Background
COPD is a chronic and debilitating disease and a leading cause of morbidity and mortality worldwide [
1]. According to the latest estimates of the World Health Organization (WHO), 210 million people have COPD and 3 million people died of COPD in 2005 [
2]. Improving patients' health status is one of the major goals in COPD treatment [
3].
Quality of life has become an important endpoint in medical care, but still there is no consensus on the definition of these concepts [
4]. Smith and colleagues consider quality of life and health status to be separate constructs, in which quality of life is more related to mental health, whereas health status is more related to physical functioning [
4]. The WHO uses a broader definition of health status, by defining health status as 'a state of complete physical, mental and social well-being, and not merely the absence of disease or infirmity'. Similarly, others [
5,
6] define health status as an overall concept covering physiological functioning, symptoms, functional impairment, quality of life, and social functioning as important main domains. These main domains were empirically found to be further divided into sixteen sub-domains [
7,
8], each sub-domain representing a unique aspect of health status. Despite differences in definitions found in the literature it has become clear that a patient's functioning consists of many conceptually distinct sub-domains. Patient-tailored treatment then requires assessment of all these sub-domains.
Questionnaires could facilitate the guidance of patient-tailored disease management by exploring which aspects of health status are problematic and which aspects are not. The past decade many questionnaires have been developed to measure health status. However, most of these instruments consists of only one or a few subscales and thus measure merely some aspects of health status. The question then rises which aspects of health status are measured by these instruments, and which aspects are not covered.
The St George's Respiratory Questionnaire (SGRQ), for instance, is one of the most frequently used and translated disease specific health status instruments in COPD [
9‐
11]. A recent Pubmed search gave 555 hits (date 06/03/2010; terms SGRQ and St George's Respiratory Questionnaire). The SGRQ has been developed to allow comparative measurement of health between patient populations and to quantify changes in health following therapy [
12]. The SGRQ consists of three sections and a total score:
Symptoms, measuring the frequency and severity of respiratory symptoms;
Activity, measuring limitation of activities by breathlessness and activities that cause breathlessness;
Impacts, measuring disturbances in social and psychological functioning due to airway disease;
Total score summarizes the impact of the disease on overall health status [
12‐
14]. The SGRQ thus measures maximally three of the sixteen aspects of health status. It is not clear which aspects of health status are measured, and which aspects of health status are not measured by the SGRQ. This question is all the important to unravel, because the SGRQ, as many other questionnaires, is subject to conceptual confusion. The SGRQ initially was conceived as a standardized self-completed questionnaire for measuring health and perceived well-being ('QoL') in airways diseases [
12]. In the literature, however, the SGRQ is interchangeably referred to as a measure of quality of life [
15], health-related quality of life [
16], health status [
17], a measure for impaired health [
18], or a measure of overall impact of the disease [
19]. Different terms are used for the concept(s) the SGRQ measures. Additionally, since the SGRQ is often used as a criterion in validity testing of other instruments [
20,
21], it is essential to clarify which aspects of health status the SGRQ measures.
In the present study, we tested which aspects of health status are measured by the SGRQ in COPD, by comparing the SGRQ sections Symptoms, Activity and Impacts with multiple aspects of the health status domains Symptoms, Functional Impairment and Quality of Life.
Discussion
The present study evaluated which aspects of health status are measured by the sections of the SGRQ, and which aspects of health status are not covered by the SGRQ.
The sections of the SGRQ correlated significantly with most sub-domains of the NIAF, indicating that the SGRQ was related to many health status aspects. However, most correlations were low to moderate and well below 0.70, indicating that shared variance was too low to conclude that sections of the SGRQ were conceptually similar to these sub-domains.
Applying the criterion of conceptual similarity, the SGRQ measured two of the ten evaluated sub-domains of health status. The SGRQ sections Symptoms and Total showed conceptual similarity with the sub-domain Subjective Symptoms (main domain Symptoms), the SGRQ sections Activity, Impacts, and Total showed conceptual similarity with the sub-domain Subjective Impairment (main domain Functional Impairment).
In a previous study [
7] we found a high correlation between the sub-domains Subjective Impairment and Subjective Symptoms. The instruments included in these sub-domains were different with respect to the content of the items, but had in common that the item-and-response format required highly subjective and general interpretations by the patient. It was argued that both sub-domains measured highly subjective notions of 'being ill', also referred to as illness perceptions [
25]. As the SGRQ reached the criterion for conceptual similarity with these two sub-domains, this would imply that the SGRQ in fact measures illness perceptions, related to symptoms (section
Symptoms and
Total) and functional impairment (sections
Activity and
Impacts). This conclusion is underlined by the high intercorrelations between the SGRQ sections, some correlations even exceeding the criterion for conceptual similarity.
Although illness perceptions related to symptoms and functional impairment are very relevant concepts, many other important aspects of health status are not covered by the SGRQ. With respect to the SGRQ as a measure of aspects of symptoms, these are restricted to the subjectively experienced severity of pulmonary symptoms. Other important aspects of symptoms, such as dyspnea-related emotions, are not measured specifically. With respect to functional impairment, only the subjectively experienced impairments are measured by the SGRQ. Impairment on the behavioural level or actual physical activity level is not measured by the SGRQ sections. Furthermore, the present study showed that the SGRQ does not measure any of the three sub-domains of quality of life evaluated in this study (General Quality of Life, Health-related Quality of Life, and Satisfaction Relation). Finally, since the SGRQ measures merely two sub-domains of the ten evaluated sub-domains, the SGRQ does not provide a detailed measurement of health status. Similarly, present data show that the SGRQ should be considered a valid measure of impaired health in COPD, as the SGRQ originally was conceived. However, the SGRQ measures only two aspects of impaired health (subjective symptoms and subjective impairment). To measure all aspects of impaired health, and thereby allowing patient-tailored treatment, other instruments need to be included as well.
Some methodological issues need to be addressed. First, the NIAF is not the definite answer to the problem of conceptual confusion in current health status instruments. Other aspects of health status not included in the framework may be relevant to COPD patients. This needs to be addressed in future studies, in which patient feedback should be incorporated. Nevertheless, this framework does provide a much more detailed definition of health status, as expressed by the many sub-domains, and is much more formulated in terms of empirical observations than found in the literature. Each sub-domain represents a (conceptually) unique health status aspect. At least 16 sub-domains are measured to provide a detailed picture of the health status of a COPD patient.
Second, using the criterion of conceptual similarity (r ≥ 0.70) as a standard for validity seems a very strict criterion. However, considering the conceptual confusion in health status, one must be carefully interpreting results of earlier validity studies. Often, much lower correlations are accepted as evidence for the validity of the instrument under scrutiny. For example, a correlation between two instruments of 0.40 may be statistically significant, but it indicates only 16% of shared variance. Unambiguous conclusions concerning conceptual similarity between two instruments can only be drawn from the results using a strict approach.
The present study focuses on the relationships between the SGRQ sections and the main domains Symptoms, Functional Impairment, and Quality of Life. Therefore, the conclusions of the present study are not applicable with respect to physiological functioning. However, from a theoretical point of view it is unlikely that a questionnaire will provide a direct measure of physiological processes. For example, studies to date [
26,
27] often show a relationship between FEV
1 and the SGRQ. However, these correlations are low to moderate and do not exceed the criterion of conceptual similarity.
With respect to generalizability of the present study, we believe that the present sample may be an adequate reflection of a the Dutch population of patients with COPD seen in an outpatient clinic. This sample may however not be representative for subgroups of COPD such as patients in pulmonary rehabilitation or patients with primary co-morbidity, which were two major exclusion criteria.
An important clinical implication of the present study is that the SGRQ could facilitate the guidance of disease management only partially. The SGRQ can only be used appropriately for exploring problems in the sub-domains Subjective Symptoms and Subjective Impairment, and not for exploring problems in other sub-domains of health status, such as aspects of quality of life.
Most instruments claiming to measure specific aspects of health status contain only two to five subscales. Thus, at best only some aspects of health status are measured by a specific instrument. This not only has implications for clinical practice, but also for research purposes. In pharmacological trials, the drug under study may have beneficial effects on some aspects of health status, but not on other aspects. If the instruments used measure only few aspects of health status beneficial effects may be missed. With respect to the use of instruments in clinical practice, the present results indicate that one single instrument cannot provide sufficient information on a patient's health status to effectively tailor treatment to the needs of the individual patient, since measuring all aspects of health status is a prerequisite for patient-tailored treatment. This requires combining different instruments into a battery of instruments measuring multiple aspects of health status. However, implementing instruments in daily practice to facilitate disease management requires that instruments are not too time consuming. The past decade a few short instruments have been developed specifically to allow measurement of health status aspects in routine care, such as the Clinical COPD Questionnaire [
28], the Respiratory Illness Questionnaire-monitoring 10 [
29], and the EuroQoL [
30]. None of these instruments provide a detailed picture of a patient's health status. Recently, we developed the Nijmegen Clinical Screening Instrument (NCSI), an instrument which can be used in routine care [
31]. The NCSI is based on the NIAF and measures eleven sub-domains of physiological functioning, symptoms, functional impairment, and quality of life. The NCSI enables a quick (15-25 minutes) and detailed assessment of health status. Also, the COPD Assessment Test (CAT) was developed [
32], 'a validated short and simple instrument for assessing the impact of COPD on health status'. The CAT is constructed as a uni-dimensional instrument, i.e. measuring one single concept, as expressed in a single score. In addition, the correlation between the CAT and the SGRQ-C was well above the criterion for conceptual similarity (r = 0.80) [
32]. Taken together, it is very likely that the CAT, like the SGRQ, measures illness perceptions. How important illness perceptions may be, patient-tailored treatment requires a detailed assessment of many aspects of health status. Therefore, the CAT also will have limited value in patient-tailored treatment.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
LD participated in the design of the study, the acquisition of the data, performed statistical analyses and interpreted the data, and drafted the manuscript. JBPe participated in the acquisition of the data, and in the critical revision of the manuscript for important intellectual content. JM participated in the design of the study, the acquisition of the data, and in the critical revision of the manuscript for important intellectual content. PNRD participated in the critical revision of the manuscript for important intellectual content. JBPr participated in the critical revision of the manuscript for important intellectual content. YFH participated in the acquisition of the data, and the critical revision of the manuscript for important intellectual content. JHV conceived the study, participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript.