The study was a prospective multi-centre study in 24 PHCCs located in the Stockholm area. The participating centres had patients with different socioeconomic backgrounds, and doctors and nurses with limited or no experience of research routines. The study was approved by the Ethics Committee of the Karolinska Institutet, Stockholm, Sweden.
Study population
131 patients diagnosed by general practitioners (GPs) as having COPD were included in the study. Exclusion criteria were age <18 years, malignant disease, severe psychiatric disease, dementia or poor understanding of written Swedish. All participants gave written informed consent. 20 patients were excluded: eight were lost to follow up, seven had incomplete SGRQs, three had spirometric recordings revealing restrictivity, one was an asthmatic included by mistake, and one had a normal spirometry. The excluded patients did not significantly differ from the study population regarding age, sex or smoking habits. The final analysis thus included 111 patients (Table
1).
Table 1
Baseline characteristics for the entire study sample with clinical COPD and the subgroup with verified COPD
Subjects n
| 111 | 83 |
Female (%) | 65.8 | 62.7 |
Age years, mean, (range) | 67.1, (42–85) | 67.1, (42–85) |
Smoking habits (%) | | |
Non-smokers | 2.8 | 1.2 |
Ex-smokers | 56.9 | 61.4 |
Smokers | 40.4 | 34.9 |
Missing data | 1.8 | 2.4 |
Disease known since, (%) | | |
< 1 yrs | 19.8 | 10.8 |
1–5 yrs | 37.8 | 39.8 |
> 5 yrs | 42.3 | 49.4 |
BMI, mean, (range) | 24.4, (17–39) | 24.2, (17–39) |
FEV1/FVC ratio – mean (SD), range | 57.8 (14.3), range (26.0–92.6) | 52.4 (11.4), range (26.0–69.0) |
FEV1, % predicted, mean (SD), range | 58.1 (20.2), range (14.8–111.5) | 52.5 (17.9), range (14.8–102.6) |
Severity classification, (%) FEV1 (post bronch. dil.) | | |
FEV1 ≥ 80% predicted | 11.0 | 7.3 |
50% ≤ FEV1 < 80% predicted | 52.3 | 46.3 |
30% ≤ FEV1 < 50% predicted | 30.3 | 37.8 |
FEV1 < 30% predicted | 6.4 | 8.5 |
Medication visit 1 (visit 2)#
| | |
- Only SABA or ipratropium as needed (%) | 4.5 (6.5) | 6.1 (6.3) |
- Ipratropium, tiotropium, LABA or SABA as regular medication (%) | 28.2 (26.2) | 26.8 (26.6) |
- Ipratropium, tiotropium, LABA or SABA and ICS as regular medication (%) | 50.0 (50.5 | 54.9 (55.7) |
- ICS without any regular brochodilators (%) | 3.6 (6.5) | 2.4 (5.1) |
- No medication (%) | 13.6 (10.3) | 9.8 (6.3) |
- Missing data (n) | 1 (4) | 1 (4) |
The COPD-population found in primary care makes up a more heterogeneous population than COPD populations usually included in treatment studies. Correct spirometric evaluations are often lacking in primary care [
6]. Nevertheless, spirometry had been performed on all but four patients in our study. Among the 111 patients in our study, 85 were diagnosed as having COPD only, whereas 26 patients were considered to have both COPD and asthma by their treating physician. However, the diagnosis of COPD with or without asthma by the GP did not always meet the spirometric criteria for COPD diagnosis according to Global Initiative for Chronic Obstructive Lung Disease (GOLD)[
1]. Nevertheless, we chose to use the GPs diagnosis as inclusion criterion, since this is how patients are diagnosed and treated in primary care. Statistical analyses were performed for the entire study sample with clinical COPD (n = 111) and for the subgroup of patients with spirometry verified COPD (n = 83) which were the major part of the study population. The results from the analyses on the subgroup with verified COPD are reported only if they differed significantly from the results of the primary analyses.
The patients were characterised with regard to age, gender, and pharmacotherapy during the week preceding each visit. Spirometry (FEV1, % of predicted) was performed with ongoing medication according to local routines, but subjected to central evaluation.
Study design
We compared the 10-item CCQ [
3,
5,
7] with the well validated, extensive SGRQ [
4,
8,
9] on two occasions 10 ± 2 weeks apart without systematic changes in treatment between visits. The time interval was chosen to allow for spontaneous change to occur. If considered needed by the GP, treatment was changed according to local routines after the first visit (Table
1).
The patients completed three questionnaires in their Swedish, self-administered versions in the following order: Short Form-36 Health Survey (SF-36) (Standardised Swedish Version 1.0) [
10,
11], SGRQ [
4,
8,
9], and finally the authorized Swedish translation of the CCQ provided by the developer [
3,
5,
7]. The questionnaires were filled in before meeting the health professional, i.e. the GP or a nurse. During the meeting The GP or a nurse (according to local routines) estimated if and how the patients' clinical status had changed between the visits.
Statistical analysis
Non-parametric methods were mainly used, as we did not assume normality of distribution for any variable. For comparison with previous validation studies, however, data in the tables are given as mean ± Standard Deviation (SD) unless otherwise indicated. The software used was SPSS version 12.0.1. (SPSS Inc., Chicago, USA).
Analysis of floor and ceiling effects in all domains in both the CCQ and the SGRQ were made. This was done by calculating the proportion of subjects that had highest possible score and the proportion of subjects that had lowest possible score in each domain.
The closeness of association of CCQ and SGRQ questionnaire data was assessed by Spearman correlation coefficients. We used the following cut-offs: 0 < | r | < 0.3 weak correlation, 0.3 < | r | < 0.7 moderate correlation, | r | > 0.7 strong correlation. The concordance between the instruments was examined with an intra class correlation coefficient.
Measurement properties, intra-class correlation (ICC) and test-retest reliability of the instruments were evaluated using data from a subgroup of stable patients according to SGRQ ratings, which has been defined ± <4 points (the MID) at visit 2.
Test-retest reliability was estimated as the ICC, i.e. the ratio of the between subjects variance and total variance.
Internal consistency, longitudinal and cross-sectional validity were evaluated using data from all patients. Internal consistency was estimated by the Cronbach α (alpha) coefficient [
20]. Commonly accepted minimal standards for reliability coefficients are 0.70 for group comparisons and 0.90 for comparisons within individuals [
2]. Reliability requirements are higher with individualized use because confidence intervals of the scores are based on the Standard Error of the Mean (SEM), and reliability coefficients <0.9 provide too wide intervals for individual monitoring [
2].
To examine cross-sectional validity, we postulated that if the SGRQ and CCQ measure the same construct, they should correlate reasonably well. The a priori expectations were that the total score of SGRQ as well as the symptoms and activity domain scores would correlate strongly with the CCQ total score and with the corresponding domains of CCQ (symptoms and functional state) respectively. For the impacts domain of SGRQ and mental health domain of CCQ, the expectation was that there would be a moderate correlation, since these domains only partially measure the same construct. Only data from the second visit was used.
Longitudinal validity is the ability of the change scores obtained with the investigated instrument to correlate highly, 0 < | r | < 0.3 weak correlation, 0.3 < | r | < 0.7 moderate correlation, | r | > 0.7 strong correlation, with change scores of the criterion/benchmark test SGRQ.