Main findings
Using a simple 20/40 rule based solely on age and smoking history identifies a high percent of individuals (74 % in the initial population and 67 % in the POC group) with lung function abnormalities (both obstructive and restrictive spirometry patterns). Further those meeting the 20/40 rule who have lung function abnormalities have an increased risk of all-cause mortality and a high rate of reported coronary heart disease, compared to individuals with normal spirometry.
Using the simple 20/40 rule in primary care could facilitate the appropriate use of spirometry evaluation and support the diagnosis of COPD in up to 40 % of patients with limited need for further testing. This spirometry testing could be accomplished in the primary care office [
28‐
30] allowing rapid diagnosis and initiation of appropriate COPD management. For the 30 % of individuals with a restrictive lung pattern, the addition of lung volumes and perhaps other elements of full pulmonary function testing may be required to determine the accuracy of the restrictive pattern and assess possible associated diagnoses.
Our evaluation of the 20/40 rule confirmed the previously reported over and under diagnosis of COPD within this population. Of those who remembered receiving a COPD or COPD plus asthma diagnosis, 11.3 % had normal spirometry assessment suggesting that a different condition may be causing their symptoms—potential over diagnosis. Conversely, 53.2 % of the individuals with an obstructive pattern did not believe they had received a COPD diagnosis (under diagnosis) likely prohibiting them from receiving any of the therapies known to improve symptom burden or prolong life [
31].
Relationship to the literature
Use of the 20/40 rule goes beyond simply “screening” for risk of COPD and includes the opportunity to evaluate for the presence of restrictive spirometry patterns which may also be seen with COPD or with other conditions in the lungs and other organ systems. Spirometric results from our study groups provide important new data on rates of restrictive spirometry patterns in long-term smokers. Published values for prevalence of a restricted pattern on spirometry in unselected populations vary from 10.3 to 12.3 % [
8,
32]. Data from the Multi-Ethnic Study of Atherosclerosis (MESA) [
33] showed that among smokers the prevalence of a restricted pattern on spirometry was 10 %, increasing by 8 % (95 % confidence interval 3–12 %) for each 10 pack-years smoked. The prevalence of a restricted pattern in MESA was 16 % for the > 20 pack year cohort, who had a smoking history significantly lower than our OME group (36 pack years, 95 % CI 27–50 vs 51.9 ± 27.8 pack years). Allowing for the additional increase in prevalence of restriction for each pack year smoked found in MESA, the 28.4 and 34.2 % prevalence of restrictive spirometry in our study populations is consistent with the MESA data and provides a useful population estimate in longer-term smokers aged 40 years and older.
Under recognition of lung function abnormalities in smokers is common. But most studies focus on the airflow obstruction reported to occur in up to 50 % of long-term smokers [
34] for which under recognition is well documented [
4,
16,
17]. However, the under-recognition of a restricted pattern in smokers is less well studied, but is clearly evident from our results which also highlights the potential consequences of under-recognition of both obstructive and restrictive lung function abnormalities. With airflow obstruction, these consequences include under treatment, that may lead to increased exacerbation frequency, symptoms, decreased quality of life, and increased medical costs [
35]. Restrictive spirometry patterns are also seen in smoking related conditions such as fibrotic interstitial lung disease, (RBILD or respiratory bronchiolitis with interstitial lung disease) [
10,
36] muscle weakness [
11], heart disease [
12], obesity and the metabolic syndrome [
13]. A restricted spirometry pattern has been reported to be associated with numerous cigarette and non-cigarette related co-morbidities, including hypertension, type II diabetes, atherosclerosis, cardiovascular disease, and all-cause mortality, in a manner that is statistically independent of confounding variables, such as diabetes, obesity, and smoking history [
37‐
47]. Our data shows that the prevalence of coronary heart disease and all-cause mortality is significantly greater for smokers age ≥ 40 years with a restricted pattern than for those with normal spirometry. In fact, there was no significant difference in coronary heart disease prevalence and all-cause mortality, between subjects with an obstructed and those with a restricted pattern on spirometry. Under-recognition of both obstruction and restriction in smokers, age ≥ 40 years, is a common occurrence with significant health consequences.
Other approaches have been developed for selecting patients at high risk of airflow limitation. These include use of questionnaires [
48‐
52], risk prediction models [
53] and handheld flow meters [
54,
55]. Most questionnaires developed for COPD are related to patient outcomes of quality of life. The COPD population screener (COPD-PS) is a 5 question tool with a positive predictive value of 56.8 % and negative predictive value of 86.4 % [
50]. van Schayck et al. [
52] and Calverley et al. [
56] developed a population-based screening questionnaire for COPD using NHANES III data. Price and coworkers [
51] published an 8-item COPD questionnaire in patients with a positive smoking history that included items related to age group, body mass index, pack-year history, and symptoms. Finally, Freeman and colleagues [
57] utilized age, cough, dyspnea, and wheezing in a questionnaire to identify patients with COPD in a primary care setting who had a positive smoking history, history of respiratory medication use, or of asthma. The COPD-PS differs from these previous questionnaires because it can be used regardless of history of respiratory problems or smoking and it contains a disease-impact item. Population screeners when used in combination with peak expiratory flow measurement (PEF) however have been shown to add little to PEF alone [
54,
58,
59]. A new questionnaire developed by the High Risk-COPD Screening Group is currently undergoing testing in combination with PEF [
60]. Preliminary data is impressive [
61]. A single published risk prediction model utilizing sex, socioeconomic status and previously recorded asthma diagnosis has performed well in an initial derivation cohort of 480,903 and validation cohort of 247,755 subjects with area under the receiver operating curve of 0.8 [
53].
Strengths and limitations
Limitations in this study include the use of a fixed threshold cut-off for the definition of obstruction (FEV
1/FVC < 70 %). This fixed ratio threshold has been demonstrated to introduce age related bias [
62]. Use of a fixed ratio can result in misclassification in more than 1/4
th of tests [
63] and over diagnosis of airflow obstruction in older subjects [
64]. However, the further qualification of FEV
1 < 80 % placed on our obstruction definition reduces over diagnosis significantly [
65]. FVC and FEV
1 were assessed before bronchodilator, allowing the possibility that a portion of the obstructed spirometries might display reversibility, suggestive of a purely asthmatic component, rather than chronic obstructive pulmonary disease [
66].
Implications for policy, practice and research
The goal of this study was to increase use of spirometry in the primary care setting. Only half of primary care providers routinely use spirometry in their practice of medicine [
67]. Among primary care physicians spirometry use was associated with agreeing that the data are necessary for accurate diagnosis, and believing that they were trained to perform and interpret the test [
68]. In view of these reports and others, we elected to use a simple rule that might increase use of spirometry in the primary care setting without significantly sacrificing accuracy. We wanted to show that use of this
simple rule could enhance diagnostic yield that had a clinical impact such as increased risk for CHD and death. Furthermore, it distinguished individuals that would benefit from lung volume measurement. Using the 20/40 rule resulted in detection of an abnormality on spirometry in nearly 70 % of individuals tested.