The latest report by the World Health Organization on the situation of smoking in the world shows that Spain, with 44.5 million inhabitants, 26% of whom are regular smokers and 3% occasional smokers, is still in phase III of the smoking epidemic curve.1 A decrease has been observed among male smokers but numbers remain steady among women, still a long way off the phase IV of the population of Western Europe. Monitoring chronic diseases related to tobacco use is a relevant, current issue. Preliminary data from December 2009 from the European Health Survey show a surprising increase in smoking among adults in Spain, reaching 31.5% at the present time.2

The EPI-SCAN study determined that the current prevalence of COPD in Spain according to the GOLD criteria is 10.2% (95% CI, 9.2–11.1) of the 40 to 80 years population.3 Previously, the IBERPOC study in 1997 determined a prevalence of COPD of 9.1% (14.3% in men and 3.9% in women), with important differences between the 7 participating areas.4 COPD is defined according to the former criteria of the European Respiratory Society as a post-bronchodilator FEV1/FVC ratio <88% of the predicted value in men and 89% in women. On this subject, the high number of cases detected in some geographical areas in the IBERPOC study corresponded with female non-smokers over 55 years with a history of childhood respiratory diseases who suffered no symptoms of expectoration or wheezing.5

The comparison of the spirometry results in the IBERPOC and EPI-SCAN studies have been described recently.6 However, the geographical variations in COPD in Spain at present are unknown, as are its relationship with smoking and other local determining factors. The so-called geographical or small area epidemiology7 encourages the local dissemination of health data compared with surrounding towns, which makes it possible to establish priorities for carrying out correction mechanisms, as recently reviewed.8 Furthermore, it should make it possible to monitor and/or confirm that community strategy recommendations are being implemented.9 Internationally, both the PLATINO and BOLD initiatives identified important differences in the distribution of the prevalence of COPD,10,11 although in both studies the samples were from one single city or area per country, so it is not possible to determine the variability within the country or in small areas. The aim of this article is to determine if there are variations in the prevalence, under-diagnosis and under-treatment of COPD among the 11 areas participating in the EPI-SCAN, and to analyze if these variations are related to some of its determining factors, including smoking on an individual level and the COPD official mortality rates published by the autonomous communities at an ecological level.

The methodology and protocol of the EPI-SCAN study have been described previously in detail.12 Briefly, the EPI-SCAN is population-based, multi-centred, cross-sectional, observational, epidemiologic study carried out on national scale with a randomized selection of participants using two-stage sampling, stratified by areas close to the participating centres. The participating centres were selected in accordance with four geographical areas (north, east, south and centre) in Spain, which are: Barcelona, Burgos, Cordoba, Huesca, Madrid (two centres), Oviedo, Seville, Valencia, Vic and Vigo. Two-stage, population-based, randomized sampling was performed using telephone sampling and including men and women in the general population aged between 40 and 80 years and resident in Spain. The field work was performed between May 2006 and July 2007. The study was authorized and presented by the corresponding ethics committees for clinical research, the Clinic i Provincial Hospital in Barcelona being the reference committee. All the participants gave their voluntary written consent to participate in the trials.

Information was collected about sociodemographic data, smoking habits, previous diagnosis of respiratory diseases and other pathologies, COPD exacerbations, dyspnoea scale, and treatment for respiratory diseases, amongst other variables. The presence of respiratory symptoms (daily morning cough, frequent sputum, and the presence at some time of dyspnoea and wheezing) was collected using the Spanish version of the CECA questionnaire.13 Forced spirometry was carried out with MasterScope CT (VIASY Healthcare®, Hoechberg, Germany, using the acceptability and reproducibility criteria and the selection of manoeuvre proposed in the most recent recommendations of the American Thoracic Society/European Respiratory Society (ATS/ERS),14 the reference values of the CECA were used.15 The manoeuvres were repeated 15–30 min after inhaling 200 mcg of salbutamol. Following the criteria of the ATS/ERS guidelines,16 the bronchodilator test was considered positive when recording an increase in FEV1 or FVC >200 ml and >12% with regard to the baseline value.

In the present analysis, COPD was defined in two ways: As a post-bronchodilator FEV1/FVC ratio<0.70 or as the lower limit of normal,14 which is the percentile 5 of the distribution of the variable FEV1/FVC in the general population, in accordance with the reference values of the CECA.15

The mortality rate by autonomous community and gender, adjusted for the European population/100,000 inhabitants was collected from official statistics.17

Of a total of 4,274 subjects contacted randomly by telephone in the 11 centres, 3,885 (90.9%) agreed to participate in the study and in the end a sample of 3,802 (88.9%) was available for analysis (minimum set of data about sex, age and lung function). As reported previously,3 the 389 (9.1%) individuals who refused to participate in the survey were slightly older and usually women, most being never-smokers and ex-smokers. The sociodemographic and clinical characteristics of the 3,802 participants by area in the EPI-SCAN study are shown in table 1. There were no differences by areas in the distribution by age, sex or packs-year, although there were significant differences by smoking status, body mass index (BMI) or educational level in some areas compared to the total (P<.05).

The prevalence of COPD according to the GOLD criteria was 10.2% (CI 95%, 9.2–11.1) and 5.6% (CI 95%, 4.9–6.4) according to the LLN criteria. The ratio of the prevalence of COPD according to the GOLD criteria among the EPI-SCAN areas was 2.7, with a peak in Asturias (16.9%) and a minimum in Burgos (6.2%) (P<.05) (table 1); likewise, the ratio of prevalences using the LLN criteria was 3.1, but with a peak in Madrid-La Princesa (10.1%) and a minimum in Burgos (3.2%) (P<.05).

Table 2 shows the characteristics of the individuals with COPD (according to GOLD) from the EPI-SCAN study. Again, there are subtle, but smaller, differences by areas in comparison with the total. Thus, with regard to the overall figures, the patients with COPD in Oviedo had a lower packs-year mean, in Cordoba and Oviedo they had a lower BMI, and differences remained in the educational level (P<.05). It is worth highlighting that the variations in the severity of COPD (fig. 1), under-diagnosis (58.6% to 72.8%) and under-treatment by areas (24.1% to 72.5%) were significant (P<.05). In contrast, it can be seen that the age and sex of the individuals with COPD is almost identical in all areas, and that the variations in height, weight, BMI and educational level are minimal. The variations in the distribution of smokers and ex-smokers are also small among both the participating sample of the population (table 1) and among those with airflow obstruction compatible with COPD (table 2).

Figure 1.

Distribution of the severity of COPD by areas.

(0.23MB).

Of the participants, 5.8% had a positive bronchodilator test, as did 20.2% of those with COPD. As can be seen in tables 1 and 2, there were wide and inconsistent variations by area in the positive bronchodilator tests. Looking at the total number of participants, there was a four-fold difference in positive bronchodilator tests between the highest number in Vigo (13%) and the lowest in Cordoba (3.2%) (table 1). However, among those with COPD the difference is only 2.1-fold between the maximum in Vigo (29.6%) and the minimum in Burgos (13.8%) (table 2).

The order of the prevalence of COPD by area did not remain the same in each area with regard to sex and age (table 3 and fig. 2) with either of the definitions used, except Burgos, which was consistently the area with the lowest prevalence of COPD, when using both the GOLD and LLN criteria, and for both sexes.

Figure 2.

Prevalence of COPD according to GOLD and the lower limit of normal (LLN) by areas, total and sex. GOLD indicates Global Initiative for Obstructive Lung Disease.

(0.27MB).

This is why it is interesting to observe that the accumulated exposure to tobacco, measured by packs-year [Spearman's correlation ?0.305 in men P=.336 and 0.210 in women P=.513] (fig. 3), does not greatly explain the geographical variations in the prevalence of COPD. Neither is a correlation observed between the prevalence in each area and the mortality rates published by the autonomous communities, with a Spearman's coefficient of 0.193 in men, P=.547 and 0.190 in women, P=.555 (fig. 4).

Figure 3.

Prevalence of COPD by area (Y axis) with regard to packs-year for the area (X axis), [Spearman's correlation -0,305 in men P=.336 and 0.210 in women P=.513].

(0.6MB).

Figure 4.

Prevalence of COPD by area (Y axis) with regard to adjusted mortality rate in the autonomous community (Strategy) (X axis) [Spearman's correlation 0.193 in men P=.547 and 0.190 in women P=.555].

(0.56MB).

This study describes important geographical variations in the current distribution of the prevalence of COPD in Spain, this being 2 or 3-fold higher in certain areas. The lack of homogeneity in the distribution of COPD in the population, both nationally and internationally, has already been described by other authors, the same occurring with asthma. However, among the novel contributions of this article is that it determines that the geographical differences in the prevalence of COPD vary for men and women, and also depend on the spirometry definition used. Furthermore there is a total lack of consistency in the data, and the individual accumulated exposure to tobacco measured by packs-year can not account for this variation by itself. No relationship was found either with the registered mortality from COPD in an ecological analysis. In a free and universal health system like the one in Spain, it is interesting to observe that there were also important variations in under-diagnosis and under-treatment which could not be explained by these factors.

There are significant variations in the distribution of COPD in Spain, both in its prevalence and its under-diagnosis and undertreatment. Confirming these variations and investigating the causes of this heterogeneity must enable not only a more sensible distribution of healthcare resources, but also the setting up of educational and corrective initiatives to reduce the effects of tobacco and other factors responsible for COPD.

The sponsor of the EPI-SCAN study is GlaxoSmithKline S.A. We would like to acknowledge the hard work of all the researchers and collaborators in each centre, and thank the participants for their voluntary, unselfish collaboration.

Julio Ancochea, La Princesa Hospital (Madrid, Spain)

Guadalupe Sánchez, Medical Department, GlaxoSmithkline S.A. (Madrid, Spain)

Enric Duran-Tauleria, Municipal Institute of Medical research (IMIM) (Barcelona, Spain)

Francisco García Río, La Paz Hospital (Madrid, Spain)

Marc Miravitlles, Hospital Clinic of Barcelona, Spain

Luis Muñoz, Reina Sofía Hospital (Cordoba, Spain)

Víctor Sobradillo-Peña, Cruces Hospital (Bilbao, Spain)

Joan B Soriano, Caubet-CIMERA Foundation, Balearic Islands (Majorca, Spain)

Julio Ancochea, La Princesa Hospital (Madrid, Spain)

Luis Borderías, San Jorge Hospital (Huesca, Spain)

Francisco García Río, La Paz Hospital (Madrid, Spain)

Jaime Martínez, Asturias Central Hospital (Oviedo, Spain)

Teodoro Montemayor, Virgen de la Macarena Hospital (Seville, Spain)

Luis Muñoz, Reina Sofía Hospital (Cordoba, Spain)

Luis Piñeiro, Xeral Cies Hospital (Vigo, Spain)

Joan Serra, Vic General Hospital (Vic, Barcelona, Spain)

Juan José Soler-Cataluña, Requena General Hospital (Requena, Valencia, Spain)

Antoni Torres, Hospital Clinic (Barcelona, Spain)

Jose Luis Viejo, Yagüe General Hospital (Burgos, Spain)