Family Practice Advance Access originally published online on November 24, 2008
Family Practice 2009 26(1):3-9; doi:10.1093/fampra/cmn084
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Is COPD a rare disease? Prevalence and identification rates in smokers aged 40 years and over within general practice in Germany
a Department of General Practice, Universitätsklinikum, Heinrich-Heine-Universität Düsseldorf, Moorenstrasse 5, D-40225 Düsseldorf
b Department of General Practice, University Witten/Herdecke, D-58448 Witten, Germany
Correspondence to Heinz-Harald Abholz, Department of General Practice, Universitätsklinikum, Heinrich-Heine-Universität Düsseldorf, Moorenstrasse 5, D-40225 Düsseldorf, Germany; Email: abholz{at}med.uni-duesseldorf.de
Received 11 March 2008; Accepted 8 October 2008.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionLimitations and strengthsDeclarationReferences |
|---|
Background. According to literature, COPD rates are high in spite of decreasing rates of main risk factors smoking and air pollution in developed countries. general practice is a good place to survey unbiased prevalence rates. Ten studies done in general practice over the last 20 years found prevalence rates among smokers between 13.1% and 92.1%.
Objective. Prevalence and detection rates of COPD in smokers in German general practice.
Methods. Twenty-eight of 34 invited and eligible GP surgeries in/around Duesseldorf, Germany, took part in the non-announced 2-day investigation of all smokers (
40 years) who visited the surgeries. Lung function test by hand-held spirometer, peak flow, sympton part of St George's Respiratory Questionnaire, and data on smoking habits were used. GOLD criteria for COPD were employed. GPs had to give their diagnosis not knowing the test results.
Results. Of 3157 patients attending the 28 surgeries, 538 were smokers. Four hundred and thirty-seven of these agreed to participate, 5 had to be excluded for medical reasons/unacceptable spirometry. Three hundred and ninety-eight patients have not been previously diagnosed with COPD or asthma. Thirty patients were disgnosed with COPD, making a prevalence of 6.9%, of which 15 patients were already known as having COPD.
Conclusion. Our result of low prevalence differs strongly from all other studies in general practice. Considering our study design which avoids selection bias found in nearly all other studies (no pre-announcement, no self-selection of patients or GPs, high participation rate and testing all patients), we strongly believe that our findings reflect the current situation of COPD in German general practice.
Keywords. COPD, detection rate, general practice, prevalence, smoking, spirometry.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionLimitations and strengthsDeclarationReferences |
|---|
Chronic obstructive pulmonary disease (COPD) has been described by the World Health Organization Global Obstructive Lung Disease (GOLD) initiative as a disease that is ‘characterized by airflow limitation that is not fully reversible’ and there are claims that this is on the increase. It is estimated that COPD will become the fifth largest cause of disability and the third most common cause of death in the world by 2020.1 Smoking is seen as the main risk factor contributing to the disease and is a habit shared by 80–90% of patients with COPD.2,3 Air pollution is another relevant risk factor.4,5 In developed countries, however, smoking rates and environmental pollution have been steadily declining over the last 40 years.
A systematic search in PubMed using the entry terms ‘[(COPD) OR (chronic bronchitis) OR (pulmonary emphysema)) AND [(prevalence) OR (cross-sectional studies)]’ (limited to people aged 19 years and over) showed that from 1986 to March 2007 over 120 studies were carried out worldwide in an effort to gauge the prevalence of COPD or associated diseases. If studies which regard occupational or environmental exposure to fumes and dust as risk factors are included, the number of projects is doubled. But only around 20 studies have been done in primary health care institutions and only 10 of them focussed on smokers. We did not include studies that had been only carried out within general practice using it only as a place for a spirometric investigation as, e.g., the Detection, Intervention and Monitoring of COPD and Asthma (DIMCA) study.6
GPs have often been accused of underestimating the number of patients with COPD because they tend to underuse spirometry in the detection of the disease's early stages.7 The 10 studies which took place in a primary healthcare setting and focussed on smokers found prevalence rates of between 13.1%8 and 92.1%.9 Information on the number of patients previously known to have COPD were presented in only two of these studies, with rates of undetected COPD of 65%10 and 63.2%,11 respectively. Only 4 of these 10 studies included patients without previously documented lung diseases.9,12–14
Taking into account that the vast majority of GPs in Germany use spirometry in their surgeries, we were interested to discover the prevalence and detection rates under such circumstances, where very little data existed hitherto. It is important to bear in mind that on average a German citizen sees a GP 12 times a year and that 90% of those in the selected age range of 40 years and over see a GP at least once a year.15 This makes epidemiological studies in German general practice quite representative for the whole population—especially when most of the patients asked by their GP agree to take part in one of the studies.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionLimitations and strengthsDeclarationReferences |
|---|
In total, 40 general practice surgeries, all of whom have been involved in student teaching at the Heinrich-Heine University (HHU) of Duesseldorf and are located in and around Duesseldorf, were invited by letter to take part.
In the end, only 28 of these surgeries actually participated, mostly because of limited time slots for investigation by the study team.
Patients
From February to April, 2007, every patient coming into the surgery over a period of 2 days who—after asking by the GP staff—was at least 40 years of age and a smoker was invited to take part in ‘a lung-testing examination’. Those willing to take part were registered by the GP staff using their in-practice identification number while a researcher was present in an observation capacity. The number of those refusing was also registered. Participants were given questionnaires concerning their smoking habits as well as the symptom section of the St George's Respiratory Disease Questionnaire (SGRQ).16 A smoker was defined as having smoked for more than 3 years during his/her lifetime and having quit in the last 3 years.
Pulmonary function testing
Participating smokers were examined by one researcher (CG) and a medical student (see Acknowledgement), both of whom were fully trained in lung function testing and the correct use of peak flow meters as well as in the interpretation of test results.
All participants, with the exception of known COPD or asthma patients, underwent three tests using the peak flow meter as well as three lung function tests using the handheld spirometer (Viasys, Jaeger, Höchberg, Germany, Flow-Screen, 2005). We took the best results according to ECSC93 criteria.17 The spirometers were calibrated every day according to the manual.
Patients with test results indicating pulmonary obstruction (measured by the ratio between forced expiratory volume in one second and forced vital capacity [FEV1/FVC] ratio below 70%) were re-tested 15 minutes after exposure to a rapid acting beta-mimetic (salbutamol, 0.2 mg) in a bid to exclude asthma as a cause: a reversibility of more than 15% or at least 200 ml FEV1 was taken as indication of a reversible obstruction. According to GOLD guidelines,18 COPD is defined as an irreversible airway obstruction using post-bronchodilator FEV1/FVC < 70%.
At the end of each day, GPs were asked to give their opinion concerning the possibility of pulmonary disease in participants without knowledge of the test results.
The study has been approved by The Human Ethics Committee of HHU Duesseldorf. There was no financial support for the study other than that provided by the HHU.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionLimitations and strengthsDeclarationReferences |
|---|
Throughout the duration of our test periods, 5020 patients contacted their health centres. In all, 3157 of these were 40 years of age or over. In total, 538 (17%) identified themselves as smokers, of which 437 were willing to take part (81% participation) (Fig. 1).
|
Five patients had to be excluded from the lung function test because they had experienced high-grade congestive heart failure, recent operations, schizophrenia, or difficulties in performing an acceptable spirometry. In total, 432 patients (80.3%) were fully examined.
In all, 398 of these (92.1%) who had no previous diagnosis of asthma or COPD and had never taken pulmonary medication underwent spirometry. The 34 patients with known COPD or asthma were investigated via questionnaires only.
The percentage of females (51.6%) was more or less equivalent to the percentage of males (48.4%) in the test sample of 432 patients. Their ages ranged from 40 to 84 years and 93.3% of all participants had been smoking for more than 10 years.
Altogether we found 30 patients (13 female and 17 male) who had COPD. This equates to a prevalence of COPD in 6.9% of the whole study group, i.e. all those taking part in the study. In the subgroup of 398 patients with no previous record of asthma or COPD who underwent spirometry, the prevalence of newly found COPD was 3.8% (15 of 398 patients). This equates to a detection rate by GPs of COPD of 50% (15 of the 30 patients were known to have COPD by the GP).
All 15 ‘newly detected’ COPD patients showed mild or moderate COPD (GOLD Stage 1 with FEV/FVC < 70% and FEV1 > 80% and GOLD Stage 2 defined as FEV1/FVC < 70% and FEV1 between 50% and 80%) (see Table 1).
|
Two of those patients previously known to have COPD had previously been diagnosed by their doctors as having ‘a combination of asthma and COPD’ and were not included in the analysis.
Evaluating the symptom section of the SGRQ, it was clear that only 4 of the 15 newly detected COPD patients showed symptoms of a chronic cough, whereas 9 of the 15 previously diagnosed COPD patients had a chronic cough (Table 2). Dyspnoea was found in only 3 of the newly detected COPD patients, but in 12 of the previously diagnosed patients. Six of the newly detected but only 3 of the COPD patients known to their GP had no chronic symptoms at all—according to the answers given in the SGRQ.
|
And those patients previously diagnosed have a larger number of symptoms when they display symptoms (more details in Tables 1 and 2).
Moreover, 19 of the participants were known to have asthma. In addition to these, we detected eight male and two female patients with previously undetected asthma. The differences in the asthmatic symptoms between the known and the newly detected patients are similar to those indicated by the COPD patients in the SGRQ. Approximately half of the newly detected asthma patients suffer from chronic cough and dyspnoea (data not shown here).
In all, 36 patients had a decreased FEV1 (<80%), 28 of whom had a decreased FVC (<80%) but ‘a normal FEV1/FVC (>70%)’. These patients did not meet the criteria for COPD or asthma and were excluded from the study.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionLimitations and strengthsDeclarationReferences |
|---|
We found a low prevalence of COPD in the sample group. There are no other comparable studies that have been based in German general practices. The 10 general practice-based studies carried out in other developed countries have shown substantially higher rates. We believe that this discrepancy is due to three mechanisms: a highly selective process when recruiting participants, relatively low participation rates19—often not stated,11,12 and different definitions of COPD,20 which failed to use the GOLD criteria.
The study from Belgium by Vandevoorde et al.11 using the same GOLD criteria for COPD as we employed showed a high prevalence (46.6%). Of the 68 COPD patients, 25 were known to have COPD, an identification rate of 36.8%. This particular study design resulted in a highly selective participation: From a total of 5755 patient contacts (within a 2-month period with six GPs!), only 166 patients were found who met the inclusion criteria and only 121 underwent spirometry.
Van Schayck et al.19 recruited patients from two semi-rural practices in The Netherlands and randomly selected 651 smokers between the ages of 35 and 70 years. He only tested 201 (30.9%) of these patients using spirometry, 169 of whom had an ‘acceptable’ spirometric curve. The criterion used to define ‘being at risk of developing COPD’ was a FEV1 < 80% based on the age, gender and height adapted norm. This definition is different from that used in our study and all other current definitions of COPD. In the group of participants aged 40 years and over, 26 (21%) had a decreased FEV1 result. However, there is no information on how many people had been previously diagnosed with COPD.
Geijer et al.13 contacted by letter all male patients aged between 40 and 65 years with no previous diagnoses of lung disease who were enlisted with GPs in Ijsselstein (The Netherlands). In total, 69% (2596 of 3763) responded, 37.7% (978 of 3763) of whom were currently smokers—a probably selective sample having a particularly high rate of smokers for The Netherlands. Of these, 918 met the inclusion criteria (only cigarette smokers) and 805 (87.6% of the eligible) underwent spirometry. The study identified that 29.9% of the sample group had COPD according to the GOLD criteria.18
This design based on inviting patients for a test increases the likelihood of selection bias because of the potential for higher participation by patients with symptoms and patients who are aware of risk factors (e.g. smokers).
An even more particular selection process seems to have been employed in the study by Frank et al.,21 where only the participants of a previously issued Respiratory Health Questionnaire Study could take part so long as they were also a ‘current smoker and/or reported four or more of the six symptoms or risk factors for obstructive airway disease'. As a result, 2646 patients from two health centres who fulfilled these criteria were invited to a lung function test. Only 871 (32.9%) attended and 825 underwent an adequate spirometry; 163 had COPD according to the GOLD criteria, a rate of nearly 20% in a highly selective sample group of smokers aged 30 years and over.
The Swedish study carried out by Stratelis et al.22 used advertising to invite volunteers for a lung function test in their primary health care centres. In total, 512 smokers aged 40 years and over were tested and a COPD prevalence of 29%, according to the GOLD criteria, was found. It is not stated, however, how many of these patients had been previously diagnosed with COPD. With regard to selection bias, the authors report that according to regional statistics, 27% of the invited 5332 people are expected to be smokers, but only 512 (9.6% of the invited) participated.
The study of Dales et al.8 in Canada presented a post-bronchodilatator prevalence of 13.1% COPD using the GOLD criteria; 26.5% had been previously known to have obstructive lung disease but this figure also includes asthma sufferers. They invited all primary healthcare practices in the Ottawa area to participate and the first eight surgeries who agreed were included in the study. Every patient aged 35 years and over who had smoked at least 20 packs of cigarettes during the course of their lifetime (a potentially very heterogeneous group) was invited to participate. In all, 1046 patients took part and 1034 underwent spirometry with ‘an acceptable’ result. No information regarding the participation rate was provided, and therefore, we cannot exclude the possibility of selection bias.
Price et al.14 found a COPD prevalence of 16% in Denver (CO) and 21.9% in Aberdeen (UK). Participants were randomly selected from primary healthcare practice rosters and invited by letter after admitting to a smoking habit and no previously known lung diseases. In all, 3161 invitations were mailed in Aberdeen and 14 200 in Denver. Only 1631 (9.4%) responded and were considered eligible and 1562 (8.9% of the invited) completed assessments. COPD was diagnosed according to the GOLD criteria.
The study from Clotet et al.12 showed a COPD prevalence of 22% in smokers aged 40–76 years in one Spanish village. Of the 1302 inhabitants in this age range, only 177 (13.6%) met the inclusion criteria—i.e. no previously diagnosed pulmonary disease and having actively smoked for at least 10 years; 164 completed the assessment in which the GOLD criteria were used. This means that the 1125 (1302 minus the 177 taking part) excluded patients were either non-smokers or had previously been diagnosed with pulmonary disease, which leaves us with an extraordinarily low rate of smokers.
Díaz et al.10 in their study from Toledo (Spain) presented a high response rate from invited participants—883 of 1003 (88.1%). The participants had to be at least 40 years old and registered with a GP; 442 (50.3%) of the 883 were smokers or ex-smokers and 427 underwent spirometry. COPD prevalence was detected in 16.4% of this group according to the GOLD criteria; 65% of patients with COPD had never previously been diagnosed with COPD. This study would appear to have no participation bias.
In a French study by Piperno et al.,9 the extremely high prevalence rate of 92.1% is probably down to the fact that GPs were allowed to select three patients who had not necessarily been previously diagnosed with COPD or asthma. Furthermore, prevalence calculation in this instance was based on the GPs’ judgement, not the GOLD criteria (only 24% of male and 18.7% of female participants underwent spirometry; peak flow measurement was carried out in 87.7% of patients).
In summary, the relatively high prevalence rates in the studies mentioned above in comparison to our own findings might be due to the different selection processes and their impact on the probability of a bias. In our study, there was no public announcement of the study before it began; therefore, no patient selection could take place, i.e. patients could neither decide to visit the practice nor avoid a visit because of the study. The only other studies avoiding this selection are those carried out by Vandevoorde et al. But we had a participation rate of 81% of smokers in contrast to only 22%—according to our calculations—in Vandevoorde’s case. As far as Dales’ and Van Schayck’s participation rates are concerned, we have no information.
In the study of Diaz, a selection by invitation could have taken place, even when the participation rate of nearly 90% makes it probably less important.
All other studies carried out in a general practice environment had even stronger selection criteria (e.g. taking only those who were ‘positive’ on a questionnaire) and this—by definition—resulted in the exclusion of potential participants.
Additionally, some studies employed definitions of COPD which would increase the likelihood of a COPD diagnosis (see study descriptions above).
Comparable studies which take place outside the environment of a general practice always suffer from a process of self-selection among participants. This is unavoidable because there is no trusted doctor trying to persuade them to participate, only an unknown study group. So even those community studies with unusually high participation rates, e.g. the study by Schirnhofer et al.23 or the whole BOLD study,24 have participation rates of only 60%.
|
Limitations and strengths |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionLimitations and strengthsDeclarationReferences |
|---|
We will consider possible bias within our study:
- Inviting only GPs who are involved in teaching is ‘a positive selection of interested and highly educated GPs’. In all probability, this would not have an impact on the low prevalence of COPD that was detected. It could explain the high identification rate within our sample. Although with regard to the identification rate, we do not think that there is a difference between the GPs who participated in our study and the general situation in German general practice, where most surgeries have spirometry: In Germany, about 80% of all GPs offer spirometry on a fee-for-service basis.
- Even with a high participation rate of 81% in our non-announced study, we ‘could have missed a number of symptomatic smokers’. A self-selection based on ‘avoiding’ the test could also have taken place when patients arrived in surgery. But this would also be a problem common to all other studies, seven of which were announced before testing took place, allowing for the possibility of even earlier and stronger self-selection.
- ‘It is possible that some smokers did not identify themselves in order to avoid testing’. Again this would be a problem common to all studies. The rate of identified smokers in our sample is close to the rate of smokers in this age group as a whole, according to a German study (22%).25
- Some of the COPD patients might have been ‘wrongly diagnosed as asthma patients’ by their GPs, leading to exclusion from spirometry according to our design of the study. But even in the unlikely event that all the patients that were previously diagnosed and treated for asthma by their GP have been incorrectly diagnosed, the prevalence rate for COPD is still fairly low, just reaching 10%.
- We used a ‘handheld spirometry’ device but we did not follow American Thoracic Society (ATS)-proposed ‘three to eight measurements’ when using spirometry, but a maximum of three. We did this for several reasons: we were not able to hire a more expensive non-handheld device and transport it to the various surgeries. Most of the other studies either also used handheld devices or do not report on the system used at all.
We also seem to be comparable to the other studies in terms of using the three measurements instead of the maximum eight. Additionally, it was demonstrated in a pilot we carried out for testing feasibility the number of patients willing to do more than three ‘trials’ was very small. We must remember that the test had not been announced beforehand so the patients came into surgery for appointments with their GPs or for getting laboratory tests or prescriptions. In summary, we put more value on a higher number of participants than the exactness suggested by ATS.
And we have to consider: with only three instead of eight measurements, we can only have increased the rate of COPD artificially, not reduced it.
Taking all these points into account, we firmly believe that the low prevalence rate of COPD within our study is an accurate reflection of the current situation in Germany. We also think that all the comparable studies have stronger selection biases and, therefore, produce higher prevalence rates. Our findings are in accord with GPs’ experiences in their everyday work, i.e. COPD is decreasing. That would indicate that after many years of combined efforts, public health measures like the Clean Air Act and Anti-Smoking Policy have resulted in a decrease in a widespread and debilitating disease in Germany—and, perhaps, in other European countries as well.
|
Declaration |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionLimitations and strengthsDeclarationReferences |
|---|
Funding: None.
Ethical approval: Ethical approval has been provided by the Ethics Committee of the Heinrich-Heine-Universität, Universitätsklinikum Düsseldorf.
Conflicts of interest: None.
|
Acknowledgments |
|---|
We thank Nadine Kroll, a medical student, for testing patients in the participating surgeries. We also thank all participating surgeries for their support.
|
Notes |
|---|
Gingter C, Wilm S and Abholz HH. Is COPD a rare disease? Prevalence and identification rates in smokers aged 40 years and over within general practice in Germany. Family Practice 2009; 26: 3–9.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionLimitations and strengthsDeclarationReferences |
|---|
1 Murray CJL, Lopez AD. Alternative projections of mortality and disability by cause. 1990–2020: global burden of disease study. Lancet (1997) 349:1498–1504.[CrossRef][Web of Science][Medline]
2 Roche N, Huchon G. Épidémiologie de la broncho-pneumopathie chronique obstructive. Rev Prat Med Gen (2004) 54:1408–1413.
3 Fletcher C, Peto R. The natural history of chronic airflow obstruction. BMJ (1977) 1:1645–1648.
4 Sunyer J. Urban air pollution and chronic obstructive pulmonary disease: a review. Eur Respir J (2001) 17:1024–1033.
5 Viegi G, Maio S, Pistelli F, et al. Epidemiology of chronic obstructive pulmonary disease: health effects of air pollution. Respirology (2006) 11:523–532.[CrossRef][Web of Science][Medline]
6 Van den Boom B, van Schayck MCP, Rutten-van Möhlken MPMH, et al. Active detection of chronic obstructive pulmonary disease and asthma in general population. Am J Respir Crit Care Med (1998) 156:1730–1738.
7 Van Schayck CP, D'Urzo A, Invernizzi G, Román M, Ställberg B, Urbina C. Early detection of chronic obstructive pulmonary disease (COPD): the role of spirometry as a diagnostic tool in primary care. Prim Care Respir J (2003) 12:90–93.
8 Dales RE, Aaron SD, Vandemheen KL, Mehdizadeh A, Clinch J. The prevalence of airflow obstruction in rural primary care. Respir Med (2006) 100:754–759.[Medline]
9 Piperno D, Bart F, Serrier P, Zureik M, Finkielsztejn L. Patients à risque de broncho-pneumopathie chronique obstructive en médecine générale. Enquête épidémiologique avec 3411 patients. Presse Med (2005) 34:1617–1622.[CrossRef][Web of Science][Medline]
10 Díaz JIJ, de Castro Mesa C, Gontán G, Salamanca MJ, López de Castro F. Prevalence of chronic obstructive pulmonary disease and risk factors in smokers and ex-smokers. Arch Bronconeumol (2003) 39:554–558.[CrossRef][Web of Science][Medline]
11 Vandevoorde J, Verbanck S, Gijssels L, et al. Early detection of COPD: a case finding study in general practice. Respir Med (2007) 101:525–530.[CrossRef][Medline]
12 Clotet J, Gómez-Arbonés X, Ciria C, Albalad JM. Spirometry is a good method for detecting and monitoring chronic obstructive pulmonary disease in high-risk smokers in primary health care. Arch Bronconeumol (2004) 40:155–159.[CrossRef][Web of Science][Medline]
13 Geijer RMM, Sachs APE, Hoes AW, Salomé PL, Lammers JWJ, Verheij TJM. Prevalence of undetected persistent airflow obstruction in male smokers 40–65 years old. Fam Pract (2005) 22:485–489.
14 Price DB, Tinkelman DG, Halbert RJ, et al. Symptom-based questionnaire for identifying COPD in smokers. Respiration (2006) 73:285–295.[CrossRef][Web of Science][Medline]
15 Grobe TG, Dörning H, Schwartz FW. GEK-Report—Ambulant-ärztliche Versorgung 2006 (2006) GEK-Edition: Schwäbisch Gmünd.
16 Jones PW, Quirk FH, Baveystock CM, Littlejohns P. A self-complete measure of health status for chronic airflow limitation. The St. George's respiratory questionnaire. Am Rev Respir Dis (1992) 145:1321–1327.[Web of Science][Medline]
17 Quanjer PH, Tammeling GJ, Cotes JE, et al. On behalf of the working party ‘Standardisation of lung function testing’. Lung volumes and forced ventilatory flows. Report working party standardisation of lung function tests, European community for steel and coal. Official statement of the European respiratory society. Eur Respir J (1993) 6(suppl):5–40.[Medline]
18 Global Initiative for Chronic Obstructive Lung Disease. Global Strategy for the Diagnosis, Management and Prevention of Chronic Obstructive Pulmonary Disease. http://www.goldcopd.com/Guidelineitem.asp?I1=2&;I2=1&intId=989. (accessed on June 06, 2006).
19 Van Schayck CP, Loozen JMC, Wagena E, Akkermans RP, Wesseling GJ. Detecting patients at a high risk of developing chronic pulmonary disease in general practice: cross sectional case finding study. BMJ (2002) 324:1370–1373.
20 Celli BR, Halbert RJ, Isonaka S, Schau B. Population impact of different definitions of airway obstruction. Eur Respir J (2003) 22:268–273.
21 Frank L, Hazell ML, Linehan MF, Frank PI. The diagnostic accuracies of COPD in general practice: the results of the MAGIC study. Prim Care Respir J (2006) 15:286–293.[CrossRef][Medline]
22 Stratelis G, Jakobsson P, Molstad S, Zetterstrom O. Early detection of COPD in primary care: screening by invitation of smokers aged 40 to 55 years. Br J Gen Pract (2004) 54:201–206.[Web of Science][Medline]
23 Schirnhofer L, Lamprecht B, Volmer WM, et al. COPD prevalence in Salzburg, Austria—results from the BOLD study. Chest (2007) 131:29–36.
24 Buist AS, McBurnie MA, Vollmer WM, et al. International variation in the prevalence of COPD (The BOLD Study): a population-based prevalence study. Lancet (2007) 370:743–750.
25 Völzke H, Neuhauser H, Moebus S, et al. Rauchen: Regionale Unterschiede in Deutschland. Deutsches Ärzteblatt (2006) 103:2784–2790.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||