Using spirometry to monitor a patient’s condition over time is often more important and more useful than an initial diagnosis.  For this reason when spirometry results are reported, they should always include Trends.

Trends are a patient’s prior pulmonary function results and are usually reported as a table, and occasionally graphed. A Trend report could look like this:

Date:	FVC(L):	%Pred FVC:	FEV1(L):	%Pred FEV1:	FEV1/FVC:
8/16/2016	2.83	90%	1.34	55%	47%
5/2/2016	2.71	86%	1.30	54%	48%
4/26/2016	2.25	72%	1.23	51%	55%
1/15/2016	2.44	78%	1.21	50%	50%

but there are no particular guidelines either for which results should be trended or how they should be reported. For this reason individual institutions and reviewers will need to decide what should be included in Trends.  As a general recommendation however, for the sake of clarity only the most pertinent results should be included.

Using Trends to assess changes over time:

A reviewer should, at a minimum, comment on any changes, or lack of them, as part of an interpretation.  The ATS/ERS guidelines for interpretation state that:

“..in subjects with relatively ‘‘normal’’ lung function, year-to-year changes in FEV1 over 1 yr should exceed 15% before confidence can be given to the opinion that a clinically meaningful change has occurred.”

They also state that:

“… short-term changes of 12% and 0.2 L in the FEV1 are usually statistically significant and may be clinically important.

They also otherwise indicate that more or less the same level of significance applies to the FVC but do not have any particular guidelines for the FEV1/FVC ratio.

A key point in the first statement is that applies to patients with reasonably normal spirometry and to year-to-year changes.  This is because this statement has it roots in occupational (health surveillance) spirometry but for this same reason it’s applicability to patients with significant lung disease is unclear.

The second statement is more pertinent and more generally applicable, but it is based on the same logic that was used to determine the significance level of post-bronchodilator changes. For a variety of reasons this may well be the correct approach, but it should be remembered that neither statement has been verified with longitudinal studies.

Finally, these statements apply only to the comparison of results separated by a relatively short time period.  When results are compared over significantly longer time periods (>5 years) then comparing changes in the percent predicted of the results is likely more relevant than comparing changes in the results themselves.

Recommendation:

Based on the ATS/ERS guidelines a change of +/- 12% and +/- 0.20 L in FVC or FEV1 within a period of <1 year should be considered significant. Changes of +/- 15% in FVC or FEV1 within a period of >1 year (with a possible maximum of 5 years) should probably also be considered significant.  For longer periods, a relative change in the FVC or FEV1 percent predicted (i.e., a change from 50% of predicted to 40% of predicted is a 20% change) of +/- 15% should likely be considered significant.

It should be noted that the ATS/ERS interpretation guidelines include more nuanced recommendations based both on time periods and lung disorders (normal versus COPD) however there are some inconsistencies in their published values.  More importantly they are based on the Coefficient of Repeatability (CR) from a limited number of studies and the guidelines explicitly recommend that any pulmonary function lab determines its own CR and substitute it.  This is well beyond the ability of most PFT Labs and attempting to adopt these recommendations adds a degree of complexity that is likely counter-productive.

Using Trends as an aid to interpretation:

When results from other pulmonary function tests, lung volume measurements in particular, are included in Trends they can serve to rule-in or rule-out certain findings.  Prior lung volume measurements that are within normal limits can indicate that a reduced FVC is more likely due to some other cause than restriction.

When only spirometry results are available however, changes over time must be assessed conditionally. A previously normal FVC in a patient with airway obstruction suggests that a reduced FVC is more likely due to gas trapping than it is to restriction but this is not definitive. When the Trended FVC is always reduced however, even if it the cause is obstructive gas trapping, a diagnosis of restriction can never be excluded.

Patterns within Trends can help assess a patient’s treatment plan or the course of their disorder but it is important to remember that trended results are limited to individual tests scattered over prolonged intervals.  What occurs within these intervals is truly unknown and for this reason changes in treatment plans should not be made solely on the basis of the apparent patterns in Trends.

It is also for this reason that graphed trends can be particularly misleading.  A common practice in graphed trends is to join the results with a line, but this line tends to strongly imply that the transition between tests is smooth, and this is often not the case.

Recommendation:

Trended results should be reported numerically and not graphically.  Prior results can help rule-in or rule-out specific diagnoses but the patterns seen in trends should be considered conditional and never definitive.

 

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