Definition of COPD: based on evidence or opinion?

R. Pellegrino, V. Brusasco, G. Viegi, R. O. Crapo, F. Burgos, R. Casaburi, A. Coates, C. P. M. van der Grinten, P. Gustafsson, J. Hankinson, R. Jensen, D. C. Johnson, N. MacIntyre, R. McKay, M. R. Miller, D. Navajas, O. F. Pedersen, J. Wanger

To the Editors:

In 1986, the American Thoracic Society (ATS) first suggested a fixed ratio of forced expiratory volume in one second (FEV1) to forced vital capacity (FVC) <0.75 to define airflow obstruction 1. Subsequent ATS documents published in 1991 2 and 1995 3 generically defined airflow obstruction as a reduction of FEV1/FVC, without recommending any numerical cut-off point.

By contrast, the European Respiratory Society (ERS) guidelines 4 suggested the diagnosis of airflow obstruction be based on a ratio of FEV1 to slow vital capacity (VC) <88 and <89% of predicted in males and females, respectively. These values were not arbitrarily chosen as they roughly correspond to the lower 95th percentiles of frequency distributions of a healthy population. More importantly, they are consistent with the well-known decrease of lung elastic recoil and, by inference, of forced expiratory flow with ageing.

In 2001, the Global Initiative for Chronic Obstructive Lung Disease (GOLD) took a step back, defining chronic obstructive pulmonary disease (COPD) by a fixed FEV1/FVC <0.70 5. Since then, the enthusiasm for having new guidelines has led the scientific community to overlook the possible consequences of such a definition, even if it was already clear that it may be a source of falsely positive cases in the general population 6. This was confirmed in a study in the USA 7 evaluating the impact of different definitions of airflow obstruction on the epidemiology of COPD. Quoting Celli et al. 7, “differences may be large, altering population prevalence estimates of COPD by >200%”. It is noteworthy that, using FEV1/FVC <0.70, the prevalence of COPD in individuals aged ≥70 yrs would be ≥40%.

At variance with the GOLD guidelines, the recent ATS/ERS guidelines on lung function testing 8 stressed the use of lower limits of normality (LLN), i.e. the lower fifth percentile of the frequency distribution of a healthy population, to define pulmonary function abnormalities.

In a recent editorial published in the European Respiratory Journal, Mannino 9 took a strong position in favour of the fixed FEV1/FVC <0.70, claiming that it is easy to keep in mind, thus helping to remove the barriers to a widespread use of spirometry, and is more sensitive than LLN to identify patients at risk of death and COPD-related hospitalisations 10.

We would like to draw the attention of the readers to the following critical issues.

First, the fixed cut-off point indicated by GOLD guidelines may have negative consequences by misclassifying healthy elderly subjects as COPD, thus possibly causing unnecessary treatment, and by misclassifying as healthy a number of subjects aged <50 yrs already affected by COPD 9, when something could be done to limit disease progression.

Secondly, the fact that risks of death and COPD-related hospitalisation 10 are predicted by FEV1/FVC <0.70 indicates that such an index may identify a proportion of individuals at risk 6, which has nothing to do with defining the diagnosis of the disease. Furthermore, it is an index which per se cannot reflect the severity of disease 8. This is clearly apparent if one keeps in mind that two patients with FEV1 of 20 and 100% pred may have the same FEV1/FVC <0.70 or even <LLN, depending on the associated reduction of FVC.

Thirdly, an FEV1/FVC >0.70 or even >LLN cannot exclude airflow obstruction with certainty because, in a minority of cases, FEV1 and FVC may be decreased proportionally as a result of an isolated increase in residual volume 8. This may lead to a false diagnosis of restriction instead of obstruction.

Fourthly, software and hardware have now changed the way of laboratory testing and there is no longer a need for manual, time-consuming calculations of predicted values, as even inexpensive spirometers can have predicting equations and statistically derived LLN values built in.

Finally, we understand that a fixed ratio might be useful where predicting equations are not available. However, the severity classification suggested by GOLD guidelines to tailor treatments, based on the percentage decrease from predicted FEV1 5, would be meaningless.

We are confident that with the world very rapidly “going global”, the advancement of technology in the medical field will help to promote a larger use of lung function testing and, with it, the generation of reference equations for different countries and ethnicities. For the time being, however, we suggest that a definition of the pulmonary defects consistent with solid principles of lung physiology is maintained.

Statement of interest

None declared.


    View Abstract