Published ahead of print on June 7, 2007, doi:10.1164/rccm.200702-337OC
© 2007 American Thoracic Society doi: 10.1164/rccm.200702-337OC
Six-Second Spirometry for Detection of Airway ObstructionA Population-based Study in Austria1 Department of Pulmonary Medicine, Paracelsus Medical University, Salzburg, Austria; 2 Department of Pulmonary and Critical Care Medicine, Oregon Health and Science University, Portland, Oregon; and 3 University of Arizona, Tucson, Arizona Correspondence and requests for reprints should be addressed to Bernd Lamprecht, M.D., Muellner Hauptstrasse 48, Department of Pulmonary Medicine, 5020 Salzburg, Austria. E-mail: b.lamprecht{at}salk.at
Rationale: The presence of airway obstruction is currently defined by Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines on the basis of the post-bronchodilator (BD) FEV1/FVC. It has been proposed that the traditional FVC can be replaced with the shorter and less demanding FEV6 for detecting airway obstruction. Objectives: A comparison of FEV1/FVC and FEV1/FEV6 for the detection of airway obstruction in population-based post-bronchodilator spirometry data. Methods: A population-based sample of 1,349 adults participated in the Burden of Obstructive Lung Disease study in Austria. Specially trained and certified technicians conducted pre-BD and post-BD spirometry according to American Thoracic Society guidelines and administered standardized questionnaires. A total of 93% of the post-BD test sessions were acceptable, and were included in this analysis. The Third National Health and Nutrition Examination Survey reference equations were used to calculate predicted values and lower limits of normal (LLN) for FEV1, FEV6, FVC, FEV1/FVC, and FEV1/FEV6. Measurements and Main Results: The post-BD FEV1/FVC was below the LLN in 199 (15.8%) subjects. The sensitivity of the FEV1/FEV6 for airway obstruction depended greatly on the threshold of percent predicted FEV1 also used in the definition. The overall sensitivity of FEV1/FEV6 for a diagnosis of airway obstruction, as defined by FEV1/FVC (including participants with an FEV1 above the LLN), was 72.9%, with 98.8% specificity. The sensitivity increased to 98.0% when a low FEV1 was also required to diagnose post-BD airway obstruction. The discordant cases had long forced expiratory times, often showed a flow–volume curve pattern consistent with two-compartment emptying, and were more often never-smokers. Conclusions: Six-second spirometry maneuvers (which measure FEV6) are as sensitive and specific for post-BD airway obstruction as traditional (prolonged exhalation time) FVC maneuvers only when the definition of airway obstruction includes a low FEV1.
Key Words: airway obstruction • chronic obstructive pulmonary disease • FEV6 • spirometry
Chronic obstructive pulmonary disease (COPD) is an important and still-increasing cause of morbidity and mortality, and therefore a relevant, yet unrecognized, public health problem. COPD is projected to rank third among all causes of death worldwide in 2020 (1). Many adult cigarette smokers have developed airway obstruction that has not been diagnosed by their physicians. Estimates of the prevalence of COPD in a country are low when only respiratory symptoms are determined by a questionnaire. Objective measurement of lung function by spirometry testing is needed to determine the true prevalence of COPD (2). Spirometry can be demanding, often requiring prolonged exhalation times to achieve a plateau on the volume–time curve and a small end-of-test volume, which indicate complete exhalation. Six-second FVC maneuvers have been proposed and studied to minimize the possible risk of syncope due to a subject's prolonged expiratory effort, to make the test less exhausting, and to enhance the reproducibility of the test (3–6). The National Lung Health Education Program recommends the use of the FEV1/FEV6 for the detection of COPD in the primary-care setting (7). The National Lung Health Education Program group defined airway obstruction when the FEV1/FEV6 ratio and the FEV1% predicted are both below their lower limit of normal (LLN). Data collected in hospital-based pulmonary function laboratories using this definition of airway obstruction prebronchodilator (pre-BD) have shown high sensitivity and specificity for the FEV1/FEV6 compared with the gold standard FEV1/FVC (4–6). However, patients referred to a hospital-based pulmonary function laboratory are much more likely to have moderate or severe airway obstruction when compared with general population-based samples where screening for COPD may be useful. The present investigation uses data from the Burden of Obstructive Lung Disease (BOLD) study (8), which is an international effort to measure the prevalence of COPD. One of the goals of BOLD is to evaluate different criteria for irreversible airflow obstruction. Given that 6-second FVC maneuvers are easier, especially in population studies, we were interested in comparing the FEV1/FEV6 and FEV1/FVC for the detection of airway obstruction. Some of the results were previously reported as an abstract (9).
Following the BOLD (8) protocol, we surveyed a sex-stratified random sample of Salzburg County, Austria, subjects aged 40 years and older. Participants were recruited by phone and invited to lung function testing (10).
Study Population
Study Measures
Definitions Smoking status (current or former smoker) was defined as smoking more than 20 packs of cigarettes in a lifetime or more than 1 cigarette/day for a year. Breathlessness was defined as being troubled by shortness of breath when hurrying on the level or walking up a slight hill. Cough was recorded if the participant did usually cough when he/she did not have a cold. For the presence of phlegm it was asked: "Do you usually bring up phlegm from your chest, or do you usually have phlegm in your chest that is difficult to bring up when you don't have a cold?" Passive smoking was present if a household member smoked cigarettes, pipes, or cigars in the home during the past 2 weeks.
Statistical Methods
The demographic characteristics and lung function data of study participants are shown in Tables 1 and 2. The overall prevalence of post-BD airway obstruction, defined by FEV1/FVC less than LLN (regardless of % predicted FEV1), was 15.8% and was significantly higher for women than for men (p = 0.0016). The prevalence of post-BD airway obstruction in the 515 never-smokers who denied having asthma was 9.7%.
For all 199 participants with FEV1/FVC less than LLN, in 145 (72.9%), the FEV1/FEV6 was also below the LLN (concordant cases). The sensitivity of FEV1/FEV6 was 72.9%, and the specificity was 98.8% when % predicted FEV1 was not considered (see Table 3). In 67 (5.3%) participants, the categorizations on the basis of FEV1/FVC were different from those made on the basis of FEV1/FEV6 (discordant cases). Of those 67 (5.3%) discordant cases, 13 had an FEV1/FVC greater than or equal to LLN (false positive), and 54 had an FEV1/FVC less than LLN (false negative).
Of the participants with airway obstruction (FEV1/FVC < LLN) who were not identified by FEV1/FEV6 (discordant, false-negative cases), only two had a low FEV1 (see Table 4). The sensitivity of FEV1/FEV6 for a diagnosis of airways obstruction increased from 65% in borderline airway obstruction to 98% in moderate obstruction and 100% in severe obstruction.
The mean difference between FVC and FEV6 was 377 ml (±213 ml [SD]) in subjects with any category of airway obstruction (FEV1/FVC < LLN), and 150 ml (±118 ml) in those with no airway obstruction (FEV1/FVC > LLN) (p < 0.001). In discordant cases, the mean difference between FVC and FEV6 was 399 ml (±228 ml). The comparison of concordant and discordant obstructive cases revealed a significant (p < 0.001) shift toward less severe disease and less symptoms in discordant cases (see Table 5). In addition, the proportion of never-smokers was significantly higher among discordant cases.
We carefully examined the spirometry results from the 54 discordant cases. The mean forced expiratory time (FET) was 11.4 seconds (5–90% confidence interval, 10.6–12.2) which was longer than the mean FET from other subjects (mean, 10.6; 5–90% confidence interval, 10.2–11.1). Many of these cases had an unusual pattern suggesting "two-compartment emptying" on the flow–volume curve (see Figure 1).
This study is one of the first to investigate post-BD FEV6 for detecting airways obstruction in a population sample. Our data demonstrate an overall prevalence of airway obstruction, defined by post-BD FEV1/FVC less than LLN, of 15.8% in subjects aged 40 years and older. However, the prevalence of severe airway obstruction (GOLD stages 3 or 4) was only 1.1% (14/1,258). The overall prevalence rate for post-BD airways obstruction was significantly higher in women than in men. This finding may be due to the change in smoking habits in Austrian men and women. In addition, women may be more susceptible to the harmful effects of smoking cigarettes when compared with men (13, 14). The 9.7% rate of airway obstruction in never-smokers without a history of asthma is higher than the expected value of around 5%. This higher prevalence may be due to COPD arising from occupational or environmental exposures, which we did not measure. The FEV1/FEV6 was a very good surrogate for the traditional FEV1/FVC in detecting post-BD airway obstruction when the FEV1 was below 80% predicted, but was less sensitive for detecting airway obstruction in borderline cases (defined as mild by some COPD guidelines). Those in the subset of smokers who slowly develop COPD progress from normal spirometry to borderline airway obstruction and then to unequivocal airway obstruction over a period of decades (15). The transition from healthy lungs to diseased lungs in these smokers is gradual, without a discrete threshold, much like the development of type 2 diabetes and hyperlipidemia. Despite financial and political imperatives, there is currently insufficient evidence to confidently set a threshold for abnormal spirometry, as reflected by the differences in published COPD guidelines. GOLD recommends the use of a fixed ratio of FEV1/FVC (post-BD) to define irreversible airflow obstruction, then stages the severity of COPD using FEV1% predicted (16). This recommendation has been challenged because the fixed ratio has been shown to overdiagnose airflow obstruction, especially in the elderly (17, 18). The challengers recommend the use of the LLN of the FEV1/FVC ratio rather than the fixed ratio (19, 18). In response, Mannino and coworkers analyzed data from a large cohort study of older adults and reported that people who had an abnormal fixed ratio had an increased risk of all-cause mortality during the 11-year follow-up (20). One can be confident that a 50-year-old smoker with an FEV1/FVC of 0.60 and an FEV1 of 60% predicted that does not change after BD has COPD; however, COPD is much less certain when the FEV1/FVC is 68% predicted (borderline low), but the FEV1 is 95% predicted (within the normal range). The only effective treatment for borderline or mild COPD is smoking cessation, so all smokers should be encouraged and helped to quit smoking (21). Studies have not shown that discussing abnormal spirometry results with a smoker substantially improves subsequent smoking cessation success (21, 22). Therefore, screening for borderline to mild COPD has not been demonstrated to have clinical benefits (23). FEV1/FEV6 was found to be as good as FEV1/FVC for predicting subsequent decline in lung function in cigarette smokers with borderline to moderate airway obstruction in the Lung Health Study, although spirometry results at study entry explained only 10% of the subsequent 5-year declines in FEV1 (24). Shortening the forced expiratory maneuver to 6 seconds may reduce both patient and technician effort, and could possibly help to prevent syncope (25). In occupational settings, FEV6 has been described as a very good surrogate for FVC in detecting airway obstruction when the FEV1 was also low (< LLN) (26). In addition, the role of FEV1/FEV6 in the detection of airway obstruction has been discussed in studies from consecutive groups of patients referred to hospital-based pulmonary function laboratories (4–6). Reported sensitivity of FEV6 varied from 86% (5) (inappropriately using the same 70% cutoff for both FEV1/FEV6 and FEV1/FVC) to 95% (6) (using appropriate LLN reference equations). The lower overall sensitivity of FEV6 in our population-based data can be explained by the much smaller proportion of subjects with severe airway obstruction, and the reduced FEV6 sensitivity in borderline airway obstruction and never-smokers. In lung function laboratories, the prevalence and severity of obstruction is much higher than for patients tested in medical settings outside of hospitals and for population-based samples. Vandervoorde and coworkers (6) analyzed spirometry data of 11,676 subjects (aged 20–80 yr) referred to their lung function laboratory. In their sample, the prevalence of airway obstruction was 39.5%, and 32.5% of all subjects with obstruction had severe airway obstruction (FEV1 < 50% predicted). The prevalence of airway obstruction in the data reported by Swanney and coworkers (4) (n = 337) was 65.6%, and 53.4% of these patients had severe obstruction (FEV1 < 50% predicted). In agreement with our results, Hansen and colleagues recently reported that FEV1/FEV6 correctly identifies moderate to severe obstruction, but is less reliable for detecting borderline airway obstruction (27). Another factor that increased the rate of discordant interpretations in our study is that our technologists vigorously obtained long-duration exhalation times: half of the FVC maneuvers had FETs of more than 10 seconds. When shorter maneuvers are considered acceptable, as in many primary-care settings, then the FVC more closely approximates the FEV6, and thus the FEV1/FEV6 is closer to the FEV1/FVC. If short 6-second maneuvers are usually performed in a given setting, then reference equations for 6-second maneuvers will provide less misclassification for airway obstruction than reference equations, which assume complete exhalations (low end-of-test volumes). However, at present, only NHANES III (12, 28) and one study of older European adults (29) provide reference equations for the FEV1/FEV6 and the FEV6. Perhaps in the future, other BOLD and PLATINO (Latin American Project for the Investigation of Obstructive Lung Disease) (30) studies will provide reference equations for 6-second maneuvers for other racial and ethnic groups. Our data suggest that a lower sensitivity for borderline airway obstruction will probably still occur when using these new reference equations. Limitations of our study include the use of reference equations from the United States that may not apply to our population. The NHANES III study also used a volume-sensing spirometer, which does not quickly determine the BTPS conditions of the exhaled gas as it cools in the bell. This may affect the FEV1/FVC measured by volume spirometers. We ran a sensitivity analysis to determine the effect of a 2% lower and 2% higher predicted FEV1/FVC LLN on the post-BD airway obstruction prevalence rate, which we report as 15.8%. The alternate rates were 12.7 and 19.7%, respectively. Another potential limitation of our study is that 200 µg of salbutamol was administered instead of 400 µg, as recommended by the GOLD guidelines. The prevalence of post-BD airway obstruction would have been slightly lower with a higher dose of BD. However, many pulmonary function laboratories have been using 200 µg and we worried about the risk of arrhythmias in our field settings. The pattern of two-compartment emptying seen on a few flow–volume curves in our study deserves better quantitation and determination of its clinical correlates. This pattern could be produced by a faulty zero-flow setting before the test begins or when drift of the flow signal occurs during prolonged exhalations (31). However, the digital ultrasonic sensor used by our study spirometer may be less susceptible to drift caused by changes in ambient and exhaled temperature when compared with older spirometers, which use analog pressure transducers. In addition, the pattern was repeatable from maneuver to maneuver, both pre-BD and post-BD in most cases. Although spirometry has long been recommended for the primary-care setting in smoking adults with any respiratory symptom (a high pretest probability) (2, 32), COPD remains a greatly underdiagnosed disease, even when considering only those with unequivocal post-BD airway obstruction (33). We support recommendations for COPD case finding in patients with a high pretest probability, using either traditional spirometric indices of airway obstruction (FEV1/FVC) or 6-second maneuvers using appropriate LLNs (7).
Originally Published in Press as DOI: 10.1164/rccm.200702-337OC on June 7, 2007 Conflict of Interest Statement: B.L. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. L.S. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. F.T. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. B.K. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. S.A.B. has served on advisory boards for GlaxoSmithKline (GSK), Altana, Schering Plough, Merck, Novartis, Pfizer, and Sepracor, has participated in chronic obstructive pulmonary disease workshops funded by AstraZeneca and GSK, and is scientific director for the Burden of Obstructive Lung Disease (BOLD) initiative that receives unrestricted educational grants to the Kaiser Permanente Center for Health Research from GSK, Pfizer, Boehringer Ingelheim, AstraZeneca, Altana, Novartis, Merck, Chiesi, Schering Plough, and Sepracor. M.S. received less than $10,000 in lecture fees from GSK, Novartis, AstraZeneca, Torrex-Chiesi, Boehringer Ingelheim, and MSD from 2004 to 2006, and is the principal investigator of the Austrian BOLD study that received grants from GSK, Boehringer Ingelheim, AstraZeneca, Altana, Novartis, and MSD. P.E. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form February 28, 2007; accepted in final form June 4, 2007
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
AT A GLANCE COMMENTARY