Software for interpreting pulmonary function
The Global Lung Function Initiative published all-age (3-95 yr) multi-ethnic reference values for FEV1, VC, FEV1/VC, FEV0.75, FEV0.75/VC, FEF25-75%, FEF75% (endorsed by the major respiratory societies: European Respiratory Society, American Thoracic Society, Australian and New Zealand Society of Respiratory Science, Asian Pacific Society for Respirology, Thoracic Society of Australia and New Zealand, and the American College of Chest Physicians). Whereas these equations have often been implemented in software made available by spirometer manufacturers, in many instancies there is a need to evaluate one's own dataset. The GLI2012_DataConversion software, which is accompanied by een extensive help file, serves the following purposes:
Transform your dataset so that it includes the Quanjer GLI-2012 predicted values, uper and lower limit of normal, z-score and percent of predicted for FEV1, VC, FEV1/VC, FEV0.75, FEV0.75/VC, FEF25-75%, FEF75%.
Check how well your reference set of healthy non-smokers fits the Quanjer GLI-2012 predicted values. One should be aware of the fact that at least 150 females and 150 females are required to obtain mainingful results. The software calculates the average z-score and its standarad deviation. In the case of a perfect fit the average z-score should be 0, and the standard deviation 1. Even then the average z-score for the indices, which should ideally be zero, may differ by up to 0.4 z-score from the GLI-2012 predicted values. Such differences are due to the sample not being fully representive of the population, but may of course also be due to technical problems. If your dataset comprises 1,000 males and 1,000 females, differences of up to 0.23 z-score may still occur. If the standard deviation of the z-score of your data is appreciably larger than 1, you will need to investigate reasons for such a large variability.
New Ethnic Group
You may wish to derive local prediction equations using your own representative dataset of healthy nonsmoking males and females, for example because they represent an ethnic group not covered by the Quanjer GLI-2012 equations. Do keep in maint the caveats outlined above. Once you have derived the new constants for the additional equation, you will have to contact the Global Lung Function Analyticial Team for instruction how to implement the new equations. Instrument manufacturers are not allowed to implement these new equations without permission from the Global Lung Function Analytical Team.
Validate implementation of the GLI-2012 equations into software
Implementation is not as easity as with conventional reference equations, so you may want to check whether your software generates correct results. To that end the software is accompanied by a test set which covers different ages, heights, sex and ethnic groups, and the resulting output generated by the software. You can, of course, make your own test set following the instructions in the accompanying help file.