PT - JOURNAL ARTICLE AU - Eva Hnizdo AU - Kanta Sircar AU - Tieliang Yan AU - Philip Harber AU - James Fleming AU - Henry W Glindmeyer TI - Limits of longitudinal decline for the interpretation of annual changes in FEV<sub>1</sub> in individuals AID - 10.1136/oem.2006.031146 DP - 2007 Oct 01 TA - Occupational and Environmental Medicine PG - 701--707 VI - 64 IP - 10 4099 - http://oem.bmj.com/content/64/10/701.short 4100 - http://oem.bmj.com/content/64/10/701.full SO - Occup Environ Med2007 Oct 01; 64 AB - Objective: Spirometry-based screening programmes often conduct annual assessment of longitudinal changes in forced expiratory volume in 1 second (FEV1) to identify individuals with excessive rates of decline. Both the American Thoracic Society (ATS) and the American College of Occupational and Environmental Medicine (ACOEM) recommend a reference limit value of ⩾15% for excessive annual decline. Neither the ATS nor the ACOEM adjust this limit for the precision of the existing spirometry data. The authors propose an improved method of defining the reference limit of longitudinal annual FEV1 decline (LLD) based on the precision of the spirometry data.Method: The authors used data from four monitoring programmes and measured their data precision using a pair-wise within-person variation statistic. They then derived programme- and gender-specific absolute and relative LLD values and validated these against the 95th percentiles for observed yearly changes in FEV1.Results: The relative limit for annual decline was more practical than the absolute limit as it adjusted for gender differences in the magnitude of FEV1. The programme-specific relative limit values were in good agreement with 95th percentiles for year-to-year FEV1 changes and ranged from 6.6% to 15.8%. For individuals with COPD and bronchial hyperreactivity the 95th percentiles for year-to-year changes were about 15% and higher.Conclusions: The relative longitudinal limit for annual FEV1 decline based upon precision of measurements is valid and can be generalised to different gender and population groups. A relative limit of approximately 10% appears appropriate for good quality workplace monitoring programmes, whereas a limit of about 15% appears appropriate for clinical evaluation of individuals with an obstructive airway disease. Computer software based on the method described is available from the corresponding author.