Considerations in the use of different spirometers in epidemiological studies

• Main Authors: Edith B. Milanzi, Gerard H. Koppelman, Marieke Oldenwening, Sonja Augustijn, Bernadette Aalders-de Ruijter, Martijn Farenhorst, Judith M. Vonk, Marjan Tewis, Bert Brunekreef, Ulrike Gehring
• Format: Article
• Language: English
• Published: BMC 2019-04-01
• Series: Environmental Health
• Subjects: Calibration; Epidemiological studies; Lung function; Spirometry; Systematic difference
• Online Access: http://link.springer.com/article/10.1186/s12940-019-0478-2

Abstract Background Spirometric lung function measurements have been proven to be excellent objective markers of respiratory morbidity. The use of different types of spirometers in epidemiological and clinical studies may present systematically different results affecting interpretation and implication of results. We aimed to explore considerations in the use of different spirometers in epidemiological studies by comparing forced expiratory volume in 1 s (FEV1) and forced vital capacity (FVC) measurements between the Masterscreen pneumotachograph and EasyOne spirometers. We also provide a correction equation for correcting systematic differences using regression calibration. Methods Forty-nine volunteers had lung function measured on two different spirometers in random order with at least three attempts on each spirometer. Data were analysed using correlation plots, Bland and Altman plots and formal paired t-tests. We used regression calibration to provide a correction equation. Results The mean (SD) FEV1 and FVC was 3.78 (0.63) L and 4.78 (0.63) L for the Masterscreen pneumotachograph and 3.54 (0.60) L and 4.41 (0.83) L for the EasyOne spirometer. The mean FEV1 difference of 0.24 L and mean FVC difference of 0.37 L between the spirometers (corresponding to 6.3 and 8.4% difference, respectively) were statistically significant and consistent between younger (< 30 years) and older volunteers (> 30 years) and between males and females. Regression calibration indicated that an increase of 1 L in the EasyOne measurements corresponded to an average increase of 1.032 L in FEV1 and 1.005 L in FVC in the Masterscreen measurements. Conclusion Use of different types of spirometers may result in significant systematic differences in lung function values. Epidemiological researchers need to be aware of these potential systematic differences and correct for them in analyses using methods such as regression calibration.