'Normal' lung function?
Your reply |
You propose to use FEV1%(F)VC as a percentage of the predicted value to assess whether a patient has or does not have airway obstruction. |
 |
It is very commendable that you want to use FEV1%(F)VC as the index for assessing obstructive lung disease. After all, the FEV1 as well as the VC may be diminished by both a restrictive and an obstructive disorder, but only an FEV1 that is disproportionately low with respect to the (F)VC signifies expiratory airflow limitation. |
 |
Unfortunately, expressing this index as a percentage of
the predicted value does not find favor. It would be appropriate
if a large predicted value were associated with a large scatter,
and a small one with a small scatter. However, this is not
the case in adults, the scatter is independent of the predicted
level. The appropriate procedure is therefore to compute how many standard deviations the observed value differs from the predicted value (the standard deviation score (SDS), also called the Z-score), as this index is independent of gender, age and standing height, and therefore unbiased. In the case of a normal distribution a SDS <-1.64 signifies that the observed value occurs in fewer than 5% of healthy subjects. For your convenience, in the SpirXpert software this is translated into a ‘lower limit of normal’. |
Please read the following chapters carefully, and then return to the question.
- Per
cent predicted usually misleading
- Expressing
spirometric test results correctly
- RSD,
standard deviation score and percentile
- The utility of the standard deviation score
- Bias
due to the use of % predicted in pulmonary function tests
- Age-specific
per cent predicted of FEV1
Back to the question