ATS Pulmonary Function Laboratory Manual

ATS Pulmonary Function Laboratory Management & Procedure Manual | 3rd Edition

2. Selection and verification of published reference equations (1) 2.1. The European Community for Coal and Steel (ECCS) published a comprehensive list of equations from studies published prior to 1983 (5). A more recent list of equations was published in the 2005 ATS/ERS Interpretative strategies for lung function tests (4). 2.2. As new reference equations are published, review the equations chosen by the laboratory. 2.3. Spirometry: When testing patients in North America, ages 8 to 80, it was recommended in the 2005 ATS/ERS recommendations (4) that the NHANES III reference equations (7) be used. The NHANES III reference equations are based on a random sample in the United States. Equations for predicted values and lower limits of the reference range are provided for individuals ages 8 to 80, three ethnic groups, and includes two new spirometric variables (FEV 6 and FEV 1 /FEV 6 ). Other reference equa- tions may be used if they provide better comparisons for a specific laboratory or patient group. One important and recent addition for spirometry are the data from the Global Lung Function Initiative (www.lungfunction.org) which extend the age ranges down to age 3.5 and up to age 95, but include data from non-U.S. populations (6). 2.4. There are no specific recommendations on reference equations for Dl CO and lung volumes. 2.5. Reference equations should be updated on a regular basis (e.g., every 10 years), taking into account the applicability of the newer reference equations and the effect on interpretation of longitudinal patient follow up (4). Compare the results of at least 40 local reference subjects (i.e., healthy, non-smoking male and fe- male subjects with a wide range of ages and heights) to the intended reference equations. 3.2. If more than 30% of the people studied fall outside the 95% confidence interval for normal limits, review all aspects of the testing procedure and then select another equation set (8). 3.3. Jensen and coworkers (7) reported that 20 to 40 subjects may not be enough to consistently confirm the selection of appropriate reference values and this method should be used cautiously. Quanjer and coworkers found that at least 150 males and 150 females would be necessary to validate values (9). Limitations of available equations (1) 4.1. Data may be limited for some ethnic groups, and even within some ethnic groups, differences may exist (10). 5. Epidemiologic issues (1) 5.1. Healthy subject reference equations that include hospital patients should not be used. 5.2. Select studies from a population free of respiratory symptoms and diseases. 5.3. Base the reference equations on a nonsmoking population. 5.4. Consider the effect of altitude on values for Dl CO and spirometric flows. 6. Statistical considerations (1) 6.1. Age and height should be the primary independent variables. 6.2. Linear equations may over-predict lung function in young adults and adolescents, and either over- predict or under-predict lung function values in the elderly. 6.3. Reference value studies should provide appropriate limits of the healthy reference range. For spiro- metric values, only lower limits of the reference range are appropriate since the question being asked is almost always: Is the measured value too low? For lung volumes and Dl CO , upper and lower limits of the reference range are recommended because the question being asked may be either: Is the measured value too low? Or, is the measured value too high? 3. Validation of selection 3.1. 4.

234