The PUMA study is a multicenter, multinational, cross-sectional, non-interventionist study. A steering committee composed of experts in epidemiology and respiratory medicine in Latin America was appointed to design the protocol and standardize the methodology. An expert pulmonologist was also appointed as national coordinator for each country.

Study sites in each country were selected to reflect the reality of national primary care practice in terms of geographical distribution (urban or rural) and healthcare sector (public or private). Participating sites included primary care centers (family doctors, general practitioners, etc.) with no direct connection with respiratory medicine specialists. The study was approved by the ethics committees for each site and all participants gave their informed consent.

The study population consisted of patients with a high risk of COPD attending participating primary care centers in Argentina, Colombia, Uruguay, and Venezuela. They were enrolled during routine spontaneous or scheduled visits and their medical appointment was unrelated to the study. To be considered high risk, patients had to meet the following inclusion criteria: ≥40 years, current or former smokers (≥10 pack-years, ≥50 pipes/year, or ≥50 cigars/year), and/or exposure to biomass smoke, such as wood or coal, for cooking or heating (exposure ≥100h/year).

Exclusion criteria were pregnancy, contraindication for spirometry (chest, abdominal or brain surgery; acute coronary syndrome; retinal detachment; hospitalization for any heart complaint during the previous 3 months), heart rate ≥120bpm, physical or mental disability rendering the patient unable to complete the study procedures, currently undergoing treatment for tuberculosis or concurrent participation in a clinical trial.

Each center was assigned a number of patients for enrollment. The recruitment period in each center depended on the number of patients with inclusion criteria presenting in the center and completing the study.

Patients were consecutively selected for participation in the study according to the order in which they were seen. Those who did not meet the inclusion criteria continued with their scheduled appointment.

The national coordinators were responsible for the training and certification of staff involved in completing questionnaires and performing spirometries. This was carried out in each country by way of a training session and use of a standardized evaluation designed specifically for this study, based on the recommendations of the American Thoracic Society (ATS).14 Investigator meetings were held to ensure compliance with the procedures and quality standards. The national coordinators standardized the spirometry procedure and reported and monitored the quality of the maneuvers performed. Personnel in the field sent the spirometries in digital format to the national coordinator within 24h after performance. Approval of the technical quality was required before recruitment in a center could be completed. Only spirometries approved by the coordinator were considered validated. If the spirometry was not validated, it was either repeated as soon as possible or the spirometry for that patient was excluded.

Investigators were responsible for selecting the study population according to the inclusion and exclusion criteria and for obtaining informed consent. Data were obtained from the physicians and from the patients on standardized questionnaires (Table 1).

The age and sex, years of experience in general medicine, professional training, smoking history, average number of patients seen per day, teaching activities, location of the study center (urban/rural), medical practice characteristics (public/private), factors limiting use of COPD guidelines and available diagnostic and therapeutic resources of each physician were recorded in the Investigator's Questionnaire.

Physicians recorded data from clinical records, including spirometries, diagnoses, previous treatments, and concomitant diseases in the Background Questionnaire.

The physician checked if the patient could perform spirometry and oximetry during their visit. Patients were then referred to personnel qualified to perform the test and complete the corresponding questionnaire. To complete the PUMA Questionnaire (modified version of the PLATINO study questionnaire), the patient was directly interviewed by trained personnel. Patients who not only met the inclusion criteria but also met any of the exclusion criteria were invited to participate in the Minimum Data Questionnaire (basic demographic data and smoking habit).

The primary outcome variable was diagnosis of COPD based on spirometry. Subjects with COPD were defined as those with an FEV1/FVC (forced expiratory volume in 1s divided by forced vital capacity) ratio of less than 0.7. COPD prevalence was also estimated on the basis of other spirometric criteria such as the GOLD classification (http://www.goldcopd.org) and lower limit of normal (LLN) criteria. Independent variables were obtained from the questionnaires and included demographic data, smoking habit, environmental/domestic pollution, symptoms/history, and treatment of respiratory diseases, concomitant diseases and use of healthcare resources.

If the patient presented any transient criterion that would prevent performance of the spirometry, the questionnaire was completed and a second visit was scheduled. Criteria for postponing spirometry included the use of short-acting bronchodilators in the previous 3h, long-acting beta-adrenergic bronchodilators in the previous 12h, long-acting anticholinergics in the previous 24h, smoking cigarettes, pipes or other tobacco products in the previous 2h, physical exercise, including working out, walking or jogging in the previous hour, and concurrent acute respiratory infection at the time of the visit.

Height and weight data were measured using calibrated equipment. Spirometries were obtained in all centers using an ultrasound spirometer (Easy One by NDD TECHNOPARK, Medical Technologies, Inc. Two Dundee Park, Andover, MA, U.S.A.) that met ATS quality control requirements.14

To determine spirometric acceptability and reproducibility, ATS criteria (150–200ml in a minimum of 8 and maximum of 15 maneuvers) were used to obtain reproducible measurements. Spirometries were carried out with the patients in a sitting position using nasal clips. Each participant completed 2 spirometric tests– baseline spirometry examination and a second test 15min after the administration of 400μg of aerosol salbutamol administered with a spacer.

Oximetry was carried out using a digital pulse oximeter after the patient had been resting for 5min and before receiving the bronchodilator for the spirometry (Riestet-ri-fox N pulse Oximeter [Rudolf Riester GmbH – Bruckstrasse 31, D-72417 Jungingen, Germany] or NONIN Onyx 9500 Oximeter [Nonin Medical, Inc., 13700 1st Avenue, North Plymouth, MN, USA]). The evaluation was performed with the patient in a sitting position breathing room air with the oximeter attached to the right index finger. The average of 6 measurements obtained at 10-s intervals was recorded. In addition, the heart rate provided by the device was recorded. The flow chart for patient selection is shown in Fig. 1, and the procedures performed during the visit are described in Table 2.

Figure 1.

Flow chart. Patient screening.

(0.29MB).

The original size of the sample calculated for studying the prevalence of COPD confirmed by spirometry in the study population of each country (estimated at 20%) with a 3% margin of error was a minimum of 715 subjects per country. The planned sample size was reached only in Venezuela, where 721 patients had completed the PUMA questionnaire by the end of enrollment. Thus, in Argentina and Colombia, with 454 and 465 subjects enrolled, respectively, the margin of error for the estimated prevalence is 3.7%. To increase the sample size for the pooled analysis, 103 patients were included from Uruguay.

In addition to country-specific analyses, some risk factor tests will be performed on the overall study population. In such cases, statistical power is at least 90%, with probability indexes of (a) 1.4 or greater for exposure factors affecting 20% of the disease-free population and (b) 1.6 or greater for exposure factors affecting 10% of the disease-free population. For secondary endpoints restricted to participants with COPD (planned n=572), margins of error for proportions of 10%, 20%, and 50% are 2.5%, 3.5%, and 4.2% respectively.

Descriptive statistics include calculations of means, medians, standard deviations and ranges for continuous variables and proportions and their respective 95% confidence intervals for categorical variables. The prevalence and treatment of COPD will be studied for each country with the exception of Uruguay, due to the small sample size. Most of the remaining analyses will be performed on the pooled sample, except when heterogeneity testing suggests the need for country-specific analyses. In bivariate analyses, proportions will be compared using the Chi-squared test for heterogeneity and linear trend. Differences in means will be investigated with the T-test or one-way ANOVA. For multivariate analyses, logistic and linear regressions will be used for categorical and continuous results, respectively. Statistical analysis will be performed using Stata, version 12 (www.stata.com).