[
Original Research COPD
]
Differential Effect of Modified Medical Research Council Dyspnea, COPD Assessment Test, and Clinical COPD Questionnaire for Symptoms Evaluation Within the New GOLD Staging and Mortality in COPD Ciro Casanova, MD; Jose M. Marin, MD; Cristina Martinez-Gonzalez, MD; Pilar de Lucas-Ramos, MD; Isabel Mir-Viladrich, MD; Borja Cosio, MD; German Peces-Barba, MD; Ingrid Solanes-García, MD; Ramón Agüero, MD; Nuria Feu-Collado, MD; Miryam Calle-Rubio, MD; Inmaculada Alfageme, MD; Alfredo de Diego-Damia, MD; Rosa Irigaray, MD; Margarita Marín, MD; Eva Balcells, MD; Antonia Llunell, MD; Juan Bautista Galdiz, MD; Rafael Golpe, MD; Celia Lacarcel, MD; Carlos Cabrera, MD; Alicia Marin, MD; Joan B. Soriano, MD; Jose Luis Lopez-Campos, MD; Juan José Soler-Cataluña, MD; and Juan P. de-Torres, MD; for the COPD History Assessment in Spain (CHAIN) Cohort
The modified Medical Research Council (mMRC) dyspnea, the COPD Assessment Test (CAT), and the Clinical COPD Questionnaire (CCQ) have been interchangeably proposed by GOLD (Global Initiative for Chronic Obstructive Lung Disease) for assessing symptoms in patients with COPD. However, there are no data on the prognostic value of these tools in terms of mortality. We endeavored to evaluate the prognostic value of the CAT and CCQ scores and compare them with mMRC dyspnea.
OBJECTIVE:
METHODS: We analyzed the ability of these tests to predict mortality in an observational cohort of 768 patients with COPD (82% men; FEV1, 60%) from the COPD History Assessment in Spain (CHAIN) study, a multicenter observational Spanish cohort, who were monitored annually for a mean follow-up time of 38 months.
Subjects who died (n 5 73; 9.5%) had higher CAT (14 vs 11, P 5 .022), CCQ (1.6 vs 1.3, P 5 .033), and mMRC dyspnea scores (2 vs 1, P , .001) than survivors. Receiver operating characteristic analysis showed that higher CAT, CCQ, and mMRC dyspnea scores were associated with higher mortality (area under the curve: 0.589, 0.588, and 0.649, respectively). CAT scores ⱖ 17 and CCQ scores . 2.5 provided a similar sensitivity than mMRC dyspnea scores ⱖ 2 to predict all-cause mortality.
RESULTS:
The CAT and the CCQ have similar ability for predicting all-cause mortality in patients with COPD, but were inferior to mMRC dyspnea scores. We suggest new thresholds for CAT and CCQ scores based on mortality risk that could be useful for the new GOLD grading classification.
CONCLUSIONS:
TRIAL REGISTRY:
ClinicalTrials.gov; No.: NCT01122758; URL: www.clinicaltrials.gov CHEST 2015; 148(1):159-168
Manuscript received October 2, 2014; revision accepted December 11, 2014; originally published Online First January 22, 2015. ABBREVIATIONS: 6MWD 5 6-min walk distance; BODE 5 BMI, measure of airflow obstruction, dyspnea score, and exercise capacity; CAT 5 COPD Assessment Test; CCQ 5 Clinical COPD Questionnaire;
journal.publications.chestnet.org
CHAIN 5 COPD History Assessment in Spain; GOLD 5 Global Initiative for Chronic Obstructive Lung Disease; mMRC 5 modified Medical Research Council; ROC 5 receiver operating characteristic; SGRQ 5 St. George’s Respiratory Questionnaire
159
Recent GOLD (Global Initiative for Chronic Obstructive Lung Disease) updates proposed important changes in the stratification of disease severity in patients with COPD.1 These recommendations were based on evidence that FEV1 is a partial descriptor of disease status. Expanding the range of evaluated symptoms by adding the assessment of dyspnea (according to the modified Medical Research Council [mMRC]) or health status (using the COPD Assessment Test [CAT] and/or the Clinical COPD Questionnaire [CCQ]) may result in more adequate assessment of patients with COPD.1 Based on expert opinions, GOLD suggests that these tools provide assessments of symptoms and that it is unnecessary and possibly confusing to use more than one scale. However, limited information is available regarding health-status tools such as the CAT and the CCQ; furthermore, the recommended cutoff points for category assignment are based
on the St. George’s Respiratory Questionnaire (SGRQ), which is a surrogate test for the CAT.2 In contrast to mMRC dyspnea,3,4 we lack information on the relationship of these tools with mortality and the possible prognostic value of the GOLD-proposed thresholds for these tools.
Materials and Methods
smoking history, dyspnea (mMRC, scale 0-4), exacerbations during the previous year, quality of life according to the Spanish versions of the CAT (scale 0-40)7 and the CCQ (scale 0-60),8 treatments, respiratory function (arterial blood gases, spirometry, lung volume, and diffusion capacity for carbon monoxide), exercise capacity (6-min walk distance [6MWD]), and BODE (BMI, measure of airflow obstruction, dyspnea score, and exercise capacity) index (scale 0-10). Data were anonymized in a database with hierarchical access control to guarantee that information was secured. All participants signed the informed consent form previously approved by the ethics committees of each participating center (Comité de Etica de la Investigación, Hospital Universitario la Candelaria, Tenerife, institutional review board no.: 258/2009).
Subjects Patients with COPD participating in this study were part of the CHAIN cohort. CHAIN is a multicenter study of 36 prospective cohorts carried out at university hospitals in Spain.5 COPD was defined by smoking history ⱖ 10 pack-years and a postbronchodilator FEV1/FVC , 0.7 after 400 mg of inhaled albuterol. Patients were stable for at least 8 weeks and received optimal medical therapy. Exclusion criteria were uncontrolled comorbidities such as malignancy at baseline or other confounding diseases that could interfere with the study. Other methodologic aspects of the CHAIN study were published previously.5 The recruitment period was January 15, 2010, to March 31, 2012 (ClinicalTrials.gov Identifier: NCT01122758). Patients are currently in the follow-up period; data analyzed in the present study came from the recruitment date up to May 15, 2014.
Therefore, in the present study, we aimed at assessing the efficacy of the CAT and the CCQ in predicting all-cause mortality in patients with COPD participating in the COPD History Assessment in Spain (CHAIN) cohort,5 a prospective Spanish multicenter study of patients with COPD. We also compared the prognostic capacities of the mMRC dyspnea, the CAT, and the CCQ. Finally, we used survival analyses to identify the best thresholds for the CAT, CCQ, and mMRC dyspnea scores to more comprehensively categorize patients with COPD by the new GOLD classification.
Clinical and Physiologic Measurements
Briefly, at baseline and each annual visit, we evaluated anthropometric data (age, sex, and BMI), comorbidities (Charlson Index, scale 0-33),6
In personal interviews, trained staff obtained information on age, sex, and BMI at the time of recruitment and at yearly appointments. BMI was calculated as weight in kilograms divided by height in meters.
AFFILIATIONS: From the Pulmonary Department (Dr Casanova), Hospital Universitario Nuestra Señora de Candelaria, Tenerife; Pulmonary Department (Dr J. M. Marin), Hospital Universitario Miguel Servet, Zaragoza; Pulmonary Department (Dr Martinez-Gonzalez), Hospital Universitario Central de Asturias, Oviedo; Pulmonary Department I (Dr de Lucas-Ramos), Hospital General Universitario Gregorio Marañón, Madrid; Pulmonary Department (Dr Mir-Viladrich), Hospital Son Llátzer, Mallorca; Pulmonary Department (Dr Cosio), Hospital Universitario Son Espases-IDISPA, Palma de Mallorca; CIBER de Enfermedades Respiratorias (CIBERES) (Drs Cosio, Peces-Barba, Balcells, and Lopez-Campos), Instituto de Salud Carlos III, Madrid; Pulmonary Department (Dr Peces-Barba), Hospital Universitario Fundación Jiménez Díaz, Madrid; Pulmonary Department (Dr Solanes-García), Hospital de la Santa Creu i Sant Pau, Barcelona; Pulmonary Department (Dr Agüero), Hospital Universitario Marqués de Valdecilla, Santander; Pulmonary Department (Dr Feu-Collado), Hospital Universitario Reina Sofía, Instituto Maimónides de Investigación Biomédica de Córdoba (IMIBIC), Universidad de Córdoba (UCO), Córdoba; Pulmonary Department (Dr Calle-Rubio), Hospital Universitario Clinico San Carlos, Madrid; Pulmonary Department (Dr Alfageme), Hospital Universitario de Valme, Sevilla; Pulmonary Department (Dr de Diego-Damia), Hospital Universitario y Politécnico de la Fe, Valencia; Pulmonary Department (Dr Irigaray), Hospital de Manacor, Mallorca; Pulmonary Department (Dr Marín), Hospital General de Castellon, Castellon; Pulmonary Department (Dr Balcells), Hospital del Mar, Barcelona;
Pulmonary Department (Dr Llunell), Hospital de Tarrasa, Tarrasa; Pulmonary Department (Dr Galdiz), Hospital Universitario de Cruces, Bilbao; Pulmonary Department (Dr Golpe), Hospital General de Calde, Lugo; Pulmonary Department (Dr Lacarcel), Hospital Ciudad de Jaén, Jaén; Pulmonary Department (Dr Cabrera), Hospital Universitario de Gran Canaria Doctor Negrín, Las Palmas de Gran Canaria; Pulmonary Department (Dr A. Marin), Hospital Universitario Germans Trias y Pujol, Badalona, Barcelona; Instituto de Investigación Hospital Universitario de la Princesa (IISP) (Dr Soriano), Universidad Autónoma de Madrid, Cátedra UAM-Linde, Madrid; Unidad Médico-Quirúrgica de Enfermedades Respiratorias, Instituto de Biomedicina de Sevilla (IBiS) (Dr Lopez-Campos), Hospital Universitario Virgen del Rocío, Sevilla; Pulmonary Department (Dr Soler-Cataluña), Hospital Universitario Arnau de Vilanova, Valencia; and Pulmonary Department (Dr de-Torres), Clínica Universidad de Navarra, Pamplona, Spain. FUNDING/SUPPORT: Financial support was provided by AstraZeneca. CORRESPONDENCE TO: Ciro Casanova, MD, Universidad de La Laguna, Tenerife, Respiratory Research Unit, Pulmonary Department, Hospital Universitario Nuestra Señora de Candelaria, Carretera del Rosario no. 145, 38010-Santa Cruz de Tenerife, Spain; e-mail: [email protected] © 2015 AMERICAN COLLEGE OF CHEST PHYSICIANS. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians. See online for more details. DOI: 10.1378/chest.14-2449
160 Original Research
[
1 4 8 # 1 C H E S T J U LY 2 0 1 5
]
A specific questionnaire was used to determine smoking status (current or former) and smoking history (in pack-years). Pulmonary function tests were performed following American Thoracic Society guidelines.9 Diffusion capacity for carbon monoxide was determined with the single-breath technique following the European Respiratory Society/American Thoracic Society guidelines.10 Pao2 was measured at rest in the sitting position while the patient breathed room air. The 6MWD test measured the better of two walks separated by at least 30 min.11 Dyspnea was evaluated with the mMRC scale.12 FEV1, BMI, 6MWD, and mMRC values were integrated into the BODE index as previously described.13 The CAT and CCQ questionnaires were self-administered by each patient with the supervision of the interviewer. Exacerbations were defined as use of antibiotics, steroids, or both or hospital admission related to worsening respiratory symptoms. All-cause mortality was recorded. Cause-specific mortality was ascertained by each site investigator to the highest detail possible and then categorized by six groups: respiratory, cardiovascular causes, lung cancer, non-lung cancer, other causes, and unknown. Information was obtained from the family and then confirmed by reviewing medical records.
TABLE 1
Statistical Analysis Data are summarized as relative frequencies for categorical variables, mean and SD for normally distributed scale variables, and median and fifth to 95th percentile for ordinal or nonnormal scale variables. Comparisons were made between groups using the Pearson x2 test, the Kruskal-Wallis H test, the Mann-Whitney U test, one-way analysis of variance, the Student t test, or the Mantel-Cox test as appropriate. A type II receiver operating characteristic (ROC) analysis with mortality as the status reference was used to determine the area under the curves for each symptom-evaluating tool, and the consistency of the differences among them was compared with the Hanley-McNeil test. The cutoff points for the CAT and the CCQ were selected using the sensitivity-specificity ratio of the mMRC dyspnea ⱖ 2 as reference. We performed Cox proportional-hazard regression analysis to determine the added precision of the CAT, CCQ, and mMRC dyspnea and its thresholds in predicting all-cause mortality in models that included all variables and those proposed in the new GOLD classification. Finally, Kaplan-Meier survival analysis for all-cause mortality using ABCD GOLD classification was performed with the current mMRC dyspnea threshold recommended by GOLD and for the new cutoffs points obtained for the CAT and the CCQ. Significance was established as two-tailed P , .05. Calculations were performed using SPSS 20.0 (IBM).
] Baseline Patient Characteristics
Variable
All Patients (N 5 768)
Age, y
68 (9)
69 (9)
62 (8)
Pack-y
56 (28)
58 (29)
47 (21)
Active smoking, %
30
a
Male Patients (n 5 634)
25
Female Patients (n 5 134)
55
BMI, kg/m2
28.1 (5.6)
28.2 (4.7)
27.1 (7.3)
FEV1, L
1.66 (0.66)
1.72 (0.67)
1.37 (0.49)
FEV1, %
60 (20)
59 (20)
62 (20)
PaO2, mm Hg
67.7 (9.9)
67.8 (9.5)
67.1 (11.7)
FVC, L
3.15 (0.95)
3.29 (0.95)
2.50 (0.63)
FVC, %
86 (23)
85 (23)
92 (24)
FEV1/FVC
53 (11)
52 (11)
54 (11)
6MWD, m
439 (105)
441 (100)
429 (122)
1 (0-4)
1 (0-4)
1 (0-4)
Dyspnea (mMRC)b
2 (0-6)
BODE index
b
2 (0-6)
1 (0-7)
IC/TLC
0.36 (0.10)
0.36 (0.11)
0.32 (0.08)
KCO, %
74 (24)
76 (24)
67 (21)
1 (0-5)
Charlson Indexb
1 (0-5)
0 (0-4)
Exacerbations
0.45 (0.03)
0.44 (0.03)
0.49 (0.07)
Hospitalizationsc
0.15 (0.02)
0.15 (0.02)
0.13 (0.03)
Inhaled anticholinergic,a %
75
74
71
Inhaled b2-agonist,a %
73
74
69
c
Inhaled corticosteroid, % a
65
66
62
CATb
11 (2-27)
11 (2-26)
14 (2-25)
CCQb
1.3 (0-3.8)
1.3 (0-3.7)
1.6 (0-3.9)
Data presented as mean (SD) unless otherwise noted. 6MWD 5 6-min walk distance; BODE 5 BMI, measure of airflow obstruction, dyspnea score, and exercise capacity; CAT 5 COPD Assessment Test; CCQ 5 Clinical COPD Questionnaire; IC 5 inspiratory capacity; KCO 5 carbon monoxide transfer coefficient; mMRC 5 modified Medical Research Council; TLC 5 total lung capacity. aNumber and/or percent. bMedian (fifth to 95th percentile). cIn the year before enrollment (percentage per patient).
journal.publications.chestnet.org
161
Results
CHAIN Cohort Follow-up and Mortality
Study Population
Median follow-up time was 38 months (fifth to 95th percentile, 22-50 months; range, 2-52 months). During follow-up, there were 73 (9.5%) deaths: 25% from respiratory causes, 16% from cardiovascular causes, 18% from lung cancer, 10% from non-lung cancer, 11% from miscellaneous causes, and 20% from unidentified causes. Baseline characteristics of survivors and nonsurvivors are shown in Table 2. Subjects who died were older, had more exacerbations and hospitalizations, had a greater degree of airway obstruction, were more hypoxic, walked less far in the 6MWD test, and had a higher BODE index than those who survived. After multivariable analysis, BODE index remained as the only independent predictor of all-cause mortality: hazard ratio 5 1.34 (95% CI: 1.14-1.56, P , .001) for BODE index.
A total of 768 patients with COPD were evaluated at baseline. The clinical and physiologic characteristics of these patients are shown in Table 1. The population was mainly men (82.5%) and included a broad range of patients with airflow obstruction: 126 mild (16.4%), 387 moderate (50.4%), 172 severe (22.4%), and 83 very severe (10.8%). Group assignment, using all metric symptoms recommended by GOLD, was as follow: 140 (18.2%) in group A, 290 (37.8%) in group B, 37 (4.8%) in group C, and 301 (39.2%) in group D. Patients reported low levels of symptoms and underwent few hospital admissions during the previous year. Approximately 75% of them used inhaled muscarinic antagonists, and a similar percentage used b2 agonists. In general, the patients displayed normal exercise capacity and little comorbidity. TABLE 2
] Characteristics of Patients With COPD at Baseline According to Mortality Outcome
Variable
Alive (n 5 695)
Sex, male (female)a
570 (125)
P Value
Dead (n 5 73) 64 (9)
.226 , .001
Age, y
67 (9)
71 (8)
Pack-y
56 (28)
56 (29)
.939
31%
.740
Active smoking, %
29%
a
BMI, kg/m2
28.1 (5.5)
27.5 (5.6)
FEV1, L
1.69 (0.65)
1.36 (0.62)
FEV1, %
60 (20)
52 (29)
.001
FVC, L
3.19 (0.94)
2.81 (0.92)
.002
FVC, %
87 (23)
80 (21)
.013
FEV1/FVC
53 (11)
PaO2, mm Hg
68.5 (9.8)
6MWD, m Dyspnea, mMRC score BODE indexb
50 (10)
376 (133)
, .001
1 (0-3)
2 (0-4)
, .001
2 (0-5)
3 (0-9)
, .001
0.36 (0.10)
0.33 (0.12)
KCO, %
75 (24)
67 (25)
Comorbidity (Charlson Index)
.046 , .001
IC/TLC
b
, .001
62.7 (9.3)
445 (94) b
.391
1 (0-5)
, .050 .029
1 (0-5)
.691
Exacerbationsc
0.43 (0.03)
0.62 (0.12)
.087
Hospitalizationsc
0.14 (0.02)
0.33 (0.08)
.025
Inhaled anticholinergic,a %
73
87
.013
Inhaled b2-agonist, %
73
76
.576
Inhaled corticosteroida
64
73
.136
a
CAT score
b
CCQ scoreb
11 (2-25)
14 (2-31)
.022
1.3 (0-3.6)
1.6 (0-4.7)
.033
Data are presented as mean (SD) unless otherwise noted. See Table 1 legend for expansion of abbreviations. aNumber and/or percent. bMedian (fifth to 95th percentile). cIn the year before enrollment (percentage per patient).
162 Original Research
[
1 4 8 # 1 C H E S T J U LY 2 0 1 5
]
Symptom-Evaluating Tools and Mortality
Subjects who died had more dyspnea and were associated with higher CAT and CCQ scores (Table 2). Figure 1 shows the ROC curves for scores from each of GOLD recommended symptom-evaluating tools (CAT, CCQ, mMRC dyspnea score). The CAT and the CCQ had similar capacities to predict mortality (C-statistic 0.589 and 0.588, respectively) in patients with COPD but inferior to that of mMRC dyspnea score (C-statistic 0.649). These differences were nonconsistent (P . .05 calculated by the Hanley-McNeil test). Both CAT scores and CCQ scores were independently associated with all-cause mortality after adjustments for age, but not when the model included all variables proposed by the GOLD multidimensional assessment and their recommended cutoffs including the two cutoff criteria for exacerbations (exacerbations ⱖ 2 or ⱖ 1 leading to hospital admission). In this model, FEV1 , 50% and ⱖ 1 hospitalizations were the most important predictors of all-cause mortality (Table 3). The current cutoffs for the CAT and CCQ recommended by GOLD were not associated with mortality (Figs 2A, 2B). mMRC dyspnea score ⱖ 2 (but not ⱖ 1, P 5 .053) was also associated with a significantly shorter all-cause survival (P 5 .001, Fig 2C). Similar results were observed with the newly determined cutoff points for the CAT (ⱖ 17) and the CCQ (. 2.5), (P 5 .004 and P 5 .003, respectively, Fig 3).
TABLE 3
] Hazard Ratios and 95% CIs for All-Cause Mortality in Patients With COPD Hazard Ratio
95% CI
FEV1 , 50%
2.05
1.27-3.31
.003
mMRC dyspnea score ⱖ 2
2.16
1.34-3.48
.001
CAT score ⱖ 10
1.32
0.81-2.17
.266
CCQ score . 1
1.38
0.84-2.25
.200
Variable
P Value
Univariate analysis
Exacerbations ⱖ 2
1.83
1.03-3.26
.039
Hospitalizationsa ⱖ 1
2.85
1.66-4.89
, .001
1.93
1.17-3.19
.010
2.59
1.47-4.55
.001
a
Multivariate analysis FEV1 , 50% Hospitalizations ⱖ 1 a
Hazard ratios and 95% CIs for all-cause mortality in patients with COPD according to univariate and multivariate logistic Cox proportionalhazard regression analyses that included all variables with their cutoff points proposed in the new GOLD combined assessment, with constant using backward stepwise method and Wald criteria per five iterations. GOLD 5 Global Initiative for Chronic Obstructive Lung Disease. See Table 1 legend for expansion of other abbreviations. aPer patient-y in the year before enrollment.
Table 4 shows mortality profiles of the ABCD classification groups using the current and new thresholds for the mMRC, CAT, and CCQ symptom measurements. Kaplan-Meier analysis including mMRC dyspnea and these new CAT and CCQ thresholds showed a significantly shorter survival in group D with all of the symptom metrics, regardless of the tools chosen. However, the overall stratification for mortality risk by GOLD categories was poor with groups A, B, and C showing important overlapping (Fig 4).
Discussion
Figure 1 – Comparative type 2 receiver operating characteristic curves for the CAT, the CCQ, and the mMRC dyspnea score, with all-cause mortality as the main outcome in patients with COPD. CAT 5 COPD Assessment Test; CCQ 5 Clinical COPD Questionnaire; mMRC 5 modified Medical Research Council.
journal.publications.chestnet.org
This observational study of patients with COPD who attended pulmonary clinics yielded several important findings. First, to our knowledge, this is the first study reporting that the CAT and the CCQ are predictors of all-cause mortality in patients with COPD. Second, these health-status questionnaires have a lower predictive capacity for all-cause mortality than the dyspnea mMRC score. This suggests that the type of tool recommended by GOLD to evaluate the symptoms domain can substantially alter the prognostic evaluation of patients, even in patients assigned to the same group. Finally, we confirmed by survival analysis that mMRC dyspnea score ⱖ 2 is a good threshold; based on these survival data, we propose new CAT (ⱖ 17) and CCQ (. 2.5) thresholds to more comprehensively categorize patients with COPD according to the new GOLD classification. 163
Figure 2 – A-C, Kaplan-Meier curves for all-cause mortality using symptom-evaluating tools and cutoff proposed by GOLD (Global Initiative for Chronic Obstructive Lung Disease): A, CAT ⱖ 10; B, CCQ . 1; C, mMRC dyspnea ⱖ 2. The survival differed significantly between groups (P 5 .001) only when the mMRC dyspnea threshold was used. See Figure 1 legend for expansion of other abbreviations.
Current guidelines for COPD management recommend multidimensional evaluation of the disease. The new GOLD strategy includes mMRC dyspnea, CAT, and the CCQ scores and states that is unnecessary to use more than one tool. However, GOLD recognized that a more comprehensive symptoms assessment should be performed, rather than generating a simple measure of breathlessness.1 The CAT and the CCQ are short and simple selfadministered questionnaires that have been developed and validated to measure clinical control in patients with COPD.14,15 Significant correlation has been demonstrated between both questionnaires and between these questionnaires and the SGRQ,4,16 but not with mMRC dyspnea.17,18 These correlations could have important implications for the new GOLD classification. A recent study reported that the use of a variety of symptomevaluating tools significantly modified grade assignment,19 but most importantly, that mMRC dyspnea, but not the CAT or the CCQ, was a predictor of mortality in patients with COPD.3,4 Therefore, one of the novel and most important findings of the present investigation was the confirmation that the CAT score and the CCQ score are predictors of mortality in patients with COPD. Both of these symptom-evaluating tools had a similar predictive 164 Original Research
capacity of all-cause mortality that was lower than the mMRC dyspnea score. Previous studies showed that the SGRQ and the Chronic Respiratory Questionnaire predicted mortality in patients with COPD independent of their degree of airway obstruction and age.20-22 According to the current investigation, the predictive capacity of those tests was lower than that of other parameters, such as exercise capacity or BODE index. On the other hand, only one study, to our knowledge, compared the predictive capacity of a health-status questionnaire (the SGRQ) with a dyspnea score, but unfortunately, in that study the latter was measured with a visual analog scale.20 The present study confirms that the instrument used to determine the level of symptoms has important consequences on the vital prognoses of these patients. This result is not surprising, as the CAT and the CCQ assess several symptoms in addition to breathlessness. Similar differences were found previously in the impact on the disease depending on the tool used to evaluate the vertical risk-assessment axis (FEV1 or exacerbations) of the new GOLD proposal.23 Although the GOLD strategy acknowledges that the new multidimensional classification was proposed for disease management and not for prognostic purposes, it would be desirable to implement a novel
[
1 4 8 # 1 C H E S T J U LY 2 0 1 5
]
Figure 3 – A, B, Kaplan-Meier curves for all-cause mortality using symptomevaluating tools by GOLD with the new cutoff that have demonstrated to predict all-causes mortality: A, CAT ⱖ 17; B, CCQ . 2.5. See Figure 1 and 2 legends for expansion of abbreviations.
grading system with the capacity to predict outcome. A review by the GOLD scientific committee recommended the implementation of studies to compare specific symptom-evaluating tools and prognosis in the each category.24 Our study sought to clarify this issue and suggest that the combined assessment should be redefined. Another important strength of the present study was our survival-based reanalysis of the cutoff points for the CAT, the CCQ, and the mMRC dyspnea proposed by GOLD. The threshold currently recommended for the
journal.publications.chestnet.org
CAT is ⱖ 10, which was previously estimated as a surrogate of the SGRQ.16 For the CCQ, the equivalent cutoff has not been precisely determined, but it appears to be in the range of 1 to 1.5 points.1 Therefore, another relevant finding of the present study was that these health-status thresholds did not predict survival. Using ROC analysis, we determined that the best predictive values of all-cause mortality were higher (CAT ⱖ 17 and CCQ . 2.5) than the thresholds currently proposed. Our proposed CAT cutoff point is similar to the one recently reported in a Canadian non-COPD population.25
165
TABLE 4
] 3-Y Mortality for Patients According to Symptom-Evaluating Tools mMRC
GOLD 2013
ⱖ2
CAT ⱖ 10
CCQ ⱖ 17
.1
. 2.5
23 (6.6)
9 (4.5)
29 (7.1)
10 (4.8)
7 (7.7)
17 (7.6)
9 (9.1)
16 (8.5)
14 (13.5)
19 (9.2)
19 (17.1)
27 (11.7)
17 (17.7)
A
13 (4.5)
14 (7)
B
13 (9.7)
C
13 (10.6)
D
28 (13.3)
30 (13.2)
4 (7.8)
Data are given as No. (%). Three-year mortality for patients according to symptom-evaluating tools with their current cutoff points by GOLD classification and the new threshold proposed in this study. See Table 1 and 3 legends for expansions of abbreviations.
Some authors have proposed to lower the cutoff point for the mMRC dyspnea score of the new GOLD classification to ⱖ 1,24 but the present study confirmed results of previous studies showing that mMRC dyspnea score ⱖ 2 is adequate for discriminating all-cause mortality. Therefore, in contrast to what we suggest for the CAT and the CCQ, we believe that the mMRC score cutoff point should not be modified. The new symptomevaluating tool thresholds proposed in the present study reached a similar mortality prediction in all GOLD categories, but most importantly showed a significant over-
lap for groups A, B, and C, mostly using the CAT score. This could be related with the great annual variability observed in the CAT scores, especially, in those with scores between 10 and 15 points.18 In fact, the low capacity of the ABCD GOLD classification to stratify mortality risk has been observed in previous studies,24 suggesting that this multidimensional assessment of patients with COPD should be refined. Another novel analysis in the present study was the result of our multivariate analysis, which included the tools recommended by the GOLD strategy using the
Figure 4 – A-C, Kaplan-Meier curves for all-cause mortality in the ABCD GOLD classification using symptom-evaluating tools with the new thresholds proposed in this study: A, CAT ⱖ 17; B, CCQ . 2.5; C, mMRC dyspnea ⱖ 2. See Figure 1 and 2 legends for expansion of abbreviations.
166 Original Research
[
1 4 8 # 1 C H E S T J U LY 2 0 1 5
]
suggested cutoffs. Our analysis further confirmed the importance of an FEV1 , 50% and a rate of annual hospitalizations ⱖ 1 as the strongest and independent predictors of all-cause mortality. Our study has several limitations. First, the CHAIN cohort is an observational study of patients with COPD attending pulmonary clinics and is not from a general medical practice or population-based study. Therefore, the cohort may not represent the true distribution of COPD severity in the general population. However, our cohort included a broad range of disease severity. Second, few women were included in the cohort, and the findings reported here cannot be extended to that sex. Nevertheless, the distribution of women into GOLD categories was similar to that of men. Third, the number of deaths is relatively small, preventing us from performing a stratified analysis of mortality according to cause. However, our all-cause mortality results associ-
Acknowledgments Author contributions: C. Casanova and J. P. d.-T. contributed to study conception and design, recruitment of patients, data analysis and interpretation, and drafting the manuscript for important intellectual content; J. M. M., C. M.-G., P. d. L.-R., I. M.-V., B. C., G. P.-B., I. S.-G., R. A., N. F.-C., M. C.-R., I. A., A. d. D.-D., R. I., M. M., E. B., A. L., J. B. G., R. G., C. L., C. Cabrera, and A. M. contributed to study conception and design, recruitment of patients, data analysis and interpretation, and review of the manuscript; and J. B. S., J. L. L.-C., and J. J. S.-C. contributed to study conception and design, data analysis and interpretation, and drafting the manuscript for important intellectual content. Financial/nonfinancial disclosures: The authors have reported to CHEST the following conflicts of interest: Dr Casanova has participated in speaking activities, industry advisory committees, and other activities related to industry sources by Almirall S.A., AstraZeneca, GlaxoSmithKline plc, and Novartis AG during the period 2012 to 2014. Dr Martinez-Gonzalez has participated in speaking activities and other related activities sponsored by Almirall S.A., Boehringer Ingelheim GmbH, Chiesi Farmaceutici S.p.A., GlaxoSmithKline plc, and Pfizer Inc during the period 2011 to 2014. Dr Solanes-García has participated in speaking activities for AstraZeneca, Biodatos Investigacion sl, Pfizer Inc, Boehringer Ingelheim GmbH, Esteve, Chiesi Farmaceutici S.p.A., GlaxoSmithKline plc, The Menarini Group, and Novartis AG. Dr Golpe has participated in speaking activities for Astra Zeneca, GlaxoSmithKline plc, Almirall S.A., Novartis, Ferrer Internacional, and Rovi Corporation. Dr de-Torres has received consultancy fees for participating in Takeda Pharmaceuticals International GmbH, The Menarini Group, and Novartis
journal.publications.chestnet.org
ated with mMRC dyspnea and other important predictors in patients with COPD, such as the BODE index, are in accordance with previous studies.3,4,13 In summary, our data, based on a large cohort of well-characterized patients, provide important information on the assessment of patients with COPD. We observed that health status, evaluated by the CAT and the CCQ, predicted mortality but with a lower capacity than the mMRC dyspnea scale. Therefore, the type of tool used to evaluate the patients’ symptoms could alter their prognostic evaluation even in those assigned to the same GOLD category. Finally, based on all-cause mortality, new cutoff points for the CAT and the CCQ were identified. According to our findings, GOLD should probably redefine the thresholds proposed for these symptom-evaluating tools, to more comprehensively assess patients with COPD. Further studies in other populations should validate our results.
AG advisory boards during the period 2010 to 2013. He has also received fees for speaking activities for GlaxoSmithKline plc, AstraZeneca, Chiesi Farmaceutici S.p.A., The Menarini Group, Novartis AG, Merck Sharp & Dohme Corp, and Takeda Pharmaceuticals International GmbH and received consultancy fees for participating in the Takeda Pharmaceuticals International GmbH and Novartis AG advisory boards during the period 2010 to 2013. Drs J. M. Marin, de Lucas-Ramos, Mir-Viladrich, Cosio, Peces-Barba, Agüero, Feu-Collado, Calle-Rubio, Alfageme, de Diego-Damia, Irigaray, Marín, Balcells, Llunell, Galdiz, Lacarcel, Cabrera, A. Marin, Soriano, Lopez-Campos, and Soler-Cataluña have reported that no potential conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article. Role of sponsors: The sponsor had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Other contributions: We thank AstraZeneca for their financial support of this study.
References 1. Vestbo J, Hurd SS, Agusti AG, et al. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: GOLD executive summary. Am J Respir Crit Care Med. 2013;187(4):347-365. 2. Jones PW, Brusselle G, Dal Negro RW, et al. Properties of the COPD assessment test in a cross-sectional European study. Eur Respir J. 2011;38(1):29-35. 3. Nishimura K, Izumi T, Tsukino M, Oga T. Dyspnea is a better predictor of 5-year survival than airway obstruction in patients with COPD. Chest. 2002;121(5): 1434-1440.
4. Almagro P, Martinez-Camblor P, Soriano JB, et al. Finding the best thresholds of FEV1 and dyspnea to predict 5-year survival in COPD patients: the COCOMICS study. Plos One. 2014;9(2):e89866. 5. López-Campos JL, Peces-Barba G, Soler-Cataluña JJ, et al; En Nombre del Grupo de Estudio CHAIN. Chronic obstructive pulmonary disease history assessment in Spain: a multidimensional chronic obstructive pulmonary disease evaluation. Study methods and organization. Arch Bronconeumol. 2012;48(12): 453-459. 6. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373-383. 7. Agustí A, Soler JJ, Molina J, et al. Is the CAT questionnaire sensitive to changes in health status in patients with severe COPD exacerbations? COPD. 2012;9(5): 492-498. 8. van der Molen T, Willemse BW, Schokker S, ten Hacken NH, Postma DS, Juniper EF. Development validity and responsiveness of the Clinical COPD Questionnaire. Health Qual Life Outcomes. 2003;1:13. 9. American Thoracic Society. Lung function testing: selection of reference values and interpretative strategies. Am Rev Respir Dis. 1991;144(5):1202-1218. 10. Macintyre N, Crapo RO, Viegi G, et al. Standardisation of the single-breath determination of carbon monoxide uptake in the lung. Eur Respir J. 2005; 26(4):720-735. 11. ATS Committee on Proficiency Standards for Clinical Pulmonary Function Laboratories. ATS statement: guidelines for the six-minute walk test. Am J Respir Crit Care Med. 2002;166(1):111-117.
167
12. Mahler DA, Wells CK. Evaluation of clinical methods for rating dyspnea. Chest. 1988;93(3):580-586. 13. Celli BR, Cote CG, Marin JM, et al. The body-mass index, airflow obstruction, dyspnea, and exercise capacity index in chronic obstructive pulmonary disease. N Engl J Med. 2004;350(10): 1005-1012. 14. Jones PW, Harding G, Berry P, Wiklund I, Chen WH, Kline Leidy N. Development and first validation of the COPD Assessment Test. Eur Respir J. 2009;34(3): 648-654. 15. van der Molen T, Willemse BW, Schokker S, ten Hacken NH, Postma DS, Juniper EF. Development, validity and responsiveness of the Clinical COPD Questionnaire. Health Qual Life Outcomes. 2003;1:13. 16. Jones PW, Tabberer M, Chen WH. Creating scenarios of the impact of COPD and their relationship to COPD Assessment Test (CAT™) scores. BMC Pulm Med. 2011;11:42.
168 Original Research
17. Han MK, Muellerova H, Curran-Everett D, et al. GOLD 2011 disease severity classification in COPDGene: a prospective cohort study. Lancet Respir Med. 2012;1(1):43-50. 18. de Torres JP, Marin JM, Martinez-Gonzalez C, et al; COPD History Assessment in Spain (CHAIN) Cohort. Clinical application of the COPD assessment test: longitudinal data from the COPD History Assessment in Spain (CHAIN) cohort. Chest. 2014;146(1):111-122. 19. Casanova C, Marin JM, Martinez-Gonzalez C, et al; COPD History Assessment In SpaiN (CHAIN) Cohort. New GOLD classification: longitudinal data on group assignment. Respir Res. 2014;15:3. 20. Domingo-Salvany A, Lamarca R, Ferrer M, et al. Health-related quality of life and mortality in male patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2002;166(5):680-685. 21. Oga T, Nishimura K, Tsukino M, Sato S, Hajiro T. Analysis of the factor related to
22.
23.
24.
25.
mortality in chronic obstructive pulmonary disease: role of exercise capacity and health status. Am J Respir Crit Care Med. 2003;167(4):544-549. Marin JM, Cote CG, Diaz O, et al. Prognostic assessment in COPD: health related quality of life and the BODE index. Respir Med. 2011;105(6): 916-921. Lange P, Marott JL, Vestbo J, et al. Prediction of the clinical course of chronic obstructive pulmonary disease, using the new GOLD classification: a study of the general population. Am J Respir Crit Care Med. 2012;186(10):975-981. Agusti A, Hurd S, Jones P, et al. FAQs about the GOLD 2011 assessment proposal of COPD: a comparative analysis of four different cohorts. Eur Respir J. 2013;42(5):1391-1401. Pinto LM, Gupta N, Tan W, et al; CanCOLD Study Group. Derivation of normative data for the COPD assessment test (CAT). Respir Res. 2014;15:68.
[
1 4 8 # 1 C H E S T J U LY 2 0 1 5
]
Original Research COPD
]
Differential Effect of Modified Medical Research Council Dyspnea, COPD Assessment Test, and Clinical COPD Questionnaire for Symptoms Evaluation Within the New GOLD Staging and Mortality in COPD Ciro Casanova, MD; Jose M. Marin, MD; Cristina Martinez-Gonzalez, MD; Pilar de Lucas-Ramos, MD; Isabel Mir-Viladrich, MD; Borja Cosio, MD; German Peces-Barba, MD; Ingrid Solanes-García, MD; Ramón Agüero, MD; Nuria Feu-Collado, MD; Miryam Calle-Rubio, MD; Inmaculada Alfageme, MD; Alfredo de Diego-Damia, MD; Rosa Irigaray, MD; Margarita Marín, MD; Eva Balcells, MD; Antonia Llunell, MD; Juan Bautista Galdiz, MD; Rafael Golpe, MD; Celia Lacarcel, MD; Carlos Cabrera, MD; Alicia Marin, MD; Joan B. Soriano, MD; Jose Luis Lopez-Campos, MD; Juan José Soler-Cataluña, MD; and Juan P. de-Torres, MD; for the COPD History Assessment in Spain (CHAIN) Cohort
The modified Medical Research Council (mMRC) dyspnea, the COPD Assessment Test (CAT), and the Clinical COPD Questionnaire (CCQ) have been interchangeably proposed by GOLD (Global Initiative for Chronic Obstructive Lung Disease) for assessing symptoms in patients with COPD. However, there are no data on the prognostic value of these tools in terms of mortality. We endeavored to evaluate the prognostic value of the CAT and CCQ scores and compare them with mMRC dyspnea.
OBJECTIVE:
METHODS: We analyzed the ability of these tests to predict mortality in an observational cohort of 768 patients with COPD (82% men; FEV1, 60%) from the COPD History Assessment in Spain (CHAIN) study, a multicenter observational Spanish cohort, who were monitored annually for a mean follow-up time of 38 months.
Subjects who died (n 5 73; 9.5%) had higher CAT (14 vs 11, P 5 .022), CCQ (1.6 vs 1.3, P 5 .033), and mMRC dyspnea scores (2 vs 1, P , .001) than survivors. Receiver operating characteristic analysis showed that higher CAT, CCQ, and mMRC dyspnea scores were associated with higher mortality (area under the curve: 0.589, 0.588, and 0.649, respectively). CAT scores ⱖ 17 and CCQ scores . 2.5 provided a similar sensitivity than mMRC dyspnea scores ⱖ 2 to predict all-cause mortality.
RESULTS:
The CAT and the CCQ have similar ability for predicting all-cause mortality in patients with COPD, but were inferior to mMRC dyspnea scores. We suggest new thresholds for CAT and CCQ scores based on mortality risk that could be useful for the new GOLD grading classification.
CONCLUSIONS:
TRIAL REGISTRY:
ClinicalTrials.gov; No.: NCT01122758; URL: www.clinicaltrials.gov CHEST 2015; 148(1):159-168
Manuscript received October 2, 2014; revision accepted December 11, 2014; originally published Online First January 22, 2015. ABBREVIATIONS: 6MWD 5 6-min walk distance; BODE 5 BMI, measure of airflow obstruction, dyspnea score, and exercise capacity; CAT 5 COPD Assessment Test; CCQ 5 Clinical COPD Questionnaire;
journal.publications.chestnet.org
CHAIN 5 COPD History Assessment in Spain; GOLD 5 Global Initiative for Chronic Obstructive Lung Disease; mMRC 5 modified Medical Research Council; ROC 5 receiver operating characteristic; SGRQ 5 St. George’s Respiratory Questionnaire
159
Recent GOLD (Global Initiative for Chronic Obstructive Lung Disease) updates proposed important changes in the stratification of disease severity in patients with COPD.1 These recommendations were based on evidence that FEV1 is a partial descriptor of disease status. Expanding the range of evaluated symptoms by adding the assessment of dyspnea (according to the modified Medical Research Council [mMRC]) or health status (using the COPD Assessment Test [CAT] and/or the Clinical COPD Questionnaire [CCQ]) may result in more adequate assessment of patients with COPD.1 Based on expert opinions, GOLD suggests that these tools provide assessments of symptoms and that it is unnecessary and possibly confusing to use more than one scale. However, limited information is available regarding health-status tools such as the CAT and the CCQ; furthermore, the recommended cutoff points for category assignment are based
on the St. George’s Respiratory Questionnaire (SGRQ), which is a surrogate test for the CAT.2 In contrast to mMRC dyspnea,3,4 we lack information on the relationship of these tools with mortality and the possible prognostic value of the GOLD-proposed thresholds for these tools.
Materials and Methods
smoking history, dyspnea (mMRC, scale 0-4), exacerbations during the previous year, quality of life according to the Spanish versions of the CAT (scale 0-40)7 and the CCQ (scale 0-60),8 treatments, respiratory function (arterial blood gases, spirometry, lung volume, and diffusion capacity for carbon monoxide), exercise capacity (6-min walk distance [6MWD]), and BODE (BMI, measure of airflow obstruction, dyspnea score, and exercise capacity) index (scale 0-10). Data were anonymized in a database with hierarchical access control to guarantee that information was secured. All participants signed the informed consent form previously approved by the ethics committees of each participating center (Comité de Etica de la Investigación, Hospital Universitario la Candelaria, Tenerife, institutional review board no.: 258/2009).
Subjects Patients with COPD participating in this study were part of the CHAIN cohort. CHAIN is a multicenter study of 36 prospective cohorts carried out at university hospitals in Spain.5 COPD was defined by smoking history ⱖ 10 pack-years and a postbronchodilator FEV1/FVC , 0.7 after 400 mg of inhaled albuterol. Patients were stable for at least 8 weeks and received optimal medical therapy. Exclusion criteria were uncontrolled comorbidities such as malignancy at baseline or other confounding diseases that could interfere with the study. Other methodologic aspects of the CHAIN study were published previously.5 The recruitment period was January 15, 2010, to March 31, 2012 (ClinicalTrials.gov Identifier: NCT01122758). Patients are currently in the follow-up period; data analyzed in the present study came from the recruitment date up to May 15, 2014.
Therefore, in the present study, we aimed at assessing the efficacy of the CAT and the CCQ in predicting all-cause mortality in patients with COPD participating in the COPD History Assessment in Spain (CHAIN) cohort,5 a prospective Spanish multicenter study of patients with COPD. We also compared the prognostic capacities of the mMRC dyspnea, the CAT, and the CCQ. Finally, we used survival analyses to identify the best thresholds for the CAT, CCQ, and mMRC dyspnea scores to more comprehensively categorize patients with COPD by the new GOLD classification.
Clinical and Physiologic Measurements
Briefly, at baseline and each annual visit, we evaluated anthropometric data (age, sex, and BMI), comorbidities (Charlson Index, scale 0-33),6
In personal interviews, trained staff obtained information on age, sex, and BMI at the time of recruitment and at yearly appointments. BMI was calculated as weight in kilograms divided by height in meters.
AFFILIATIONS: From the Pulmonary Department (Dr Casanova), Hospital Universitario Nuestra Señora de Candelaria, Tenerife; Pulmonary Department (Dr J. M. Marin), Hospital Universitario Miguel Servet, Zaragoza; Pulmonary Department (Dr Martinez-Gonzalez), Hospital Universitario Central de Asturias, Oviedo; Pulmonary Department I (Dr de Lucas-Ramos), Hospital General Universitario Gregorio Marañón, Madrid; Pulmonary Department (Dr Mir-Viladrich), Hospital Son Llátzer, Mallorca; Pulmonary Department (Dr Cosio), Hospital Universitario Son Espases-IDISPA, Palma de Mallorca; CIBER de Enfermedades Respiratorias (CIBERES) (Drs Cosio, Peces-Barba, Balcells, and Lopez-Campos), Instituto de Salud Carlos III, Madrid; Pulmonary Department (Dr Peces-Barba), Hospital Universitario Fundación Jiménez Díaz, Madrid; Pulmonary Department (Dr Solanes-García), Hospital de la Santa Creu i Sant Pau, Barcelona; Pulmonary Department (Dr Agüero), Hospital Universitario Marqués de Valdecilla, Santander; Pulmonary Department (Dr Feu-Collado), Hospital Universitario Reina Sofía, Instituto Maimónides de Investigación Biomédica de Córdoba (IMIBIC), Universidad de Córdoba (UCO), Córdoba; Pulmonary Department (Dr Calle-Rubio), Hospital Universitario Clinico San Carlos, Madrid; Pulmonary Department (Dr Alfageme), Hospital Universitario de Valme, Sevilla; Pulmonary Department (Dr de Diego-Damia), Hospital Universitario y Politécnico de la Fe, Valencia; Pulmonary Department (Dr Irigaray), Hospital de Manacor, Mallorca; Pulmonary Department (Dr Marín), Hospital General de Castellon, Castellon; Pulmonary Department (Dr Balcells), Hospital del Mar, Barcelona;
Pulmonary Department (Dr Llunell), Hospital de Tarrasa, Tarrasa; Pulmonary Department (Dr Galdiz), Hospital Universitario de Cruces, Bilbao; Pulmonary Department (Dr Golpe), Hospital General de Calde, Lugo; Pulmonary Department (Dr Lacarcel), Hospital Ciudad de Jaén, Jaén; Pulmonary Department (Dr Cabrera), Hospital Universitario de Gran Canaria Doctor Negrín, Las Palmas de Gran Canaria; Pulmonary Department (Dr A. Marin), Hospital Universitario Germans Trias y Pujol, Badalona, Barcelona; Instituto de Investigación Hospital Universitario de la Princesa (IISP) (Dr Soriano), Universidad Autónoma de Madrid, Cátedra UAM-Linde, Madrid; Unidad Médico-Quirúrgica de Enfermedades Respiratorias, Instituto de Biomedicina de Sevilla (IBiS) (Dr Lopez-Campos), Hospital Universitario Virgen del Rocío, Sevilla; Pulmonary Department (Dr Soler-Cataluña), Hospital Universitario Arnau de Vilanova, Valencia; and Pulmonary Department (Dr de-Torres), Clínica Universidad de Navarra, Pamplona, Spain. FUNDING/SUPPORT: Financial support was provided by AstraZeneca. CORRESPONDENCE TO: Ciro Casanova, MD, Universidad de La Laguna, Tenerife, Respiratory Research Unit, Pulmonary Department, Hospital Universitario Nuestra Señora de Candelaria, Carretera del Rosario no. 145, 38010-Santa Cruz de Tenerife, Spain; e-mail: [email protected] © 2015 AMERICAN COLLEGE OF CHEST PHYSICIANS. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians. See online for more details. DOI: 10.1378/chest.14-2449
160 Original Research
[
1 4 8 # 1 C H E S T J U LY 2 0 1 5
]
A specific questionnaire was used to determine smoking status (current or former) and smoking history (in pack-years). Pulmonary function tests were performed following American Thoracic Society guidelines.9 Diffusion capacity for carbon monoxide was determined with the single-breath technique following the European Respiratory Society/American Thoracic Society guidelines.10 Pao2 was measured at rest in the sitting position while the patient breathed room air. The 6MWD test measured the better of two walks separated by at least 30 min.11 Dyspnea was evaluated with the mMRC scale.12 FEV1, BMI, 6MWD, and mMRC values were integrated into the BODE index as previously described.13 The CAT and CCQ questionnaires were self-administered by each patient with the supervision of the interviewer. Exacerbations were defined as use of antibiotics, steroids, or both or hospital admission related to worsening respiratory symptoms. All-cause mortality was recorded. Cause-specific mortality was ascertained by each site investigator to the highest detail possible and then categorized by six groups: respiratory, cardiovascular causes, lung cancer, non-lung cancer, other causes, and unknown. Information was obtained from the family and then confirmed by reviewing medical records.
TABLE 1
Statistical Analysis Data are summarized as relative frequencies for categorical variables, mean and SD for normally distributed scale variables, and median and fifth to 95th percentile for ordinal or nonnormal scale variables. Comparisons were made between groups using the Pearson x2 test, the Kruskal-Wallis H test, the Mann-Whitney U test, one-way analysis of variance, the Student t test, or the Mantel-Cox test as appropriate. A type II receiver operating characteristic (ROC) analysis with mortality as the status reference was used to determine the area under the curves for each symptom-evaluating tool, and the consistency of the differences among them was compared with the Hanley-McNeil test. The cutoff points for the CAT and the CCQ were selected using the sensitivity-specificity ratio of the mMRC dyspnea ⱖ 2 as reference. We performed Cox proportional-hazard regression analysis to determine the added precision of the CAT, CCQ, and mMRC dyspnea and its thresholds in predicting all-cause mortality in models that included all variables and those proposed in the new GOLD classification. Finally, Kaplan-Meier survival analysis for all-cause mortality using ABCD GOLD classification was performed with the current mMRC dyspnea threshold recommended by GOLD and for the new cutoffs points obtained for the CAT and the CCQ. Significance was established as two-tailed P , .05. Calculations were performed using SPSS 20.0 (IBM).
] Baseline Patient Characteristics
Variable
All Patients (N 5 768)
Age, y
68 (9)
69 (9)
62 (8)
Pack-y
56 (28)
58 (29)
47 (21)
Active smoking, %
30
a
Male Patients (n 5 634)
25
Female Patients (n 5 134)
55
BMI, kg/m2
28.1 (5.6)
28.2 (4.7)
27.1 (7.3)
FEV1, L
1.66 (0.66)
1.72 (0.67)
1.37 (0.49)
FEV1, %
60 (20)
59 (20)
62 (20)
PaO2, mm Hg
67.7 (9.9)
67.8 (9.5)
67.1 (11.7)
FVC, L
3.15 (0.95)
3.29 (0.95)
2.50 (0.63)
FVC, %
86 (23)
85 (23)
92 (24)
FEV1/FVC
53 (11)
52 (11)
54 (11)
6MWD, m
439 (105)
441 (100)
429 (122)
1 (0-4)
1 (0-4)
1 (0-4)
Dyspnea (mMRC)b
2 (0-6)
BODE index
b
2 (0-6)
1 (0-7)
IC/TLC
0.36 (0.10)
0.36 (0.11)
0.32 (0.08)
KCO, %
74 (24)
76 (24)
67 (21)
1 (0-5)
Charlson Indexb
1 (0-5)
0 (0-4)
Exacerbations
0.45 (0.03)
0.44 (0.03)
0.49 (0.07)
Hospitalizationsc
0.15 (0.02)
0.15 (0.02)
0.13 (0.03)
Inhaled anticholinergic,a %
75
74
71
Inhaled b2-agonist,a %
73
74
69
c
Inhaled corticosteroid, % a
65
66
62
CATb
11 (2-27)
11 (2-26)
14 (2-25)
CCQb
1.3 (0-3.8)
1.3 (0-3.7)
1.6 (0-3.9)
Data presented as mean (SD) unless otherwise noted. 6MWD 5 6-min walk distance; BODE 5 BMI, measure of airflow obstruction, dyspnea score, and exercise capacity; CAT 5 COPD Assessment Test; CCQ 5 Clinical COPD Questionnaire; IC 5 inspiratory capacity; KCO 5 carbon monoxide transfer coefficient; mMRC 5 modified Medical Research Council; TLC 5 total lung capacity. aNumber and/or percent. bMedian (fifth to 95th percentile). cIn the year before enrollment (percentage per patient).
journal.publications.chestnet.org
161
Results
CHAIN Cohort Follow-up and Mortality
Study Population
Median follow-up time was 38 months (fifth to 95th percentile, 22-50 months; range, 2-52 months). During follow-up, there were 73 (9.5%) deaths: 25% from respiratory causes, 16% from cardiovascular causes, 18% from lung cancer, 10% from non-lung cancer, 11% from miscellaneous causes, and 20% from unidentified causes. Baseline characteristics of survivors and nonsurvivors are shown in Table 2. Subjects who died were older, had more exacerbations and hospitalizations, had a greater degree of airway obstruction, were more hypoxic, walked less far in the 6MWD test, and had a higher BODE index than those who survived. After multivariable analysis, BODE index remained as the only independent predictor of all-cause mortality: hazard ratio 5 1.34 (95% CI: 1.14-1.56, P , .001) for BODE index.
A total of 768 patients with COPD were evaluated at baseline. The clinical and physiologic characteristics of these patients are shown in Table 1. The population was mainly men (82.5%) and included a broad range of patients with airflow obstruction: 126 mild (16.4%), 387 moderate (50.4%), 172 severe (22.4%), and 83 very severe (10.8%). Group assignment, using all metric symptoms recommended by GOLD, was as follow: 140 (18.2%) in group A, 290 (37.8%) in group B, 37 (4.8%) in group C, and 301 (39.2%) in group D. Patients reported low levels of symptoms and underwent few hospital admissions during the previous year. Approximately 75% of them used inhaled muscarinic antagonists, and a similar percentage used b2 agonists. In general, the patients displayed normal exercise capacity and little comorbidity. TABLE 2
] Characteristics of Patients With COPD at Baseline According to Mortality Outcome
Variable
Alive (n 5 695)
Sex, male (female)a
570 (125)
P Value
Dead (n 5 73) 64 (9)
.226 , .001
Age, y
67 (9)
71 (8)
Pack-y
56 (28)
56 (29)
.939
31%
.740
Active smoking, %
29%
a
BMI, kg/m2
28.1 (5.5)
27.5 (5.6)
FEV1, L
1.69 (0.65)
1.36 (0.62)
FEV1, %
60 (20)
52 (29)
.001
FVC, L
3.19 (0.94)
2.81 (0.92)
.002
FVC, %
87 (23)
80 (21)
.013
FEV1/FVC
53 (11)
PaO2, mm Hg
68.5 (9.8)
6MWD, m Dyspnea, mMRC score BODE indexb
50 (10)
376 (133)
, .001
1 (0-3)
2 (0-4)
, .001
2 (0-5)
3 (0-9)
, .001
0.36 (0.10)
0.33 (0.12)
KCO, %
75 (24)
67 (25)
Comorbidity (Charlson Index)
.046 , .001
IC/TLC
b
, .001
62.7 (9.3)
445 (94) b
.391
1 (0-5)
, .050 .029
1 (0-5)
.691
Exacerbationsc
0.43 (0.03)
0.62 (0.12)
.087
Hospitalizationsc
0.14 (0.02)
0.33 (0.08)
.025
Inhaled anticholinergic,a %
73
87
.013
Inhaled b2-agonist, %
73
76
.576
Inhaled corticosteroida
64
73
.136
a
CAT score
b
CCQ scoreb
11 (2-25)
14 (2-31)
.022
1.3 (0-3.6)
1.6 (0-4.7)
.033
Data are presented as mean (SD) unless otherwise noted. See Table 1 legend for expansion of abbreviations. aNumber and/or percent. bMedian (fifth to 95th percentile). cIn the year before enrollment (percentage per patient).
162 Original Research
[
1 4 8 # 1 C H E S T J U LY 2 0 1 5
]
Symptom-Evaluating Tools and Mortality
Subjects who died had more dyspnea and were associated with higher CAT and CCQ scores (Table 2). Figure 1 shows the ROC curves for scores from each of GOLD recommended symptom-evaluating tools (CAT, CCQ, mMRC dyspnea score). The CAT and the CCQ had similar capacities to predict mortality (C-statistic 0.589 and 0.588, respectively) in patients with COPD but inferior to that of mMRC dyspnea score (C-statistic 0.649). These differences were nonconsistent (P . .05 calculated by the Hanley-McNeil test). Both CAT scores and CCQ scores were independently associated with all-cause mortality after adjustments for age, but not when the model included all variables proposed by the GOLD multidimensional assessment and their recommended cutoffs including the two cutoff criteria for exacerbations (exacerbations ⱖ 2 or ⱖ 1 leading to hospital admission). In this model, FEV1 , 50% and ⱖ 1 hospitalizations were the most important predictors of all-cause mortality (Table 3). The current cutoffs for the CAT and CCQ recommended by GOLD were not associated with mortality (Figs 2A, 2B). mMRC dyspnea score ⱖ 2 (but not ⱖ 1, P 5 .053) was also associated with a significantly shorter all-cause survival (P 5 .001, Fig 2C). Similar results were observed with the newly determined cutoff points for the CAT (ⱖ 17) and the CCQ (. 2.5), (P 5 .004 and P 5 .003, respectively, Fig 3).
TABLE 3
] Hazard Ratios and 95% CIs for All-Cause Mortality in Patients With COPD Hazard Ratio
95% CI
FEV1 , 50%
2.05
1.27-3.31
.003
mMRC dyspnea score ⱖ 2
2.16
1.34-3.48
.001
CAT score ⱖ 10
1.32
0.81-2.17
.266
CCQ score . 1
1.38
0.84-2.25
.200
Variable
P Value
Univariate analysis
Exacerbations ⱖ 2
1.83
1.03-3.26
.039
Hospitalizationsa ⱖ 1
2.85
1.66-4.89
, .001
1.93
1.17-3.19
.010
2.59
1.47-4.55
.001
a
Multivariate analysis FEV1 , 50% Hospitalizations ⱖ 1 a
Hazard ratios and 95% CIs for all-cause mortality in patients with COPD according to univariate and multivariate logistic Cox proportionalhazard regression analyses that included all variables with their cutoff points proposed in the new GOLD combined assessment, with constant using backward stepwise method and Wald criteria per five iterations. GOLD 5 Global Initiative for Chronic Obstructive Lung Disease. See Table 1 legend for expansion of other abbreviations. aPer patient-y in the year before enrollment.
Table 4 shows mortality profiles of the ABCD classification groups using the current and new thresholds for the mMRC, CAT, and CCQ symptom measurements. Kaplan-Meier analysis including mMRC dyspnea and these new CAT and CCQ thresholds showed a significantly shorter survival in group D with all of the symptom metrics, regardless of the tools chosen. However, the overall stratification for mortality risk by GOLD categories was poor with groups A, B, and C showing important overlapping (Fig 4).
Discussion
Figure 1 – Comparative type 2 receiver operating characteristic curves for the CAT, the CCQ, and the mMRC dyspnea score, with all-cause mortality as the main outcome in patients with COPD. CAT 5 COPD Assessment Test; CCQ 5 Clinical COPD Questionnaire; mMRC 5 modified Medical Research Council.
journal.publications.chestnet.org
This observational study of patients with COPD who attended pulmonary clinics yielded several important findings. First, to our knowledge, this is the first study reporting that the CAT and the CCQ are predictors of all-cause mortality in patients with COPD. Second, these health-status questionnaires have a lower predictive capacity for all-cause mortality than the dyspnea mMRC score. This suggests that the type of tool recommended by GOLD to evaluate the symptoms domain can substantially alter the prognostic evaluation of patients, even in patients assigned to the same group. Finally, we confirmed by survival analysis that mMRC dyspnea score ⱖ 2 is a good threshold; based on these survival data, we propose new CAT (ⱖ 17) and CCQ (. 2.5) thresholds to more comprehensively categorize patients with COPD according to the new GOLD classification. 163
Figure 2 – A-C, Kaplan-Meier curves for all-cause mortality using symptom-evaluating tools and cutoff proposed by GOLD (Global Initiative for Chronic Obstructive Lung Disease): A, CAT ⱖ 10; B, CCQ . 1; C, mMRC dyspnea ⱖ 2. The survival differed significantly between groups (P 5 .001) only when the mMRC dyspnea threshold was used. See Figure 1 legend for expansion of other abbreviations.
Current guidelines for COPD management recommend multidimensional evaluation of the disease. The new GOLD strategy includes mMRC dyspnea, CAT, and the CCQ scores and states that is unnecessary to use more than one tool. However, GOLD recognized that a more comprehensive symptoms assessment should be performed, rather than generating a simple measure of breathlessness.1 The CAT and the CCQ are short and simple selfadministered questionnaires that have been developed and validated to measure clinical control in patients with COPD.14,15 Significant correlation has been demonstrated between both questionnaires and between these questionnaires and the SGRQ,4,16 but not with mMRC dyspnea.17,18 These correlations could have important implications for the new GOLD classification. A recent study reported that the use of a variety of symptomevaluating tools significantly modified grade assignment,19 but most importantly, that mMRC dyspnea, but not the CAT or the CCQ, was a predictor of mortality in patients with COPD.3,4 Therefore, one of the novel and most important findings of the present investigation was the confirmation that the CAT score and the CCQ score are predictors of mortality in patients with COPD. Both of these symptom-evaluating tools had a similar predictive 164 Original Research
capacity of all-cause mortality that was lower than the mMRC dyspnea score. Previous studies showed that the SGRQ and the Chronic Respiratory Questionnaire predicted mortality in patients with COPD independent of their degree of airway obstruction and age.20-22 According to the current investigation, the predictive capacity of those tests was lower than that of other parameters, such as exercise capacity or BODE index. On the other hand, only one study, to our knowledge, compared the predictive capacity of a health-status questionnaire (the SGRQ) with a dyspnea score, but unfortunately, in that study the latter was measured with a visual analog scale.20 The present study confirms that the instrument used to determine the level of symptoms has important consequences on the vital prognoses of these patients. This result is not surprising, as the CAT and the CCQ assess several symptoms in addition to breathlessness. Similar differences were found previously in the impact on the disease depending on the tool used to evaluate the vertical risk-assessment axis (FEV1 or exacerbations) of the new GOLD proposal.23 Although the GOLD strategy acknowledges that the new multidimensional classification was proposed for disease management and not for prognostic purposes, it would be desirable to implement a novel
[
1 4 8 # 1 C H E S T J U LY 2 0 1 5
]
Figure 3 – A, B, Kaplan-Meier curves for all-cause mortality using symptomevaluating tools by GOLD with the new cutoff that have demonstrated to predict all-causes mortality: A, CAT ⱖ 17; B, CCQ . 2.5. See Figure 1 and 2 legends for expansion of abbreviations.
grading system with the capacity to predict outcome. A review by the GOLD scientific committee recommended the implementation of studies to compare specific symptom-evaluating tools and prognosis in the each category.24 Our study sought to clarify this issue and suggest that the combined assessment should be redefined. Another important strength of the present study was our survival-based reanalysis of the cutoff points for the CAT, the CCQ, and the mMRC dyspnea proposed by GOLD. The threshold currently recommended for the
journal.publications.chestnet.org
CAT is ⱖ 10, which was previously estimated as a surrogate of the SGRQ.16 For the CCQ, the equivalent cutoff has not been precisely determined, but it appears to be in the range of 1 to 1.5 points.1 Therefore, another relevant finding of the present study was that these health-status thresholds did not predict survival. Using ROC analysis, we determined that the best predictive values of all-cause mortality were higher (CAT ⱖ 17 and CCQ . 2.5) than the thresholds currently proposed. Our proposed CAT cutoff point is similar to the one recently reported in a Canadian non-COPD population.25
165
TABLE 4
] 3-Y Mortality for Patients According to Symptom-Evaluating Tools mMRC
GOLD 2013
ⱖ2
CAT ⱖ 10
CCQ ⱖ 17
.1
. 2.5
23 (6.6)
9 (4.5)
29 (7.1)
10 (4.8)
7 (7.7)
17 (7.6)
9 (9.1)
16 (8.5)
14 (13.5)
19 (9.2)
19 (17.1)
27 (11.7)
17 (17.7)
A
13 (4.5)
14 (7)
B
13 (9.7)
C
13 (10.6)
D
28 (13.3)
30 (13.2)
4 (7.8)
Data are given as No. (%). Three-year mortality for patients according to symptom-evaluating tools with their current cutoff points by GOLD classification and the new threshold proposed in this study. See Table 1 and 3 legends for expansions of abbreviations.
Some authors have proposed to lower the cutoff point for the mMRC dyspnea score of the new GOLD classification to ⱖ 1,24 but the present study confirmed results of previous studies showing that mMRC dyspnea score ⱖ 2 is adequate for discriminating all-cause mortality. Therefore, in contrast to what we suggest for the CAT and the CCQ, we believe that the mMRC score cutoff point should not be modified. The new symptomevaluating tool thresholds proposed in the present study reached a similar mortality prediction in all GOLD categories, but most importantly showed a significant over-
lap for groups A, B, and C, mostly using the CAT score. This could be related with the great annual variability observed in the CAT scores, especially, in those with scores between 10 and 15 points.18 In fact, the low capacity of the ABCD GOLD classification to stratify mortality risk has been observed in previous studies,24 suggesting that this multidimensional assessment of patients with COPD should be refined. Another novel analysis in the present study was the result of our multivariate analysis, which included the tools recommended by the GOLD strategy using the
Figure 4 – A-C, Kaplan-Meier curves for all-cause mortality in the ABCD GOLD classification using symptom-evaluating tools with the new thresholds proposed in this study: A, CAT ⱖ 17; B, CCQ . 2.5; C, mMRC dyspnea ⱖ 2. See Figure 1 and 2 legends for expansion of abbreviations.
166 Original Research
[
1 4 8 # 1 C H E S T J U LY 2 0 1 5
]
suggested cutoffs. Our analysis further confirmed the importance of an FEV1 , 50% and a rate of annual hospitalizations ⱖ 1 as the strongest and independent predictors of all-cause mortality. Our study has several limitations. First, the CHAIN cohort is an observational study of patients with COPD attending pulmonary clinics and is not from a general medical practice or population-based study. Therefore, the cohort may not represent the true distribution of COPD severity in the general population. However, our cohort included a broad range of disease severity. Second, few women were included in the cohort, and the findings reported here cannot be extended to that sex. Nevertheless, the distribution of women into GOLD categories was similar to that of men. Third, the number of deaths is relatively small, preventing us from performing a stratified analysis of mortality according to cause. However, our all-cause mortality results associ-
Acknowledgments Author contributions: C. Casanova and J. P. d.-T. contributed to study conception and design, recruitment of patients, data analysis and interpretation, and drafting the manuscript for important intellectual content; J. M. M., C. M.-G., P. d. L.-R., I. M.-V., B. C., G. P.-B., I. S.-G., R. A., N. F.-C., M. C.-R., I. A., A. d. D.-D., R. I., M. M., E. B., A. L., J. B. G., R. G., C. L., C. Cabrera, and A. M. contributed to study conception and design, recruitment of patients, data analysis and interpretation, and review of the manuscript; and J. B. S., J. L. L.-C., and J. J. S.-C. contributed to study conception and design, data analysis and interpretation, and drafting the manuscript for important intellectual content. Financial/nonfinancial disclosures: The authors have reported to CHEST the following conflicts of interest: Dr Casanova has participated in speaking activities, industry advisory committees, and other activities related to industry sources by Almirall S.A., AstraZeneca, GlaxoSmithKline plc, and Novartis AG during the period 2012 to 2014. Dr Martinez-Gonzalez has participated in speaking activities and other related activities sponsored by Almirall S.A., Boehringer Ingelheim GmbH, Chiesi Farmaceutici S.p.A., GlaxoSmithKline plc, and Pfizer Inc during the period 2011 to 2014. Dr Solanes-García has participated in speaking activities for AstraZeneca, Biodatos Investigacion sl, Pfizer Inc, Boehringer Ingelheim GmbH, Esteve, Chiesi Farmaceutici S.p.A., GlaxoSmithKline plc, The Menarini Group, and Novartis AG. Dr Golpe has participated in speaking activities for Astra Zeneca, GlaxoSmithKline plc, Almirall S.A., Novartis, Ferrer Internacional, and Rovi Corporation. Dr de-Torres has received consultancy fees for participating in Takeda Pharmaceuticals International GmbH, The Menarini Group, and Novartis
journal.publications.chestnet.org
ated with mMRC dyspnea and other important predictors in patients with COPD, such as the BODE index, are in accordance with previous studies.3,4,13 In summary, our data, based on a large cohort of well-characterized patients, provide important information on the assessment of patients with COPD. We observed that health status, evaluated by the CAT and the CCQ, predicted mortality but with a lower capacity than the mMRC dyspnea scale. Therefore, the type of tool used to evaluate the patients’ symptoms could alter their prognostic evaluation even in those assigned to the same GOLD category. Finally, based on all-cause mortality, new cutoff points for the CAT and the CCQ were identified. According to our findings, GOLD should probably redefine the thresholds proposed for these symptom-evaluating tools, to more comprehensively assess patients with COPD. Further studies in other populations should validate our results.
AG advisory boards during the period 2010 to 2013. He has also received fees for speaking activities for GlaxoSmithKline plc, AstraZeneca, Chiesi Farmaceutici S.p.A., The Menarini Group, Novartis AG, Merck Sharp & Dohme Corp, and Takeda Pharmaceuticals International GmbH and received consultancy fees for participating in the Takeda Pharmaceuticals International GmbH and Novartis AG advisory boards during the period 2010 to 2013. Drs J. M. Marin, de Lucas-Ramos, Mir-Viladrich, Cosio, Peces-Barba, Agüero, Feu-Collado, Calle-Rubio, Alfageme, de Diego-Damia, Irigaray, Marín, Balcells, Llunell, Galdiz, Lacarcel, Cabrera, A. Marin, Soriano, Lopez-Campos, and Soler-Cataluña have reported that no potential conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article. Role of sponsors: The sponsor had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Other contributions: We thank AstraZeneca for their financial support of this study.
References 1. Vestbo J, Hurd SS, Agusti AG, et al. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: GOLD executive summary. Am J Respir Crit Care Med. 2013;187(4):347-365. 2. Jones PW, Brusselle G, Dal Negro RW, et al. Properties of the COPD assessment test in a cross-sectional European study. Eur Respir J. 2011;38(1):29-35. 3. Nishimura K, Izumi T, Tsukino M, Oga T. Dyspnea is a better predictor of 5-year survival than airway obstruction in patients with COPD. Chest. 2002;121(5): 1434-1440.
4. Almagro P, Martinez-Camblor P, Soriano JB, et al. Finding the best thresholds of FEV1 and dyspnea to predict 5-year survival in COPD patients: the COCOMICS study. Plos One. 2014;9(2):e89866. 5. López-Campos JL, Peces-Barba G, Soler-Cataluña JJ, et al; En Nombre del Grupo de Estudio CHAIN. Chronic obstructive pulmonary disease history assessment in Spain: a multidimensional chronic obstructive pulmonary disease evaluation. Study methods and organization. Arch Bronconeumol. 2012;48(12): 453-459. 6. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373-383. 7. Agustí A, Soler JJ, Molina J, et al. Is the CAT questionnaire sensitive to changes in health status in patients with severe COPD exacerbations? COPD. 2012;9(5): 492-498. 8. van der Molen T, Willemse BW, Schokker S, ten Hacken NH, Postma DS, Juniper EF. Development validity and responsiveness of the Clinical COPD Questionnaire. Health Qual Life Outcomes. 2003;1:13. 9. American Thoracic Society. Lung function testing: selection of reference values and interpretative strategies. Am Rev Respir Dis. 1991;144(5):1202-1218. 10. Macintyre N, Crapo RO, Viegi G, et al. Standardisation of the single-breath determination of carbon monoxide uptake in the lung. Eur Respir J. 2005; 26(4):720-735. 11. ATS Committee on Proficiency Standards for Clinical Pulmonary Function Laboratories. ATS statement: guidelines for the six-minute walk test. Am J Respir Crit Care Med. 2002;166(1):111-117.
167
12. Mahler DA, Wells CK. Evaluation of clinical methods for rating dyspnea. Chest. 1988;93(3):580-586. 13. Celli BR, Cote CG, Marin JM, et al. The body-mass index, airflow obstruction, dyspnea, and exercise capacity index in chronic obstructive pulmonary disease. N Engl J Med. 2004;350(10): 1005-1012. 14. Jones PW, Harding G, Berry P, Wiklund I, Chen WH, Kline Leidy N. Development and first validation of the COPD Assessment Test. Eur Respir J. 2009;34(3): 648-654. 15. van der Molen T, Willemse BW, Schokker S, ten Hacken NH, Postma DS, Juniper EF. Development, validity and responsiveness of the Clinical COPD Questionnaire. Health Qual Life Outcomes. 2003;1:13. 16. Jones PW, Tabberer M, Chen WH. Creating scenarios of the impact of COPD and their relationship to COPD Assessment Test (CAT™) scores. BMC Pulm Med. 2011;11:42.
168 Original Research
17. Han MK, Muellerova H, Curran-Everett D, et al. GOLD 2011 disease severity classification in COPDGene: a prospective cohort study. Lancet Respir Med. 2012;1(1):43-50. 18. de Torres JP, Marin JM, Martinez-Gonzalez C, et al; COPD History Assessment in Spain (CHAIN) Cohort. Clinical application of the COPD assessment test: longitudinal data from the COPD History Assessment in Spain (CHAIN) cohort. Chest. 2014;146(1):111-122. 19. Casanova C, Marin JM, Martinez-Gonzalez C, et al; COPD History Assessment In SpaiN (CHAIN) Cohort. New GOLD classification: longitudinal data on group assignment. Respir Res. 2014;15:3. 20. Domingo-Salvany A, Lamarca R, Ferrer M, et al. Health-related quality of life and mortality in male patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2002;166(5):680-685. 21. Oga T, Nishimura K, Tsukino M, Sato S, Hajiro T. Analysis of the factor related to
22.
23.
24.
25.
mortality in chronic obstructive pulmonary disease: role of exercise capacity and health status. Am J Respir Crit Care Med. 2003;167(4):544-549. Marin JM, Cote CG, Diaz O, et al. Prognostic assessment in COPD: health related quality of life and the BODE index. Respir Med. 2011;105(6): 916-921. Lange P, Marott JL, Vestbo J, et al. Prediction of the clinical course of chronic obstructive pulmonary disease, using the new GOLD classification: a study of the general population. Am J Respir Crit Care Med. 2012;186(10):975-981. Agusti A, Hurd S, Jones P, et al. FAQs about the GOLD 2011 assessment proposal of COPD: a comparative analysis of four different cohorts. Eur Respir J. 2013;42(5):1391-1401. Pinto LM, Gupta N, Tan W, et al; CanCOLD Study Group. Derivation of normative data for the COPD assessment test (CAT). Respir Res. 2014;15:68.
[
1 4 8 # 1 C H E S T J U LY 2 0 1 5
]
