted VC came from the NHANESIII study and the predicted TLC and RV came from the ERS 1993 Statement. In my PFT lab our equipment manufacturer made the decision to use the predicted RV from whatever source the end-user selected (which in our case is the ERS93 study as well) but to re-calculate the predicted TLC using the predicted FVC, again from whatever source the end-user selected (which in our case was also NHANESIII). What this means is that for my lab:

predicted TLC = predicted VC + predicted RV.

What I saw in the report however, was that the predicted TLC and RV came from the ERS93 study and the predicted VC came from NHANESIII but that meant that:

predicted TLC ≠ predicted VC + predicted RV.

In fact the predicted TLC was almost a half a liter less than if it had been calculated from the predicted RV and predicted VC. What I also saw was that:

predicted TLC ≠ predicted FRC + predicted IC

predicted RV ≠ predicted FRC – predicted ERV

Leaving aside the problems with IC and ERV, this discrepancy brings up some interesting questions. First, is it right to substitute the VC from a set of spirometry reference equations with the VC from a set of lung volume reference equations? Presumably manufacturers and PFT labs do this so there is no confusion about the predicted VC when a report has both spirometry and lung volume results. But the predicted TLC and RV were derived using the VC from the same set of reference equations, which means that if you’re going to insert a VC from a different set of reference equations and if you want everything to add up properly you’re going to have to re-calculate either the predicted TLC or the predicted RV and is that right?

If the predicted VC from the lung volume reference equations is kept (and it was about 8% less than the predicted VC from the NHANESIII reference equation for the individual in the report) then there is going to be a different percent predicted when the VC is the same as the FVC. This means that a VC that was mildly reduced in spirometry can be within normal limits in lung volumes and is that right?

Or, you can leave the predicted TLC and RV alone and plug in the predicted VC without altering anything, but that means that you can have a normal TLC and VC with an abnormal RV (the original problem) or even a normal RV and VC with an abnormal TLC and does that make sense?

[BTW, I still don’t have any idea where the predicted IC and ERV came from. They are different by a couple percent from those calculated from the original reference equations and no matter whether I use the VC from ERS93 or NHANESIII they still don’t add up so at this point I remain somewhat baffled.]

In one sense this is a reference equation problem and is exacerbated by the fact that there are relatively few lung volume reference equations to choose from and all of them are derived from relatively small populations. When I reviewed James Hansen’s textbook on Pulmonary Function testing, he felt there were only three sets of reference equations that are in common use. Of these studies, the largest had 627 participants (Gutierrez et al, plethysmography), the smallest had 176 (ERS93, helium dilution) and the middle one had 245 (Crapo, single breath helium dilution). Compare these to 3975 participants in the NHANESIII study and 74,187 in the GLI Study.

So, the final question is should reference equations be mixed at all? The ATS/ERS standards are mute on this issue so there are no guidelines, which leaves it in the hands of equipment manufacturers and end-users. I think the most important point is that the sub-volumes should always add up properly (i.e., TLC = VC + RV, TLC = FRC + IC and RV = FRC – ERV) and for me this is pretty much non-negotiable. That leaves only two alternatives:

  • Leave the original reference equations for lung volumes alone and accept the fact that the predicted FVC and VC are going to be different. On the minus side this is going to lead to a certain amount of confusion from whoever reads the report. On the plus side it would mean that the ratios between different volumes such as RV/TLC, IC/ERV and RV/VC found in the original study population would remain correct
  • Choose either RV or TLC as the “real” lung volume and re-calculate the other lung volumes using the new VC accordingly. A significant question about this approach is what then happens to the other lung volumes? For example, if TLC is recalculated from RV and a new VC, then the predicted RV/TLC ratio decreases. If FRC is unchanged, then the predicted IC will also increase.

RV_Stable_VC_Changes

What’s wrong with this is that numerous investigators have shown that the RV/TLC ratio is primarily dependent on age and is independent of the actual TLC so the new RV/TLC ratio is incorrect. The relationship between FRC, IC and TLC is more complex and we don’t use an FRC/TLC ratio or an IC/TLC ratio, but once you’ve “corrected” the TLC these relationships have also changed and are now different from that of the original reference equation. So the real question is that once you start “correcting” lung volumes with a new VC, where do you stop?

It seems to me that the least number of problems are caused by leaving the lung volume reference equations alone and accepting the fact that the predicted VC and the predicted FVC may well be different. Having different predicteds for VC and FVC is what seems to bother people the most however, and I think this points to a lack of understanding of reference equations.

Reference equations for spirometry, lung volumes and DLCO are (with only exceedingly rare exceptions) always derived from different populations, in different locations, using different types of equipment, at different times and are statistically analyzed in different ways. Trying to dovetail these different reference equations into a coherent whole is an ongoing issue for all PFT Labs whether they realize it or not (and I suspect that many don’t). The real solution would be to have reference equations for spirometry, lung volumes and DLCO derived from the same (diverse and very large) population but this isn’t going to happen anytime soon, so the final answer is that we need to continue to be aware of the limitations of the reference equations we use.

The lab the report came from was recently completely re-equipped from the ground up with new equipment and software from a different manufacturer than they had previously so the report and reference equations date at least from the changeover. I know that the lab’s previous reference equations were transferred to the new equipment and it is possible that an error was made along the way and that these aren’t the default settings of the software. But that also means that either by error or intent it is possible to make reference equation settings that don’t add up. This fact should be taken as a reminder that whenever new equipment or new software (even if it is just a new version of existing software) is acquired that you shouldn’t assume that everything adds up and that results always need to be scrutinized carefully.

My biggest concern about this situation is that I think that we all expect, with good reason, for lung volumes to add up correctly. I know that if I hadn’t noticed the discrepancy between the percent predicted VC, TLC and RV it would never have occurred to me that it was necessary to check the math of the predicted values. In this case, for reasons that remain unclear, none of the predicted lung volumes in the report added up correctly and this makes interpreting the results more than a bit problematic.

Manufacturers and end-users are making decisions about VC reference equations and their use with lung volume reference equations in whatever way seems to be right to them. This issue has a critical bearing on the interpretation of lung volume measurements and badly needs to be addressed in the next ATS/ERS statement on interpretation.

References:

Crapo RO, Morris AH, Clayton PD, Nixon CR. Lung volume in healthy nonsmoking adults. Bull Eur Physiopathol Respir 1983; 18: 419-425

Gutierrez C, et al. Reference values of pulmonary function tests for Canadian caucasians. Can Respir J 2004; 6: 414-424.

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

Hansen JE. Pulmonary Function Testing and Interpretation. Published by Jaypee Brothers Medical Publishers, 2011.

Quanjer PH, Tammeling GJ, Cotes JE, Pedersen OF, Peslin R, Yernault J-C. Lung volumes and forced ventilatory flows. Report working party standardization of lung function tests European community for steel and coal. Official statement of the European Respiratory Society. Eur Respir J 1993; 6: Supplement 16, 5-40.

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

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.