I have been taking a close look at the raw data from all lung volume tests lately in large part because N2 washouts are still relatively new to my PFT Lab and we’re continuing to learn from our mistakes. When I saw this N2 washout test I knew that there was something wrong with it. The patient had performed three N2 washout tests and the TLC, FRC and RV for this one test were significantly larger than for the other tests. The most common problem we’ve been having with N2 washouts has been with patient leaks during the washout period which almost always show up as an upwards drift in the tidal baseline. This test did not show any drift however, and it took me a little while before I could see what was wrong with it.

The N2 washout maneuver has the patient start by breathing tidally for a short period of time in order to determine where end-exhalation (FRC) is located. The patient then performs a slow vital capacity maneuver by steadily inhaling maximally to TLC and then exhaling maximally to RV. The technician then switches the patient into the washout breathing circuit at maximal exhalation and the patient resumes breathing tidally for the remainder of the test.

The key assumption in this procedure is that the patient is switched at their maximal exhalation and that all of the nitrogen that is subsequently washed out comes solely from the patient’s RV. TLC is then calculated by adding the SVC volume to the measured RV. FRC is calculated by adding ERV to RV.

In this case the patient paused during the maximal exhalation portion of the SVC and the technician switched them into the washout breathing circuit. The patient however, had not really reached RV and with the next breath actually exhaled past the point at which the switch-in occurred. The test system software then altered the volume position of RV according to this more maximal exhalation. When the test was completed the RV volume was then calculated from the amount of nitrogen that had been washed out, but this volume was actually the point where the patient was switched-in and not RV. RV was therefore overestimated and because of this, TLC and FRC were also overestimated.

This is not really a technician error or a patient error, but is instead a software error. The software assumes that switch-in occurs at RV but then allows RV to be determined from a maximal exhalation occurring at any point during the entire test. If RV can be set at any time during the test then there needs to be a distinction made between switch-in volume and RV just like the one between TGV and FRC in plethysmography.

This is a moderately unusual error because it required two different things to go wrong. First the patient did not exhale maximally and was switched-in to the washout circuit at a volume above RV. Second, and more importantly, they performed a maximal exhalation during the washout period when they should have been breathing tidally. Without this second part even though RV would have been overestimated, TLC and FRC would probably have been reasonably accurate.

I have notified the manufacturer and hopefully this will be fixed in the next software release we receive but until that time (and probably even after) I will continue to review the raw data for all lung volume tests and will be on the lookout for the next time it occurs. This problem may be specific to the brand of equipment we are using or may be to the version of software we are using but I would recommend that if you use N2 washout to measure lung volumes that you try this yourself to see if you get a similar error.

UPDATE (11/30/2012):  I have talked with the manufacturer and they claim that they do calculate the “switch-in” volume.  FRC is calculated from the initial exhalation to the “switch-in” volume and TLC and RV are then calculated from FRC, and not from presuming the “switch-in” volume is RV.  “Switch-in” volume is not reported and the way that lung volumes are calculated is not documented in their manual.  The overestimation of TLC may have been caused by patient leak or other problems.

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