Measuring results Copy

Measuring results

There are two methods of comparing a worker’s measured values against predicted values and these are:

  1. The classic older way:  Percentage of predicted (%pred) = measured / predicted x 100
  2. The new way: Z-score = (measured – predicted) / standard deviation

Figure 5.1: Example spirogram showing results expressed as %pred as well as a Z-score

Understanding percentage of predicted values

Where other physiologic parameters, like blood pressure or heart rate, are not dependent on other physiological properties, spirometry values can greatly differ depending on the worker’s specific properties like age, sex, ethnicity and height. To allow an interpretation of a spirometry test we compare the known values with the predicted values. The predicted values are average values of a large population of healthy workers. The software that comes with most spirometers calculates these predicted values.

Predicted values were used in the past because we did not always have better information at hand. In the ideal situation spirometry values should be compared to historical data of the same worker. Studies have shown that healthy workers will lose up to 25 ml of FEV1 every year from the age of 25. Losing more volume can be pathological.

Comparing the spirometry values only with the predicted values can lead to errors. If the FVC and FEV1 were lower than 80% of predicted values, the values were considered to be too low. But if a worker that has blown 120% of his predicted values and blows 100% of his predicted values one year later may have a very big problem, even if comparison with his calculated predicted values shows no signs of abnormality.

Further misdiagnosis occurs when fixed cut-offs, such as 80% predicted forced expiratory volume in 1 s (FEV1) or 0.70 FEV1/FVC, are used; particularly in young children and elderly adults. While per cent predicted has historically been used to interpret lung function results, Z-scores are more appropriate as they take into account the predicted value, as well as the between-worker variability of measurements.

Experience has shown that predicted values for spirometry are dynamic as the population changes. Years after publication of predicted values it becomes clear that the values do no longer correspond to the present-day population.

Understanding Lower Level of Normal (LLN)

The Lower Limit of Normal (LLN) Z-score is a more accurate way to assess spirometric values than percentage of predicted (%pred). A Z score is a number of standard deviations a score is above or below the mean. In the Standard Normal Distribution, the mean is always equal to 0 and the standard deviation is equal to 1.0. reference values

Figure 5.2: This pictogram shows the percentage of the population expected within a certain standard deviation in spirometry

(http://www.ltcconline.net/greenl/courses/201/probdist/zScore.htm)

Figure 5.3: This 3-part diagram shows the percent of a normal distribution that lies between 1, 2, and 3 standard deviations from the mean: between -1 and 1 you can find approximately 68%; between -2 and 2 is approximately 95%; and between -3 and 3 is approximately 99.7% of the population

(http://jukebox.esc13.net/untdeveloper/RM/Stats_Module_4/mobile_pages/Stats_Module_47.html)

Figure 5.4: This pictogram shows the Z-score relative to a normal range allowing a simple interpretation of the spirometric results

(https://www.spirometry.com/ENG/lln-z-score/lln-z-score.asp)

Advantages of using LLN

The recommendation for interpreting results is to compare the measured value with the LLN which corresponds to a value of Z-score of -1.64

The true LLN, when expressed as a % of predicted, varies considerably with age.  A FEV1 value of 66% of the predicted may be normal in a person who is 80 years old, but it would be abnormal for a person who is 20 years old. Therefore, evaluating spirometry results using the percentage of predicted is very popular but should be replaced by the method of using the Z-score.  The benefit of using a Z-score is that for any parameter a LLN Z score of –1.64 signifies in males, females, children and adults that the measured value is at the 5th percentile (5% of people will still have normal lung function even though they measure below 100%).  

The American Thoracic Society (ATS) and European Respiratory Society (ERS) both recommend the use of LLN to express between health and suspected disease. At this time, all spirometer software is enabled to report results as percentage of predicted but not all spirometer software will report results as LLN yet. Only the new software offers results as Z-scores. In this training course we address both ways of measuring lung function understanding that you may continue to use the older method of interpretation being percent of predicted should your spirometer not provide reports in LLN and Z-scores.

LLN and Global Lung Initiative

In 2012 the Global Lung Initiative published new predicted values that correspond better to the present-day population. For the first time in history there was a cooperation between 6 international respiratory societies:

  • ERS: European Respiratory Society
  • ATS – American Thoracic Society
  • ANZSRS – Australian and New Zealand Society of Respiratory Science
  • APSR – Asian Pacific Society for Respirology
  • TSANZ – Thoracic Society of Australia and New Zealand
  • CHEST – American College of Chest Physicians

Lung function of more than 74000 healthy non-smoking people, spanning all ages and ethnic groups was measured and from this vast amount of data the new GLI predicted values were calculated. It is advised to use these new GLI predicted values and no longer use other sources of predicted values.