Abstract
Our aim was to describe the population-based distribution of several COPD multi-dimensional indices and to evaluate their relationship with daily physical activity, co-morbidity, health status and systemic inflammatory biomarkers. From a population-based sample of 3,802 subjects aged 40–80 from the EPI-SCAN study, 382 subjects (10.2%) with a post-bronchodilator FEV1/FVC<0.7 were identified as COPD. Smoking habits, respiratory symptoms, quality of life, co-morbidities, lung function and inflammatory biomarkers were recorded. Health status and daily physical activity were assessed using the EQ-5D and LCADL questionnaires, respectively. The new GOLD grading and the BODE, ADO, DOSE, modified DOSE, e-BODE, BODEx, CPI, SAFE and HRS indices were determined.
A notable dispersion in the total scores was observed, although 83–88% of the COPD patients were classified into the mildest level and 1–3% in the most severe. The SAFE index was the best independent determinant of daily physical activity; the SAFE and ADO indices were associated with presence of co-morbidity; and the SAFE and modified DOSE indices were independently related to health status. The systemic biomarkers showed a less consistent relation with several indices. In a population-based sample of COPD patients, the SAFE index reaches the highest relation with physical activity, co-morbidity and health status.
Introduction
Chronic obstructive pulmonary disease (COPD) is a major public health problem characterized by chronic progressive airflow limitation that results in dyspnoea, reduced physical activity, co-morbidities and impaired health status. Hence, degree of airflow obstruction is generally considered the key factor to stage COPD severity and to guide and monitor treatment (Citation1). However, the degree of airflow limitation is admittedly a poor predictor of health status or exercise tolerance (Citation2), and therefore this limits its use as the sole indicator of severity in COPD.
In a seminal study, Celli et al. (Citation3) developed the BODE index, a predictor for mortality that comprises four domains (body mass index, airflow obstruction, dyspnoea and exercise capacity). It also predicts hospitalizations (Citation4), reflects the deleterious effects of exacerbations (Citation5) and shows disease modification after interventions (Citation6). In the last years, a number of multi-factorial grading scales have been developed in an attempt to better assess severity, disability and prognosis in COPD (Citation7–10). A recent systematic review identifies 15 different multi-dimensional prognostic COPD indices, several of which have been validated appropriately (Citation11).
Multi-dimensional indices for assessing COPD activity have the advantage of including non-pulmonary markers that indicate the impact of systemic factors on COPD outcome. However, it can be questioned which are the most important variables to include in a multi-dimensional index in order to most appropriately characterize the activity or prognosis of systemic factors associated with COPD. In turn, little information about their population-based distribution is available, since all these indices have been developed and validated in selected samples from outpatient clinics or hospitals. This makes their integration difficult in the estimation of the socioeconomic burden of COPD.
It has been recently suggested that, in order to better assess COPD severity, it is suitable to incorporate (in addition to prognostic variables) markers chosen to characterize disease activity as biological and freeliving physiological variables, particularly daily physical activity (Citation12). In addition to these outcomes, it has been proposed to evaluate the time spent in certain health states. In Europe, the EuroQol 5-dimension (EQ-5D) questionnaire is the instrument most widely used for generating health state utility scores and it is used to calculate the quality-adjusted life-years (Citation13). Moreover, EQ-5D utility scores have been used to evaluate the cost-effectiveness of interventions by state transition models that simulate the progression of COPD over different stages of disease severity (Citation14).
Our aim was to identify the population-based distribution of several multi-dimensional indices to assess COPD activity and to evaluate their relationship with cross-sectional clinical characteristics, such as daily physical activity, presence of co-morbidity, health status burden and systemic inflammatory biomarkers.
Material and methods
Study population
We used data from the EPI-SCAN study, a multicentre, cross-sectional, population-based, observational study, in order to estimate the prevalence of COPD. Detailed descriptions of the EPI-SCAN study are available elsewhere (Citation15). From a population-based sample of 3,802 non-institutionalised participants, aged 40–80 years, we selected those with a post-bronchodilator FEV1/FVC ratio < 0.70. The study was approved by the corresponding ethics committees (Hospital Clinic, Barcelona, Spain). All participants gave written informed consent to participate in the study.
Procedures
Demographic characteristics, smoking habits, educational level, domestic and occupational exposures, respiratory history and symptoms, previous medication and use of health services were collected. Any respiratory exacerbation that required a change in regular medication was considered to be mild, while a respiratory exacerbation treated with a course of oral corticosteroids or antibiotics was considered moderate. Self-reported co-morbidity was documented using the Charlson index. The presence of heart failure, ischemic heart disease, peripheral vascular disease or cerebrovascular disease was coded as cardiovascular disease. Baseline dyspnoea level was assessed by the modified Medical Research Council (mMRC) scale.
Quality of life was assessed using the validated Spanish version of the St. George Respiratory Questionnaire. The EQ-5D was employed to estimate health status. Social tariffs for EQ-5D health states were obtained based on self-rated health and the rating of hypothetical health states using a visual analogic scale and time trade-off techniques. The London Chest Activity of Daily Living (LCADL) scale was used to evaluate daily physical activity. According the Spanish validation study, a total score higher than 22 was considered as a very low level of physical activity (Citation16).
Serum levels of tumour necrosis factor (TNF)-α, interleukin (IL)-6, IL-8, C-reactive protein (CRP), albumin, fibrinogen and nitrites/nitrates (NOx) were determined as previously described (Citation17). Post-bronchodilator spirometries were performed at each site using the same equipment according to current recommendations (Citation15). A 6-minute walk test was performed twice, with an interval between tests of 30 minutes, in accordance with the ATS guidelines (Citation18).
In accordance with the new GOLD severity grades (Citation2), COPD patients were classified in groups A (low risk, less symptoms), B (low risk, more symptoms), C (high risk, less symptoms) or D (high risk, more symptoms), using predicted FEV1, mMRC and exacerbation history.
The severity of COPD was assessed using the following multi-dimensional indices: BODE (Citation3), ADO (Citation9), DOSE (Citation8), modified DOSE (Citation19), e-BODE (Citation10), BODEx (Citation10), COPD Prognostic Index (CPI) (Citation20), SAFE (Citation7), and Hospitalization Risk Score (HRS) (Citation21). A more detailed description of the indices calculation is contained in the online data supplement.
Statistical analysis
Values are expressed as mean ± standard deviation or percentage. Differences between study groups were analysed using the Chi-squared test or ANOVA with post-hoc analysis by the Bonferroni test. Relationships among variables were evaluated by the Pearson correlation analysis and using multiple linear regression models with adjustment for age, sex, BMI and smoking status. To examine associations between variables, odds ratios in the univariate and multivariate analyses were calculated by logistic regression. We developed multiple logistic regression models with adjustment for age, sex, BMI and smoking habit. Statistical significance was assumed for p < 0.05.
Results
From the population-based sample of 3,802 subjects, 386 (10.2%) had post- bronchodilator FEV1/FVC < 0.7. They had a mean age of 63 years, 70.5% were men, mean BMI 28.0 Kg/m2 and had a substantial smoking exposure (mean 42 pack-years from 34% current smokers and 39.9% former smokers). Tables and E1 (see online supplementary material) show their main characteristics.
Table 1. General characteristics of COPD patients, by GOLD airflow limitation
presents the distribution of the multi-dimensional indices in our population-based sample of COPD patients, both for the total number of patients as well as by GOLD grading. The frequency distribution of the total scores from the indices analysed show a notable dispersion () with a predominance of lower scores, which is more pronounced for the BODE and BODEx indices and less so for the ADO index. The CPI and HRS indices are not included in the figure, since the range of their scale (0 to 100) do not allow for comparison.
Figure 1. Frequency distribution of the total scores of several multi-dimensional indices on a population-based sample of COPD patients.
Table 2. Distribution of the multi-dimensional indices in the population-based sample of COPD patients, by GOLD airflow limitation
shows the frequencies of presentation for the indices that have categories or levels of classification. Despite the fact that very severe COPD patients (old GOLD stage IV) were not included, all the indices are able to identify a small percentage of patients in the more advanced stages or grades. Of all these, the stratification of new GOLD grades covers a wider range in the distribution of the patients.
Table 3. Population-based distribution of new GOLD grades, BODE, e-BODE, BODEx and SAFE scores
To determine the variables related with the level of daily physical activity, as estimated by the total score of the LCADL questionnaire, we included variables for sex, age, BMI, pack-years, GOLD grade and all the multi-dimensional indices in a multiple linear regression model. Out of all these variables, the only ones that remained as independent factors were SAFE stage and sex (). shows the discriminative capacity of the different multi-dimensional indices for identifying COPD patients with a very low level of physical activity (total LCADL score > 22). Except in the case of HRS, which did not reach a significant association, the highest scores of the remaining multi-dimensional indices are associated with the presence of limited physical activity. From among these variables, the multiple logistic regression model only retained the SAFE index as an independent predictor (adjusted odds ratio: 2.127 [95%CI: 1.380–3.278], p = 0.001).
Figure 2. Discriminative capacity of several multi-dimensional indices to identify COPD patients with a very low level of daily physical activity. Adjusted odds ratio for sex, age and smoking habit.
Table 4. Multivariate model to predict the daily physical activity in COPD patients
In the analysis of the relationship between the multi-dimensional indices and co-morbidity, CPI and HRS were excluded because these indices score the presence of cardiovascular diseases. Out of the remainder, the ADO index and SAFE stages were independent predictors for the Charlson index (r2 = 0.170, p < 0.001) (). At the same time, BMI, ADO index and SAFE stage were identified as independent risk factors for the presence of cardiovascular morbidity ().
Table 5. Independent multi-dimensional indices related with co-morbidity (Charlson index) in COPD patients
Table 6. Adjusted risk factors for the presence of cardiovascular morbidity in COPD patients
From among the multi-dimensional indices for evaluating COPD severity, the modified DOSE index was the only one to maintain an independent relationship with the general health status assessed by the EuroQoL visual analogic score (r = –0.562, p < 0.001) (Figure E1-A in the online supplementary material). Both EQ-5D utility scores were only independently related to SAFE index (r = –0.500, p < 0.001 and r = –0.522, p < 0.001, respectively) (Figure E1).
Finally, Table E2 shows the results of the multiple regression analysis between systemic inflammatory biomarkers and multi-dimensional indices. C-reactive protein level was related to BODE quartile and CPI, whereas TNF-α and IL-8 were related to DOSE index and to modified DOSE index, respectively. No independent determinants were identified by the serum levels of IL-6, fibrinogen or nitrites/nitrates.
Discussion
Our results provide information about the population distribution of the most often used multi-dimensional indices used to evaluate COPD severity. In most indices, between 83–88% of COPD patients from a population sample are situated at the lowest level (first quartile or stage I), 8–12% at level 2, 2–3% at level 3 and 1–3% at level 4. In comparison, the new classification established by GOLD provides a grade distribution with a wider range, assigning 53% of the patients to group A, 34% to group B, 2% to group C and 10% to group D.
Meanwhile, in COPD patients from our population, the SAFE index is best for estimating daily physical activity; the SAFE and ADO indices are best for the presence of co-morbidity, and the SAFE and modified DOSE indices are best for general health status. Systemic inflammation markers show a less consistent relationship with several indices (BODE, CPI, DOSE, mDOSE and SAFE).
The population distribution of the COPD severity levels should be considered when applying multi-dimensional indices in epidemiological studies. In reality, the majority of these indices have been elaborated and validated based on selected COPD hospital-based series in which there is abundance of severe patients. The samples that generated the BODE, SAFE, CPI, e-BODE and BODEx indices had a 30–35% of patients with mild-moderate COPD (3,7,10,20) and the DOSE index 52% (Citation8), while in our population sample these severity levels represent 94% of the patients. Therefore, it is necessary to evaluate in detail the performance of the different multi-dimensional indices for assessing patient population cohorts with different severity distributions. Even in clinical series, differences have been demonstrated in the mortality-predicting capability of multi-dimensional indices when applied to patients with either mild-moderate or severe-very severe disease (Citation22).
With the aim to evaluating severity according to prognosis, several indices have been proposed to predict mortality (Citation3,Citation9,Citation19), hospitalizations (Citation8,Citation21), exacerbations (Citation7) or all of these (Citation20). However, less attention has been given to their relationship with disease activity indicators, such as physical activity, co-morbidity, health status or systemic inflammation, which can be more relevant in patients with mild-moderate forms of the disease.
Out of all the multi-dimensional indices evaluated, in our population-based sample of COPD patients it was the SAFE index that reached a better relationship with daily physical activity, and it was even able to discriminate very sedentary patients as well. The elements that integrate the SAFE index (health-related quality of life, airflow obstruction and exercise tolerance) maintain a known relationship with the daily physical activity of COPD patients (Citation23). Likewise, the tool used in our study to evaluate physical activity has also been shown to be related with tolerance to exercise, estimated by the 6-minute walked distance (Citation16) and by the shuttle walk test (Citation24).
The relationship between the total score of the LCADL questionnaire in COPD patients and FEV1 is less evident. While in the study to develop the questionnaire it was not identified (Citation24), in the validation of the Spanish version a significant, although discrete, relationship was observed (r = –0.45) (Citation16). The third component of the SAFE index, quality of life evaluated by the SGRQ, has also been shown to be related with the total score of the LCADL questionnaire (Citation16,Citation24).
Keeping in mind that the main determinants of daily physical activity in COPD patients are exercise tolerance and hyperinflation (Citation23), it may be considered surprising for a multi-dimensional index to include quality of life instead of dyspnoea. Although quality of life is an immediate consequence of dyspnoea, there are other manifestations that are better assessed by the SGRQ than by a single dyspnoea scale. Among the other consequences of hyperinflation not related with dyspnoea are the development of cardiovascular co-morbidity (Citation25) and the greater risk for exacerbations (Citation26). Specifically, the SAFE index was developed in order to evaluate the risk of exacerbations (Citation7).
From among the different multi-dimensional indices contemplated in our population-based sample of COPD patients, the modified DOSE and SAFE indices were determinants that were independent of health status. In addition to FEV1, the impact on health status of the other three determinants of the modified DOSE index –dyspnoea, smoking and exacerbations –should be highlighted. Dyspnoea represents the most disabling symptom of COPD and the degree of dyspnoea provides information regarding the patients perception of illness and can be measured (Citation27). There is a known close relationship between several dyspnoea scales (Citation27) as well as health status scores (Citation28).
Exacerbations are also an important determinant in health status in COPD. In a population-based, cross-sectional survey in 5 EU countries and the USA, it was noted that COPD patients with three or more exacerbations during the previous year have a poorer health status, as assessed by the EQ-5D (Citation29).
In our results, we have detected a certain disagreement between the multi-dimensional indices that best discriminate health status (SAFE and mDOSE indices) and co-morbidity (SAFE and ADO indices). At least partially, this disagreement could be justified by the tool used to evaluate health status (EQ-5D), which is not very sensitive to co-morbidities (Citation13). Some specific health-related quality-of-life questionnaires seem to better detect the effect of co-morbidities (Citation30), although their impact on the global health status of mild COPD patients is perhaps less pronounced. In any event, the data provided in these analyses are merely informational and are not meant to imply using one index over another, since longitudinal data would be needed to make such a comparison.
Our study presents a number of advantages and limitations. The big sample size, the standardized collection of data, and the complete absence of knowledge about the primary hypothesis of this analysis are advantages. On limitations, it was not possible to include in the analysis all the multi-dimensional indices reported to date, as HADO, TARDIS, COPDSS or i-BODE (Citation11), because some of the variables necessary for their calculation had not been systematically compiled in the EPI-SCAN study. However, the calculated indices probably represent those which are most widely used in standard practice and, in general, it is simpler for them to be applied systematically.
We do not have a longitudinal follow-up of the patients; therefore, it is not possible to evaluate the relationship between multi-dimensional indices and daily physical activity or health status throughout the progression of the disease. Furthermore, our sample presents a very small number of patients with severe or very severe disease in whom multi-dimensional indices are especially useful for estimating prognoses.
Unlike previous studies with clinical samples, our study provides a representative sample of the population distribution seen in COPD, with a predominance of mild patients. It also provides information about the use of multi-dimensional indices in epidemiological studies. In any case, it should also be kept in mind that applying multi-dimensional indices in patient samples other than the one in which it was developed may require recalibration or modification (Citation11).
Conclusions
Our study provides information about the population distribution of the multi-dimensional indices designed to evaluate COPD, and it identifies those that maintain a better relationship with disease activity indicators. Particularly, the SAFE index is an independent predictor for daily physical activity, co-morbidity and health status burden.
Declaration of Interest Statement:
The EPI-SCAN study has been funded by an unrestricted grant from GlaxoSmithKline Spain. The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
Supplementary Material
Download Zip (459.7 KB)References
- Rabe KF, Hurd S, Anzueto A, Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: GOLD executive summary. Am J Respir Crit Care Med 2007; 176:532–535.
- Global initiative for chronic obstructive lung disease. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease (Revised 2011). www.goldcopd.org. Date last updated: December 2011. Date last accessed: April 2, 2012.
- Celli BR, Cote CG, Marín JM, The body-mass index, airflow obstruction, dyspnea, and exercise capacity index in chronic obstructive pulmonary disease. N Engl J Med 2004; 350:1005–1012.
- Ong KC, Earnest A, Lu SJ. A multi-dimensional grading system (BODE index) as predictor of hospitalization for COPD. Chest 2005; 128:3810–3816.
- Cote CG, Dordelly LJ, Celli BR. Impact of exacerbations on patient-centered outcomes. Chest 2007; 131:696–704.
- Martinez FJ, Foster G, Curtis JL, Predictors of mortality in patients with emphysema and severe airflow obstruction. Am J Respir Crit Care Med 2006; 173:1326–1334.
- Azarisman MS, Anshar Fauzi M, Ahmad Faizal MP, The SAFE (SGRQ score, air-flow limitation and exercise tolerance) index: A new composite score for the stratification of severity in chronic obstructive pulmonary disease. Postgrad Med J 2007; 83:492–497.
- Jones RC, Donaldson GC, Chavannes NH, Derivation and validation of a composite index of severity in chronic obstructive pulmonary disease. The DOSE index. Am J Respir Crit Care Med 2009; 180:1189–1195.
- Puhan MA, Garcia-Aymerich J, Frey M, Expansion of the prognostic assessment of patients with chronic obstructive pulmonary disease: the updated BODE index and the ADO index. Lancet 2009; 374:704–711.
- Soler-Cataluña JJ, Martínez-García MA, Sánchez Sánchez L, Severe exacerbations and BODE index: two independent risk factors for death in male COPD patients. Respir Med 2009; 103:692–699.
- Van Dijk WD, Bemt L, Haak-Rongen S, Multi-dimensional prognostic indices for use in COPD patient care. A systematic review. Respir Res 2011; 12:151.
- Mahler DA, Criner GJ. Assessment tools for chronic obstructive pulmonary disease. Do newer metrics allow for disease modification? Proc Am Thorac Soc 2007; 4:507–511.
- Rutten-van Mölken MP, Oostenbrink JB, Tashkin DP, Does quality of life of COPD patients as measured by the generic EuroQol five-dimension questionnaire differentiate between COPD severity stages? Chest 2006; 130:1117–1128.
- Hoogendoorn M, Rutten-van Mölken MP, Hoogenveen RT, A dynamic population model of disease progression in chronic obstructive pulmonary disease. Eur Respir J 2005; 26:223–233.
- Miravitlles M, Soriano JB, Garcia-Rio F, Prevalence of COPD in Spain and impact of undiagnosed COPD on quality of life and daily life activities. Thorax 2009; 64:863–868.
- Vilaró J, Gimeno E, Sánchez Férez N, Actividades de la vida diaria en pacientes con enfermedad pulmonar obstructiva crónica: validación de la traducción española y análisis comparativo de 2 cuestionarios. Med Clin (Barc) 2007; 129:326–332.
- Garcia-Rio F, Miravitlles M, Soriano JB, Systemic inflammation in chronic obstructive pulmonary disease: A population-based study. Respir Res 2010; 11:63.
- ATS Statement: Guidelines for the six-minute walk test. Am J Respir Crit Care Med 2002; 166:111–116.
- Oga T, Tsukino M, Hajiro T, Predictive properties of diferent multi-dimensional staging systems in patients with chronic obstructive pulmonary disease. Int J COPD 2011; 6:521–526.
- Briggs A, Spence M, Wang H, Development and validation of a prognostic index for health outcomes in chronic obstructive pulmonary disease. Arch Intern Med 2008; 168:71–79.
- Niewoehner DE, Lokhnygina Y, Rice K, Risk indexes for exacerbations and hospitalizations due to COPD. Chest 2007; 131:20–28.
- Esteban C, Quintana JM, Moraza J, BODE-index vs HADO-score in chronic obstructive pulmonary disease: which one to use in general practice? BMC Med 2010; 8:28.
- Garcia-Rio F, Lores V, Mediano O, Daily physical activity in patients with chronic obstructive pulmonary disease is mainly associated with dynamic hyperinflation. Am J Respir Crit Care Med 2009; 180:506–512.
- Garrod R, Bestall JC, Paul EA, Development and validation of a standardized measure of activity of daily liging in patients with severe COPD: the London Chest Activity of Daily Living scale (LCADL). Respir Med 2000; 94:589–596.
- Watz H, Waschki B, Meyer T, Decreasing cardiac chamber sizes and associated heart dysfunction in COPD. Role of hyperinflation. Chest 2010; 138:32–38.
- Garcia-Rio F, Rojo B, Casitas R, Prognostic value of the objective measurement of daily physical activity in COPD patients. Chest 2012; 142:338–46.
- Mahler DA, Weinberg DH, Wells CK, The measurement of dyspnea: contents, interobserver agreement, and physiologic correlates of two new clinical indexes. Chest 1984; 85:751–758.
- Hajiro T, Nishimura K, Tsukino M, Comparison of discriminative properties among disease-specific questionnaires for measuring health-related quality of life in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1998; 157:785-90.
- Punekar YS, Rodriguez-Roisin R, Sculpher M, Implications of chronic obstructive pulmonary disease (COPD) on patients health status: A western view. Respir Med 2007; 101:661–669.
- Van Manen JG, Bindels PJ, Dekker FW, The influence of COPD on health-related quality of life independent of the influence of co-morbidity. J Clin Epidemiol 2003; 56:1177–1184.