What is Abnormal?    Page 1


   When an individual has their lung function measured for the first time the clinician will want to know if the result is unusual for that individual. The clinician needs to know if the result is within the expected range for someone of their age, sex, height and ethnicity (all factors known to influence their lung function results).

   The best method to determine whether the result is unusual is to calculate how far the subject's result is from their predicted value and then express this in terms of the number of standard deviations their value is from their predicted. This result is called the "Standardised Residual" (SR) or the z-score and is given by:

Plot of FEV1 data showing definition of a residual

                SR = (Recorded Value - Predicted Value) / RSD

   RSD is the residual standard deviation from the prediction equation used. The term "residual" refers to the difference between a given subject's result and their predicted value.

   The figure to the right shows over 1000 data points for normal healthy male subjects' FEV1 values used to create predicted values. A male patient's FEV1 value, point A, is 2.1 L and this subject has a predicted value, point B, of 3.8L and so the patient's residual is 2.1 - 3.8 = -1.7L. This residual is then standardised by the spread (standard deviation) of all the residuals found in this male population used for making the prediction. For these data the RSD = 0.51L so the SR value is -1.7/0.51 = -3.3

   Calculating the predicted value is usually done using a prediction equation taken from a large population of comparable subjects who have never smoked. The GLI-2012 equations are derived from over 70,000 subjects and give the most reliable predictions for European subjects based on their sex, age, height and ethnicity [6]. The GLI-2012 equations return the z-scores (SR values) given the subject's sex, age, height and ethnicity.