Background: Chronic obstructive pulmonary disease (COPD) is the third leading cause of death in the United States.
Purpose: We conducted this systematic review to support the U.S. Preventive Services Task Force in updating its recommendation on screening for COPD. Our review addressed eight questions: 1) Does screening for COPD in asymptomatic adults age 40 years and older with prebronchodilator spirometry improve health-related quality of life or reduce morbidity or mortality? 2) Do prescreening questionnaires reliably identify high-risk asymptomatic adults who are more likely to test positive on screening for COPD? 3) What is the test performance of screening pulmonary function tests in predicting diagnosis of COPD in asymptomatic adults, based on confirmation with postbronchodilator spirometry to identify fixed airflow obstruction? 4) What are the adverse effects of screening for COPD with prescreening questionnaires or screening pulmonary function tests? 5) Does identifying asymptomatic adults with fixed airflow obstruction through screening improve the delivery and uptake of targeted preventive services? 6) What are the adverse effects of COPD screening, including the impact of targeted preventive services in this population? 7) Does treatment of asymptomatic adults identified with mild to moderate COPD through screening improve health-related quality of life or reduce morbidity or mortality? 8) What are the adverse effects of COPD treatment in this population?
Data Sources: We searched MEDLINE, PubMed Publisher-Supplied Records, and the Cochrane Collaboration Registry of Controlled Trials to identify literature published from January 2000 or 2005 through January 2015, depending on key question. We supplemented our searches with reference lists from the previous review, relevant existing systematic reviews, suggestions from experts, and Clinicaltrials.gov to identify ongoing trials.
Study Selection: Two investigators independently reviewed identified abstracts and full-text articles against a set of a priori inclusion and quality criteria.
Data Analysis: One investigator abstracted data into an evidence table and a second investigator checked these data. We qualitatively synthesized the data for each key question; quantitative synthesis was not appropriate due to heterogeneity and few trials for any given intervention and outcome.
Results: We identified three externally validated COPD questionnaires, the COPD Diagnostic Questionnaire (CDQ), the Lung Function Questionnaire, and the COPD Population Screener. The CDQ, an eight-item self-administered, symptom- and risk factor–based questionnaire, was externally validated in two good- and three fair-quality diagnostic accuracy studies (n=3,048). Validation populations recruited exclusively or at least partly from primary care practices, excluding participants with known lung disease, and most studies recruited at least half of their participants with a smoking history. Most external validation studies reported that a CDQ score of greater than 16.5 had a sensitivity in the low 90 percent range and specificity in the high-30 to mid-40 percent range for diagnosing spirometrically-confirmed COPD. The Lung Function Questionnaire, a five-item self-administered, risk factor– and symptom-based questionnaire, was externally validated in one fair-quality multicenter primary care study (n=1,288) in the United States of ever smokers with at least a 10 pack-year exposure. The study reported a high rate of unacceptable spirometry (31%), as well as an estimated sensitivity of 88 percent and specificity of 25 percent. The COPD Population Screener, a five-item self-administered, risk factor– and symptom-based questionnaire, was externally validated in a single fair-quality population-based study (n=2,357) in a rural Japanese town and reported a sensitivity of 67 percent and specificity of 73 percent.
We identified two fair-quality Burden of Obstructive Lung Disease population-based studies of prebronchodilator peak flow. These studies used different index test and gold standard thresholds for defining COPD in both low- and high-index countries without exclusion of known COPD; these studies do not provide sufficient information to make conclusions regarding peak flow screening accuracy. We identified one good- and one fair-quality study of prebronchodilator microspirometry measuring the ratio of forced expiratory flow in 1 second to forced expiratory flow in 6 seconds (FEV1/FEV6) and reporting consistent sensitivities in the low 50 percent range and specificities in the 90 percent range. We identified one fair-quality study of postbronchodilator microspirometry measuring FEV1/FEV6 in a population of approximately half ever smokers, which reported a higher sensitivity (80%) and specificity (95%).
One fair-quality study examined a staged approach whereby the screening test was considered positive only if both the CDQ and FEV1/FEV6 tests were positive. Sensitivity and specificity were 72 and 97 percent, respectively, in the entire population and similar in a subset of smokers only.
Evidence of screening harms from diagnostic accuracy studies was limited; only false positives and false negatives associated with screening were reported, and few studies reported data for the calculation of number of missed cases.
We identified five randomized, controlled trials (RCTs) (n=1,620) addressing the effectiveness of COPD screening in influencing smoking cessation rates. Of the three RCTs reporting biochemically confirmed abstinence, only one fair-quality U.K. primary care–based RCT (n=561) reported a statistically significant difference in smoking cessation at 1 year, with a number needed to treat of 14; this trial measured the incremental value of adding “lung age” to standardized counseling. The other two underpowered RCTs of biochemically validated abstinence reported no difference or a nonstatistically significant trend favoring reduction in the spirometry group. No studies examined the effectiveness of screening to increase vaccination rates.
There were no treatment trials identified in screen-detected patients; thus, we included trials with either subanalyses of participants with mild to moderate COPD or trials in which the mean FEV1 percent predicted was 60 percent or greater. We identified a total of one good- and 13 fair-quality RCTs meeting these criteria providing analysis of mild to moderate COPD patients; two long-acting β-agonist (LABA) studies (n=3,174), one inhaled corticosteroid (ICS)-LABA combination study (n=1,097), five tiotropium studies (n=4,592), and six ICS studies (n=3,983). Overall, subanalyses were limited due to post hoc timing, underpowering for subgroups, lack of data to confirm baseline comparability for the subgroup, lack of interaction testing, and lack of control for confounders. However, available subanalyses suggest no benefit in all-cause mortality but a decrease in annual rate of exacerbations with LABAs, LABA-ICS, tiotropium, and ICS. Because absolute rates of exacerbations were less than 1 in patients with mild to moderate COPD, the clinical magnitude of this benefit is uncertain. Data were too limited to make conclusions regarding other patient-focused outcomes (e.g., exercise capacity, dyspnea, and quality of life).
We identified eight effectiveness RCTs reporting harms data, but few trials reported harms for any individual drug class, making conclusions about treatment-related adverse events challenging. Concerns about pneumonia and bone demineralization with ICS medications could not be confirmed because few trials reported these outcomes. U.S. Food and Drug Administration drug labels for the considered drug classes report side effects as generally mild, ranging from dry mouth and coughing to vomiting and pneumonia.
Conclusions: There is no direct evidence to quantify the benefits and harms of COPD screening with questionnaires or handheld spirometry, nor is there evidence to estimate the treatment benefits in screen-detected populations. The evidence gaps identified in this systematic review suggest that there is a need for future research examining the treatment benefit in asymptomatic screen-detected populations or populations with mild disease.