I had finished reviewing a pre- and post-BD spirometry report yesterday and was about to toss it on my out pile when I noticed something a bit odd about the post-BD results. I pulled it back and spent some time trying to decide if the interpretation needed to be changed but after a lot of internal debate I finally let it go as it was. I’ve continued to think about it however, and although I’m not sure that was the right decision I still haven’t come up with a clear answer.
Here’s what I saw:
| Observed: | %Predicted: | Post-BD: | %Predicted: | %Change: | |
| FVC: | 3.70 | 97% | 3.91 | 103% | +6% |
| FEV1: | 2.82 | 94% | 2.79 | 93% | -1% |
| FEV1/FVC: | 76 | 95% | 71 | 89% | -6% |
| PEF: | 6.62 | 94% | 7.19 | 102% | +9% |
| Exp. Time: | 10.92 | 11.15 |
The reported pre-BD and post-BD results were from good quality tests and met the criteria for repeatability. My problem is that the baseline results were normal but if I had seen the post-BD results by themselves I would have considered them to show mild airway obstruction.
So what’s going on here? The results from baseline spirometry aren’t anywhere close to the borderline for airway obstruction, no matter which guidelines you use. There was a small post-BD increase in FVC (+6%) but since the ATS/ERS standards for a significant post-bronchodilator change in FEV1 or FVC is ≥12% and ≥200 ml it isn’t significant. The post-BD decrease in FEV1 was also small (-1%) and well within normal test-to-test variability, but when you put the small increase in FVC together with the small decrease in FEV1 suddenly the FEV1/FVC ratio below the LLN.
Does that mean these results show airway obstruction? One argument in favor of this is that the GOLD standards for spirometry only consider the post-BD results. The basic point of this as I see it is that since COPD is expected to have little or no bronchodilator response, this allows asthma to be ruled out as the cause of obstruction. But obstruction was only in the post-BD results, not the baseline results, so I’m not sure this applies.
The ATS/ERS interpretation algorithm also only looks at the baseline results when assessing results for obstruction and only assesses post-BD results for significant improvement.
Could the decrease in the FEV1/FVC ratio be a side-effect of albuterol? We usually use albuterol as our bronchodilator and some individuals are sensitive to it and bronchoconstrict instead of bronchodilating (read the product insert and you’ll find that this is one of the possible side-effects). This occurs in probably less than 1% of the population but I see it a couple times a year. In this case though, even though there was a small post-BD decrease in FEV1 there was a larger increase in FVC without any significant change in expiratory time, and this means there was probably more bronchodilation than bronchoconstriction.
One final point is that post-BD changes are only assessed using the FEV1 and FVC, not the FEV1/FVC ratio. If the pre-BD and post-BD results were reversed, even though the individual would have gone from mild airway obstruction to normal, the bronchodilator response still wouldn’t have been called a significant.
This situation doesn’t seem to be covered by any of the guidelines. On the one hand, if the pre-BD and post-BD results were reversed I’d have no hesitation about saying the results showed mild airway obstruction. On the other hand, the baseline spirometry was normal, the decrease in FEV1 was well within normal test-to-test variability and the increase in FVC was not significant. In the end I had to let it go as normal spirometry mostly because I couldn’t find any way to legitimately use the post-BD FEV1/FVC ratio. That doesn’t mean I haven’t second-guessed that decision a half dozen times since then, but since I keep second-guessing the second-guessing, I’ve let it stand.
Even though I couldn’t come up with any clarity for this situation what this does show is the effect that FVC can have on the FEV1/FVC ratio. When individuals respond to a bronchodilator it’s usually the FEV1 that increases and any increases in FVC are relatively small. Post-BD increases in FVC without any increase in FEV1 are relatively rare and when seen it’s almost always in individuals who’s baseline shows severe airway obstruction. When individuals have post-BD bronchoconstriction it’s usually fairly clear and usually both the FVC and FEV1 are affected. In this case however, the baseline spirometry was normal and the post-BD changes were just ambiguous enough that I don’t think there ever could be a clear answer.
PFT Blog by Richard Johnston is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License
Great example for discussion. It points out some of the limitations of interpreting tests results. When we interpret results, we are always subject to the possibility of making a statistical error.
As you know, in statistical hypothesis testing, a type I error is the incorrect rejection of a true null hypothesis (a “false positive”), while a type II error is incorrectly retaining a false null hypothesis (a “false negative”)type 1 error. In my opinion, we need to be honest and recognize that our cut offs for “normal” are arbitrary. We would do well honoring that in the wording of every interpretation.
For example, these results could be honestly interpreted as “mild reversible expiratory airflow obstruction may be present” or “mild reversible expiratory airflow obstruction cannot be excluded”. The wording is honest and probably more informative than “normal”.
What you have found is pretty common. A few points:
1. The ATS criteria of 12% and 200 ml suffer from the intrinsic defect of judging the same absolute increase as significant and not significant depending on the baseline. Thus, an increase in FEV1 of 300 ml from 3L baseline is to be interpreted as nonsignificant even though 300 ml increase is substantial and way above the permitted repeatability criteria and thus represent a true bronchodilator response. Using the percent increase alone is also problematic. 100 ml increase over a baseline of 500ml is 20% increase but not significant. I have a paper on expression on bronchodilator response in Respiratory Medicine where I have argued that 200 ml increase is good enough. Thus, this patient had a significant increase in FVC representing a bronchodilator response.
2. FVC response is seen frequently in COPD without changing FEV1. This is a volume response. Some very severe asthmatics also show this. I have a paper on this in Indian Journal of Chest Diseases. The ATS-ETS guidelines also recognize that the response to bronchodilator may be in FVC or FEV1. In this case, the FEF25-75 would have decreased.
3. It is fallacious to assume that lung function should show obstruction to have a bronchodilator response. I frequently see large increases even when pre-bronchodilator test in apparently normal by the algorithms. This is because the ceiling for a person is not the predicted value but the “personal best” that can be higher than the predicted value. An asthmatic’s FEV1 may increase from 90% to 110%. Again, the 200 ml increase is good enough to call this good response to bronchodilator.
4. The use of the term “reversibility” should be discouraged. I call it “a significant response to bronchodilator”. That takes care of all these odd situations. I am sure your subject had asthma.
Dr. Chhabra –
The ATS/ERS guidelines of 12% and 200 ml have been criticized numerous times. In particular it has been noted that they (and for that matter the criteria for FVC and FEV1 reproducibility) are biased towards higher values by a small number of subjects whose spirometry has a high degree of variability and that smaller increases would serve to demonstrate a significant bronchodilator response. To some extent I would agree with this but also have to point out that the validity of a lower threshold hinges on optimal pre- and post-BD test quality and for at least this reason it cannot be universally applicable. You could exclude subjects with less than optimal test quality from a lower threshold but then you have to define (and justify) the inclusion/exclusion criteria and the threshold used on those who are excluded. If this tiered approach fits the reality of routine spirometry testing better than a single threshold it also means that it’s not possible to make blanket statements about what constitutes a significant response.
There is likely a great deal of crossover between individuals with COPD whose FVC increases significantly post-BD and those whose IC also increases significantly post-BD. But a discussion of post-BD increases in FVC cannot occur without also discussing increases in expiratory time. When the FVC and the expiratory time both increase significantly post-BD, is the increase really significant? Maybe yes, and at least a couple of studies have indicated that a post-BD increase in expiratory time is in itself significant, but when routine spirometry is considered I’d have to say this is a lot less clear. For this reason I’d suggest that increases in FEV6, which are not tied to expiratory time, should be considered instead of FVC.
A phrase that seems be making the rounds a lot lately is that “extraordinary claims require extraordinary evidence”. A diagnosis of asthma may not be extraordinary but it certainly has significant implications in an individual’s clinical management and therefore requires significant evidence. There are many individuals for whom a post-BD increase in FEV1 or FVC (regardless of how you define the criteria) is either markedly significant or markedly absent. There are also many individuals whose response is equivocal and I believe the case I cited falls within this category and would therefore hesitate to diagnose asthma with any certainty.
One final point and that is that I tend to review all of the pre- and post-BD spirometry efforts made by an individual patient and often see that the reported post-BD increase depends on which specific results were selected. The ATS/ERS criteria states that the largest FEV1 and FVC be reported regardless of which effort they came from (assuming they have adequate test quality, of course), but we (and many other labs) also base the selection of FEV1 on the peak expiratory flow as well. We believe this is the correct approach but it also means that given the exact same set of results a lab that does not use PEF in the selection process could report a completely different outcome. Like it or not there is an element of subjectivity in the selection process, particularly when test quality is suboptimal in any way, and this also has a bearing on the significance of any post-BD changes.
Regards, Richard