PURPOSE: Spirometric diagnosis of chronic obstructive pulmonary disease (COPD) is commonly based on Global Initiative for Chronic Obstructive Lung Disease (GOLD) criteria using a fixed ratio of FEV1/FVC < 70% as abnormal. However, false positive in elderly patients and false negative in young patients may result from this assumption. The use of the <5% lower limit of normal (LLN) for FEV1/FVC as a cut-off, has a more solid scientific foundation and is advocated in order to increase diagnostic accuracy.
METHODS: Using spirometry results from a pool of healthy local population, we calculated regression equations predicting normal values for FEV1 & FEV1/FVC. The 5% LLN for these parameters were then calculated. A total of 721 patients were subsequently classified as obstructive or not by using both the GOLD and the LLN criteria, and results were compared.
RESULTS: In total, 149/721 (20.7%) patients were deemed obstructive by GOLD versus 177/721 (24.5%) patients by the LLN criterion. 35/177 (19.8%) patients considered obstructive by the LLN were normal by GOLD. Conversely, 7/149 (4.7%) patients considered obstructive by GOLD were normal by the LLN.
CONCLUSION: In this unselected population, we confirmed a considerable discrepancy between the GOLD and the LLN criteria for the diagnosis of COPD. Although simple, the use of a fixed ratio of FEV1/FVC as diagnostic of obstruction may be inaccurate in certain circumstances. On the other hand, the use of the LLN requires calculation of regression equation for FEV1/FVC (and FEV1 separately for severity staging). It is possible that variation of results will be greater compared to the GOLD criteria, if using different sets of prediction equations. This could be a possible obstacle in the adoption of the LLN criterion in every day practice.
CLINICAL IMPLICATIONS: Calculation of LLN form various populations and subsequent comparison of classification between different sets of prediction equations would be important in order to examine the eventual necessity of applying local instead of universal prediction equations in the diagnosis of COPD.
DISCLOSURE: George Tatsis, No Financial Disclosure Information; No Product/Research Disclosure Information