NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Wilt TJ, Niewoehner D, Kim CB, et al. Use of Spirometry for Case Finding, Diagnosis, and Management of Chronic Obstructive Pulmonary Disease (COPD). Rockville (MD): Agency for Healthcare Research and Quality (US); 2005 Sep. (Evidence Reports/Technology Assessments, No. 121.)

  • This publication is provided for historical reference only and the information may be out of date.

This publication is provided for historical reference only and the information may be out of date.

Cover of Use of Spirometry for Case Finding, Diagnosis, and Management of Chronic Obstructive Pulmonary Disease (COPD)

Use of Spirometry for Case Finding, Diagnosis, and Management of Chronic Obstructive Pulmonary Disease (COPD).

Show details



Prevalence of Airflow Obstruction, Chronic Obstructive Lung Disease, and Use of Spirometry for Diagnosis and Case-Finding

COPD is a major health problem resulting in considerable morbidity, mortality, loss of productivity, and utilization of health care resources. Individuals with respiratory symptoms compatible with COPD are often not diagnosed or are misdiagnosed. Compared to clinical examination alone, spirometry, in combination with clinical examination, improves diagnostic accuracy of clinically significant disease in adults with respiratory symptoms (especially dyspnea) that are compatible with COPD. No single item or combination of items from the clinical examination rules out spirometrically determined airflow limitation. The best clinical finding associated with decreased likelihood of airflow limitation is a history of never having smoked cigarettes (especially in patients without a history of wheezing and without wheezing on examination). The best findings associated with increased likelihood of airflow limitation are objective wheezing, barrel chest, positive match test result, rhonchi, hyperresonance, forced expiratory time greater than 9 seconds, and subxyphoid apical impulse. A finding of two of the following virtually rules in airflow limitation: 70-pack years or more of smoking, decreased breath sounds, or history of COPD. Three findings predict the likelihood of airflow limitation in men: years of cigarette smoking, subjective wheezing, and either objective wheezing or peak expiratory flow rate. The clinical history, respiratory symptom status and physical examination are of limited value in determining whether an individual has airflow obstruction.

Based on NHANES results, 12.8 percent of the adult population reported a current or past clinical diagnosis of OLD. However, only 17.4 percent had 1987-ATS defined low lung function, suggesting that most individuals who report a diagnosis of emphysema or chronic bronchitis of COPD do not have airflow obstruction. Of individuals reporting a diagnosis of COPD, 25.6 percent reported chronic phlegm and 48 percent reported shortness of breath; the symptom most likely to affect quality of life and predict mortality.

The prevalence and severity of airflow obstruction in general populations vary across countries. The biggest factor in varying prevalence estimates is the criteria used to define airflow obstruction and clinically significant COPD. Within the same population the prevalence of disease defined as “at risk” or having air flow obstruction can vary more than three fold by altering definition thresholds. The prevalence of airflow obstruction and symptoms increases with age and a history of smoking. There are relatively few differences according to race or gender after accounting for age and smoking status. Increasing severity of spirometrically determined airflow obstruction is associated with respiratory symptom prevalence. However, respiratory symptoms are not unique to COPD and may be due to other medical conditions (e.g., heart failure, deconditioning) even in the presence of airflow obstruction. Many individuals with respiratory symptoms have normal airflow and nearly one-quarter of individuals with severe to very severe airflow obstruction have no respiratory symptoms. Impairment in health status is most commonly associated with dyspnea and typically not evident until individuals have postbronchodilator airflow obstruction of GOLD 3,4 severity (FEV1 <50 percent predicted). Less than 5 percent of the U.S. population has respiratory symptoms and moderate, severe, or very severe airflow obstruction. A substantial portion of these individuals may not have been diagnosed with COPD and many who have reported a clinical diagnosis of COPD do not have airflow obstruction. Spirometry performed in the absence of bronchodilator testing (a method likely to be encountered in primary care clinics) identifies over 20 percent of the U.S. adult population and 25 percent of current smokers as having “abnormal airflow” or being “at-risk.” Prevalence increases with age and a broader definition of what constitutes airflow obstruction. The vast majority of individuals with airflow obstruction detected by case finding with spirometry have mild airflow obstruction and no dyspnea.

Spirometry, while important in determining prognosis, whether respiratory symptoms are likely due to COPD, and whether these symptoms would improve with COPD specific therapy is not an ideal test for establishing a diagnosis of clinically significant COPD. Using physiologic variables to define clinically significant COPD differs from other chronic conditions such as hypertension, diabetes, or hyperlipidemia that use laboratory values to define clinically significant disease and evaluate treatment effectiveness even in the absence of symptoms. Unlike those conditions, interventions for COPD, except for oxygen therapy in individuals with resting hypoxemia and smoking cessation have not been shown to be effective in asymptomatic adults, do not alter the laboratory parameter used to determine disease status (spirometry) acutely or over prolonged followup, and do not reduce mortality. Additionally, in subjects with COPD clinical outcomes are not associated with spirometric response to treatment and the symptom of dyspnea is a better predictor of clinical outcomes than spirometry. Instead, the benefits of COPD interventions are to improve patient's existing symptoms and functional status. Spirometric testing is of value to improve diagnostic accuracy in individuals reporting bothersome respiratory symptoms. Individuals should not be labeled as having COPD or treated with COPD-specific medications in the absence of respiratory symptoms and spirometric testing that demonstrates airflow obstruction.

Spirometry for Smoking Cessation

Smoking cessation is the most important intervention to reduce the development and/or progression of airflow obstruction and symptomatic COPD. Quitting smoking is also an important factor in reducing a wide range of other medical conditions that result in considerable morbidity, mortality, and health care costs. Thus, relatively small improvements in smoking cessation rates due to feasible interventions would be beneficial. Except for smoking cessation, no interventions have been demonstrated to reduce spirometric decline in lung function or prevent the development of respiratory symptoms in asymptomatic individuals within a 3-year period.

However, all adults should have smoking status assessed regardless of symptom or spirometric status. Counseling strategies and interventions, including pharmacologic therapy, should be offered for those willing to quit. Smoking cessation rates and motivation to quit may differ slightly according to spirometric and symptom status. However, results are inconsistent and the variability and magnitude of the difference according to these categories is unlikely to provide independent aid for clinicians in determining an individual patient's likelihood of quitting or whether targeted programs would be beneficial.

The evidence from non-randomized studies indicates that biological markers, including spirometry, may have some potential as a motivational tool as part of a multidisciplinary approach to assist patients and clinicians improve smoking cessation rates. The lack of controls makes assessment of the independent contribution of spirometry problematic. Randomized trials of other biomarkers for improving smoking cessation have generally been negative. Improvements in smoking cessation rates are generally of small magnitude and generally require multimodality therapy. Thus, there is little evidence for the biologic plausibility that spirometry would provide more than small improvements in smoking cessation.

Baseline symptom or spirometry status appears to be of limited clinical use in risk stratification and in assisting clinicians' target smoking cessation strategies. Efforts to improve smoking cessation rates in subjects with COPD have led to a modest increase in abstinence. However, because smoking has a wide range of serious adverse effects, even fairly small differences in cessation rates may be clinically important if they could be achieved feasibly in clinical settings. The only randomized trial to demonstrate a long-term improvement in smoking cessation rates among subjects with mild to moderate COPD or judged to be at increased risk used a pharmacologic intervention provided free of charge in combination with an intensive program of cessation and maintenance counseling. All subjects were provided their spirometric results. The intensity of this type of smoking cessation program may not be generalizable to primary care clinics. Differences in symptom status and baseline spirometric values between subjects who quit and those who continued to smoke were small and inconsistent in direction.

The evidence from randomized controlled trials assessing the effectiveness of obtaining spirometry and discussing results with current smokers in order to improve smoking cessation is limited and flawed. However, the evidence indicates that spirometry is unlikely to provide more than small improvements in smoking cessation rates. The intervention arms of six of the seven studies involved multiple components that are known to alter smoking cessation rates or had control groups that did not receive smoking cessation advice/therapies.8085 Therefore they do not allow for the independent assessment of the effects of spirometry. The only study that assessed the independent effect of spirometry failed to demonstrate a benefit (abstinence rate difference of 1.0 percent).79 This study was relatively small, suffered from poor physician and patient compliance, and did not obtain spirometry directly in the primary care setting. Two studies approximate the independent effects of spirometry on smoking cessation. Their results are conflicting.80,81 One showed an absolute point-prevalent abstinence rate difference of 13.3 percent at 12 months that favored the spirometry group.80 The other had an absolute point-prevalent abstinence rate difference of 5 percent at 9 months that favored the control group.81 None of the study results were statistically significant.

There is no information whether spirometry improves the prognosis of a subject's willingness to quit and/or addiction to tobacco. The only study of a mandated program that stratified quit rates by spirometry results reported less abstinence in patients with abnormal spirometry.85 This suggests the possibility of recidivism among patients with abnormal spirometry. Spirometric results may theoretically provide information that enhances physician compliance and/or effectiveness in providing smoking cessation therapies. Additionally, it may motivate smokers to quit. However, there is little empiric evidence from randomized controlled trials that assesses the effectiveness and potential adverse effects of spirometry for smoking cessation.

Spirometry for Initiating, Monitoring, and Modifying Treatment

Results from NHANES indicate that the majority of individuals reporting a clinical diagnosis of COPD have normal prebronchodilator airflow on spirometry. In the absence of spirometric testing, these individuals likely were incorrectly diagnosed and may have received unnecessary and ineffective treatment. Initiating COPD specific interventions in subjects with respiratory symptoms should not be done unless spirometric testing is performed and confirms airflow obstruction.

Treatment trials typically were of short duration and enrolled subjects with an established clinical diagnosis of COPD, activity limiting and bothersome respiratory symptoms (especially frequent exacerbations), and moderate to very severe airflow obstruction on baseline spirometry. No trials adjusted interventions according to an individual's baseline or followup spirometry, spirometric response to treatment, slope of spirometric values over time, or crossing a “threshold” spirometric value. Compared to placebo inhaled corticosteroids and long-acting bronchodilators reduced the absolute percentage of individuals having at least one exacerbation over a 3 month to 5 year time period by 5–6 percent. Comparative studies suggest that long acting β agonist and long-acting anticholinergics are of similar efficacy in preventing COPD exacerbations, but inhaled corticosteroids were slightly more effective than LABA. Short-acting anticholinergics are not superior to placebo, slightly less effective than long-acting anticholinergics, and comparable to short acting β agonists. These benefits were almost exclusively limited to individuals with a previous clinical diagnosis of COPD who had activity limiting or bothersome respiratory symptoms and baseline spirometry indicating severe to very severe airflow obstruction (GOLD Stage 3,4). Treatment effectiveness did not vary according to dose of pharmacologic interventions. Hospitalization rates were rarely reported and were lower compared to placebo by 4–7 percent.

The average improvement in respiratory health status due to inhaled corticosteroids and bronchodilators did not achieve a previously determined level of clinical significance even in individuals with severe airflow obstruction. However, individual patients may obtain a large and noticeable benefit and studies of tiotropium indicated that a greater percentage of subjects receiving tiotropium achieved a clinically significant improvement than those receiving placebo. Inhaled bronchodilators and corticosteroids did not alter spirometric decline or reduce mortality in subjects with baseline spirometry indicating airflow obstruction, though the number of subjects and duration of studies may be inadequate to conclusively conclude that they are ineffective for mortality.

Interventions other than smoking cessation do not prevent the development of respiratory symptoms among individuals not reporting these symptoms at baseline. These interventions also do not reduce mortality or spirometric decline in lung function. Therefore, treatment benefits are almost exclusively due to improvement in bothersome respiratory symptoms and possibly respiratory related health status among individuals with activity limiting respiratory symptoms. Many subjects enrolled in trials with mild to moderate airflow obstruction did not have activity limiting respiratory symptoms (or reported no symptoms) or a prior diagnosis of COPD. Most were detected based on spirometry in a fashion likely to occur with broad based primary care testing. The longest trial had a followup of 5 years, and thus the effectiveness of these agents on COPD outcomes at longer duration is not known. Pooled analysis of three trials of inhaled corticosteroids enrolling approximately 2,500 subjects with a mean FEV1 >2L (GOLD Stage 0–2) and followed for 3 or more years failed to demonstrate a benefit in clinical outcomes, although there was a trend towards a reduction of mortality. One of these studies114 demonstrated a statistically significant but clinically small improvement in respiratory symptoms and physician visits. In the COPE trial analysis of the subgroup of patients with a FEV1 value less than 50 percent predicted (low FEV1 group) suggests that the improvement in time to first exacerbation due to fluticasone is driven by this group. In subjects who smoke at baseline and have normal to moderate airflow obstruction (GOLD Normal-Stage 2), ipratropium did not prevent the development of dyspnea, cough, and sputum, or respiratory hospitalizations at 3 years regardless of presence or absence of baseline respiratory symptoms.

Long-acting monotherapies provide similar reductions in COPD exacerbations among symptomatic individuals with severe to very severe airflow obstruction. There are differences in their adverse effects. Five trials compared monotherapy using either long-acting β agonists or inhaled corticosteroids versus combination therapy and versus placebo. Compared with placebo the absolute difference of having at least one COPD exacerbation was: 3.7 percent for long-acting beta agonists, 5.2 percent for inhaled corticosteroids, and 6 percent for combination therapy of long acting β agonists and corticosteroids. Combination therapy with LABA and inhaled corticosteroids did not significantly reduce exacerbations or mortality compared to corticosteroid monotherapy. The combination of short-acting anticholinergic plus short- or long-acting β agonist is not superior to short-acting anticholinergics alone. No studies are available to determine if adding long-acting anticholinergics to inhaled corticosteroids or β agonists reduces exacerbations or improves respiratory symptoms compared to monotherapy. Pulmonary rehabilitation provides a small improvement in clinical outcomes including respiratory health status measures during the period of the rehabilitation in individuals with respiratory symptoms and severe to very severe airflow obstruction.

In symptomatic patients with severe to very severe airflow obstruction the choice of pharmacologic agents depends primarily on costs and adverse effects because effectiveness is similar. In studies that compared different doses of the same drug treatment effectiveness did not vary. The primary demonstrated benefit of these interventions is in reducing exacerbations (and possibly hospitalizations) rather than an average clinically noticeable benefit in dyspnea. Exacerbations are relatively rare and it is difficult to assess whether an average patient is achieving clinical improvement. Thus, treatment should be continued even if patients do not report symptomatic improvement. This indicates that dose titration or modification is not beneficial. However, the long-acting inhaled anti-cholinergic agent, tiotropium, is superior to the short-acting anti-cholinergic, ipratropium, in individuals with moderate to severe respiratory symptoms and airflow obstruction.

Spirometry may be useful for identifying a threshold value to initiate treatment in adults with bothersome respiratory symptoms (especially dyspnea and frequent exacerbations) with inhaled corticosteroids, bronchodilators, or pulmonary rehabilitation. This threshold appears to be at a postbronchodilator FEV1 below approximately 50 percent predicted (GOLD Stage 3,4). There is evidence to suggest that monitoring subjects spirometric response to therapy or change over time while on therapy does not improve outcomes. Limited data suggest that an individual's response to inhaled bronchodilators is quite variable and that spirometric response to treatment is not associated with improvement in clinical outcomes. An individual's spirometric change over time is also quite variable and except for identifying a spirometric threshold to initiate therapy does not improve treatment outcomes. Modification of therapies in the absence of adverse effects or compliance issues is not supported by evidence.

Spirometry for Prognosis

Spirometry provides independent prognostic value regarding health status, rate of exacerbations, morbidity, and mortality. However, degree of dyspnea appears to be a better predictor of mortality than FEV1 and a multidimensional grading system that assessed body-mass index, spirometry, dyspnea, and exercise capacity (the BODE index) predicted death better than spirometry alone. Baseline spirometry predicts rate of spirometric decline over time in male smokers. The probability of survival at 28 months of followup in subjects with established COPD was 90 percent and 75 percent in subjects with ATS-1995 Stage I, II, and III disease. Four factors, when combined, provide an index that predicted the risk of death better than FEV1 alone. These include (B) body mass index; (O) airflow obstruction; (D) dyspnea and (E) exercise capacity on 6-minute walk. The presence of current respiratory symptoms is a better predictor than spirometric value of having future respiratory symptoms. Subjects with chronic sputum production and normal spirometry (Stage GOLD 0 condition) are not at higher risk for developing airflow obstruction than individuals without COPD. Over half of these GOLD 0 subjects did not have sputum production at 10 years of followup.

Estimating the Number Needed to Evaluate with Spirometry and Symptom Assessment

The number that would need evaluation by spirometry and symptom assessment to provide clinical benefit was estimated based on data from NHANES III, as well as efficacy data from intervention trials. If a primary care clinic population was comprised of 10,000 adults with similar characteristics as NHANES III respondents (47 percent never smokers) then approximately 6,588 would undergo spirometric testing for either the presence of symptoms or because they were judged to be at increased risk due to smoking status. Thirty-nine “never smoking” adults (0.8 percent), 42 “previous smokers” (1.7 percent), and 48 “current smokers” (1.6 percent) have both respiratory symptoms and airflow obstruction severity (approximately GOLD Stage 3,4) that might make them candidates for COPD-specific treatment in addition to smoking cessation and influenza vaccination (129/10,000 or 1.3 percent of the total clinic population). Using the pooled efficacy data from treatment trials of inhaled bronchodilators or corticosteroids, it can be estimated that approximately 2,043 “never smoking” adults, 960 “previous smokers,” and 1,010 “current smokers” would have to have respiratory symptom and spirometry evaluation with subsequent selective treatment to prevent one subject from having one or more COPD exacerbations over a 6–36 month period. A total of 7 out of 10,000 primary care adults would have prevention of one or more COPD exacerbations. The pooled efficacy data indicate that treatment would not reduce mortality (except for oxygen in subjects with resting hypoxemia). The average improvement in respiratory health status among treated subjects would not be of clinical significance though approximately 18 of these 129 treated patients (14 percent) would have a clinically noticeable improvement in health status. Treatment with combination therapy would not be superior to inhaled monotherapy, and, on average, therapy in asymptomatic individuals or those with mild to moderate airflow obstruction would not improve or prevent symptoms. If subjects with moderate airflow obstruction (approximately GOLD Stage 2) were also assumed to benefit in a similar fashion, then approximately 529 adults would be candidates and 32 (0.3 percent) would have reductions in exacerbations and 76 subjects (0.8% of all adults) would have noticeable improvements in respiratory health status.

The number of eligible candidates for COPD therapy would increase from these NHANES estimates if spirometric testing and symptom assessment were limited to middle age or older adults (e.g., age 50 or greater) because the risk of airflow obstruction and symptoms increases with age. However, our estimates are otherwise optimistic because they assume that adults with severe airflow obstruction and any respiratory symptom including symptoms limited to wheeze, cough, or sputum production would benefit in an amount similar to subjects enrolled in treatment trials who had known COPD, dyspnea, and experienced frequent exacerbations. Cost associated with testing and treatment would be large and include bronchodilator testing not typically performed in primary care settings as well as assessing individuals with risk exposure beyond a personal smoking history. Costs could be reduced considerably with no apparent reduction in clinical effectiveness by targeting spirometry to individuals with respiratory symptoms, especially current and former smokers 40–50 years of age or older who have bothersome dyspnea. Spirometry could improve treatment costs if it led to treatment being targeted towards individuals with bothersome respiratory symptoms, especially dyspnea and exacerbations who have severe to very severe airflow obstruction. The existing evidence indicates that spirometry is unlikely to provide more than a small improvement in sustained smoking abstinence and that it is of limited clinical use in predicting subsequent smoking cessation rates.

Spirometric testing in combination with clinical examination is useful in symptomatic individuals for improving diagnostic accuracy compared to clinical examination alone. It helps to ensure that COPD-specific therapy is not initiated in individuals who do not have at least moderate airflow obstruction. Among adults with bothersome respiratory symptoms, spirometry may be useful for determining at what threshold level of airflow obstruction to initiate therapy. Spirometric testing in symptomatic adults could improve physician use of COPD-specific treatments to subjects likely to benefit (i.e., those with bothersome respiratory symptoms and severe to very severe airflow obstruction) while reducing the cost and side effects of unnecessary or ineffective treatments. In subjects with COPD acute spirometric response to bronchodilators is variable, potentially misleading, does not predict long-term spirometric decline, and is not associated with clinical response to treatment. Responsiveness to bronchodilators in younger adults with respiratory symptoms is likely to be beneficial if asthma is suspected. However, it is not useful for assessing clinical response to therapy or determining treatment options in subjects with COPD. Periodic spirometric testing to monitor and modify treatment has not been evaluated. However, this method is unlikely to be beneficial because different types of pharmacologic management have similar efficacy, relative treatment effectiveness cannot be determined by baseline spirometry or spirometric response to treatment, there is considerable intra-individual variation in spirometric results, pharmacologic therapies do not alter the rate of spirometric decline, clinical outcomes are not associated with spirometric response to therapy, and dose titration or combination therapy is not more effective than fixed dose monotherapy. Choice of therapy should be determined by patient preference, cost, and adverse effects.


Irreversible airflow obstruction as determined by spirometry in individuals with respiratory symptoms is the most widely established criterion for establishing the diagnosis of COPD. It is useful for determining whether treatment is likely to be beneficial and estimating prognosis. While respiratory symptoms are quite common in adults, the vast majority of these individuals do not have clinically significant airflow obstruction, and many who have moderate or worse airflow obstruction do not have bothersome respiratory symptoms. Spirometry in combination with clinical examination improves diagnostic accuracy in adults with respiratory symptoms compared to clinical examination alone. It is useful in determining the presence and severity of airflow obstruction prior to establishing a diagnosis or initiating disease-specific therapy. Spirometry is likely to demonstrate that some adults with a previous clinical diagnosis of COPD do not have airflow obstruction and should not be labeled or receive COPD specific therapy. Increased use of spirometry primary care settings for adults with bothersome respiratory symptoms, especially dyspnea, would identify the small percentage of individuals with severe to very severe airflow obstruction who have not received a clinical diagnosis of COPD and might benefit from disease specific therapies.

A strategy of conducting spirometric testing of all at-risk adults would require testing a large number of asymptomatic individuals or those with nonspecific and nonbothersome respiratory symptoms. It would result in considerable testing costs and health care personnel time and resources. Some individuals with abnormal airflow will have other medical conditions causing respiratory symptoms (e.g., heart failure). As criteria for defining disease expand, the number of adults labeled with disease markedly increases. If spirometric measures of airflow obstruction are used as sufficient criteria to establish disease, then the vast majority of adults newly diagnosed by spirometric testing will be asymptomatic or have nondisabling respiratory symptoms. Some may be treated unnecessarily or not receive effective interventions for other medical conditions.

Spirometric testing is unlikely to alter smoking cessation rates or be useful for monitoring response to therapy or modifying treatments. The average benefits of therapy are primarily seen in those with severe to very severe airflow obstruction (FEV1 < 50% predicted, GOLD Stage 3,4 disease) and related to reduction in COPD exacerbations. Treatment, beyond smoking cessation and influenza vaccination, does not prevent symptom development in asymptomatic individuals over a 3 year period. None of the interventions other than smoking cessation alter the rate of decline of spirometry and clinical response to treatment is not associated with spirometric changes. Spirometry provides independent prognostic value for predicting respiratory and overall morbidity and mortality in individuals with established COPD. However, the degree of dyspnea appears to be a better predictor than spirometry. Patients with normal spirometry and chronic sputum production (GOLD 0) do not appear to be a group “at increased risk” for development of clinically significant airflow obstruction.

Future studies are required to determine if spirometry improves smoking cessation rates, if treatment effectiveness in patients with established COPD varies according to an individual's baseline or followup spirometric value, and if treatment is effective in individuals with airflow obstruction who do not report respiratory symptoms.


Our report has limitations. We used NHANES III population-based spirometry data performed without bronchodilator testing to estimate prevalence of airflow obstruction, symptom status, and previous reported clinical diagnosis of COPD. We did not assess the benefits or harms of spirometry (including use of bronchoresponsiveness) for other respiratory conditions including asthma and restrictive lung disease. NHANES is a national probability sampling of adults and may not directly reflect the population to be evaluated in primary care clinics. Many NHANES respondents were younger and thus at lower risk of having COPD. We were unable to determine prevalence of specific respiratory symptoms by postbronchodilator GOLD stage category in subgroups of interest (smoking status, age, race, gender). We used estimates derived from spirometry done in the absence of bronchodilators for the total population sample. Furthermore, we could not determine the number of individuals with a diagnosis of COPD by GOLD stage nor the accuracy or methods used for diagnosis. Our report was limited to subjects with COPD. We did not assess patients with asthma or restrictive lung disease.

Failing to find a benefit that spirometry improves smoking cessation does not mean that a benefit does not exist. Available RCT evidence was limited and of poor quality. There was also no evidence that spirometric testing led to adverse effects such as lower smoking cessation, poorer quality of life, or misuse of smoking cessation interventions. As noted, even a relatively small improvement in smoking cessation could have large population benefits due to the high prevalence and large and diverse adverse health effects of smoking.

Data regarding COPD-specific treatments typically provided outcomes for the whole population enrolled and did not report results for subgroups according to respiratory or spirometric status. However, results from the few studies that provided this information suggest that interventions are most effective in individuals with the combination of activity limiting respiratory symptoms and severe to very severe airflow obstruction. While average improvement in respiratory symptoms was less than considered clinically significant (especially for dyspnea) it is likely that individual patient's response to therapy varies. Secondary analyses determined that some individuals found a clinically significant improvement in respiratory health status and likely dyspnea, cough, and sputum production. However, based on the available data, interventions appear to be most effective at reducing exacerbations rather than the patient's perception of dyspnea that most affects day-to-day health status. While mortality was not improved with these interventions, studies were typically of short duration and the confidence intervals around the point estimate for effectiveness were wide. Clinically significant improvements in mortality due to interventions beyond oxygen therapy may exist.

Our report is not a formal cost effectiveness analysis. A previous cost-effectiveness analysis concluded that inhaled corticosteroids were cost effective in subjects with ATS Stage 2–3 disease (GOLD Stage 3,4). We used the available information from NHANES regarding airflow obstruction performed in the absence of bronchodilator testing and respiratory symptom status prevalence assessed by responses to survey questions, optimistic assumptions regarding treatment efficacy, and conducted sensitivity analysis incorporating treatment of subjects with moderate airflow obstruction (FEV1 50–80 percent predicted). Information from population based studies indicated that failure to use postbronchodilator spirometry likely resulted in only a small misclassification of subjects. We also employed widely available estimates for costs of one time spirometry and pharmacologic interventions. Our cost estimates did not include the medical and societal costs for COPD exacerbations or hospitalization that might be prevented. Nor do they consider the benefits that might occur by targeting COPD treatments to individuals who have both bothersome respiratory symptoms and severe to very severe airflow obstruction.

Future Research Needs

  • Conduct randomized trials to determine if spirometry in primary care office-based settings results in improved rates of smoking cessation and long-term abstinence. Studies should evaluate rates of smoking cessation; types of smokers likely to benefit (based on smoking intensity, readiness to quit, symptom, and spirometric status); types of smoking cessation counseling; and pharmacologic interventions as well as other interventions specifically for airflow obstruction or respiratory symptoms. A conceptual trial design is shown in Figure 18 on page 99.
  • Determine if inhaled treatments prevent the development of respiratory symptoms and/or improve health status in individuals with airflow obstruction not reporting bothersome respiratory symptoms. Studies should evaluate subjects across the full spirometric range of airflow obstruction severity and be at least several years in duration.
  • Conduct randomized trials to determine if therapeutic thresholds exist for specific interventions according to spirometric and symptom status (especially in subjects with mild to moderate airflow obstruction).
  • Conduct randomized trials to determine if therapy based on spirometric level, response to therapy, or change over time provides better clinical outcomes compared to clinical examination, fixed-dose, or symptom-driven therapy.
  • Improve physician recognition of respiratory symptoms, especially dyspnea, that are compatible with COPD and may benefit from earlier detection via a combination of clinical history, physical examination, and measures of airflow (spirometry).
  • Conduct long-term longitudinal cohort studies to better assess the associations between spirometric values, symptom status, and clinical outcomes, especially in individuals with mild disease or GOLD 0, or those who are asymptomatic.
  • Estimate the costs, adverse effects, time, and personnel involved with spirometry for casefinding, diagnosis, and management including the possible harms from COPD-specific therapies or disease labeling.
  • Identify better diagnostic markers for clinically significant COPD.
  • Develop new therapies that can improve clinical outcomes, especially dyspnea, as well as alter the decline in spirometry.
Figure 18. Potential study design of a randomized trial to evaluate the impact of spirometric testing to alter smoking cessation rates.


Figure 18. Potential study design of a randomized trial to evaluate the impact of spirometric testing to alter smoking cessation rates.


  • PubReader
  • Print View
  • Cite this Page

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...