The Body-Mass Index, Airflow Obstruction, Dyspnea, and Exercise Capacity Index in Chronic Obstructive Pulmonary Disease
http://www.100md.com
《新英格兰医药杂志》
ABSTRACT
Background Chronic obstructive pulmonary disease (COPD) is characterized by an incompletely reversible limitation in airflow. A physiological variable — the forced expiratory volume in one second (FEV1) — is often used to grade the severity of COPD. However, patients with COPD have systemic manifestations that are not reflected by the FEV1. We hypothesized that a multidimensional grading system that assessed the respiratory and systemic expressions of COPD would better categorize and predict outcome in these patients.
Methods We first evaluated 207 patients and found that four factors predicted the risk of death in this cohort: the body-mass index (B), the degree of airflow obstruction (O) and dyspnea (D), and exercise capacity (E), measured by the six-minute–walk test. We used these variables to construct the BODE index, a multidimensional 10-point scale in which higher scores indicate a higher risk of death. We then prospectively validated the index in a cohort of 625 patients, with death from any cause and from respiratory causes as the outcome variables.
Results There were 25 deaths among the first 207 patients and 162 deaths (26 percent) in the validation cohort. Sixty-one percent of the deaths in the validation cohort were due to respiratory insufficiency, 14 percent to myocardial infarction, 12 percent to lung cancer, and 13 percent to other causes. Patients with higher BODE scores were at higher risk for death; the hazard ratio for death from any cause per one-point increase in the BODE score was 1.34 (95 percent confidence interval, 1.26 to 1.42; P<0.001), and the hazard ratio for death from respiratory causes was 1.62 (95 percent confidence interval, 1.48 to 1.77; P<0.001). The C statistic for the ability of the BODE index to predict the risk of death was larger than that for the FEV1 (0.74 vs. 0.65).
Conclusions The BODE index, a simple multidimensional grading system, is better than the FEV1 at predicting the risk of death from any cause and from respiratory causes among patients with COPD.
Chronic obstructive pulmonary disease (COPD), a common disease characterized by a poorly reversible limitation in airflow,1 is predicted to be the third most frequent cause of death in the world by 2020.2 The risk of death in patients with COPD is often graded with the use of a single physiological variable, the forced expiratory volume in one second (FEV1).1,3,4 However, other risk factors, such as the presence of hypoxemia or hypercapnia,5,6 a short distance walked in a fixed time,7 a high degree of functional breathlessness,8 and a low body-mass index (the weight in kilograms divided by the square of the height in meters),9,10 are also associated with an increased risk of death. We hypothesized that a multidimensional grading system that assessed the respiratory, perceptive, and systemic aspects of COPD would better categorize the illness and predict the outcome than does the FEV1 alone. We used data from an initial cohort of 207 patients to identify four factors that predicted the risk of death: the body-mass index (B), the degree of airflow obstruction (O) and functional dyspnea (D), and exercise capacity (E) as assessed by the six-minute–walk test. We then integrated these variables into a multidimensional index — the BODE index — and validated the index in a second cohort of 625 patients, with death from any cause and death from respiratory causes as the outcome variables.
Methods
Between January 1997 and June 2002, a total of 859 outpatients with a wide range in the severity of COPD were recruited from clinics in the United States, Spain, and Venezuela. The study was approved by the human-research review board at each site, and all patients provided written informed consent. COPD was defined by a history of smoking that exceeded 20 pack-years and a ratio of FEV1 to forced vital capacity (FVC) of less than 0.7 measured 20 minutes after the administration of albuterol.1 All patients were in clinically stable condition and receiving appropriate therapy. Patients who were receiving inhaled oxygen had to have been taking a stable dose for at least six months before study entry. The exclusion criteria were an illness other than COPD that was likely to result in death within three years; asthma, defined as an increase in the FEV1 of more than 15 percent above the base-line value or of 200 ml after the administration of a bronchodilator; an inability to take the lung-function and six-minute–walk tests; a myocardial infarction within the preceding four months; unstable angina; or congestive heart failure (New York Heart Association class III or IV).
Variables Selected for the Bode Index
We determined the following variables in the first 207 patients who were recruited between 1995 and 1997: age; sex; pack-years of smoking; FVC; FEV1, measured in liters and as a percentage of the predicted value according to the guidelines of the American Thoracic Society11; the best of two six-minute–walk tests performed at least 30 minutes apart12; the degree of dyspnea, measured with the use of the modified Medical Research Council (MMRC) dyspnea scale13; the body-mass index9,10; the functional residual capacity and inspiratory capacity11; the hematocrit; and the albumin level. The validated Charlson index was used to determine the degree of comorbidity. This index has been shown to predict mortality.14 The differences in these values between survivors and nonsurvivors are shown in Table 1.
Table 1. Characteristics of the First 207 Patients, According to Whether They Survived.
Each of these possible explanatory variables was independently evaluated to determine its association with one-year mortality in a stepwise forward logistic-regression analysis. A subgroup of four variables had the strongest association — the body-mass index, FEV1 as a percentage of the predicted value, score on the MMRC dyspnea scale, and the distance walked in six minutes (generalized r2=0.21, P<0.001) — and these were included in the BODE index (Table 2). All these variables predict important outcomes, are easily measured, and may change over time. We chose the post-bronchodilator FEV1 as a percent of the predicted value, classified according to the three stages identified by the American Thoracic Society, because it can be used to predict health status,15 the rate of exacerbation of COPD,16 the pharmacoeconomic costs of the disease,17 and the risk of death.18,19 We chose the MMRC dyspnea scale because it predicts the likelihood of survival among patients with COPD8 and correlates well with other scales and health-status scores.20,21 We chose the six-minute–walk test because it predicts the risk of death in patients with COPD,7 patients who have undergone lung-reduction surgery,22 patients with cardiomyopathy,23 and those with pulmonary hypertension.24 In addition, the test has been standardized,12 the clinically significant thresholds have been determined,25 and it can be used to predict resource utilization.26 Finally, there is an inverse relation between body-mass index and survival9,10 that is not linear but that has an inflection point, which was 21 in our cohort and in another study.10
Table 2. Variables and Point Values Used for the Computation of the Body-Mass Index, Degree of Airflow Obstruction and Dyspnea, and Exercise Capacity (BODE) Index.
Validation of the Bode Index
The BODE index was validated prospectively in two ways in a different cohort of 625 patients who were recruited between January 1997 and January 2003. First, we used the empirical model: for each threshold value of FEV1, distance walked in six minutes, and score on the MMRC dyspnea scale shown in Table 2, the patients received points ranging from 0 (lowest value) to 3 (maximal value). For body-mass index the values were 0 or 1, because of the unique relation between body-mass index and survival described above. The points for each variable were added, so that the BODE index ranged from 0 to 10 points, with higher scores indicating a greater risk of death. In an exploratory analysis, the various components of the BODE index were assigned different weights, with no corresponding increase in predictive value.
Study Protocol
In the cohort, patients were evaluated with the use of the BODE index within six weeks after enrollment and were seen every three to six months for at least two years or until death. The patient and family were contacted if the patient failed to return for appointments. Death from any cause and from specific respiratory causes was recorded. The cause of death was determined by the investigators at each site after reviewing the medical record and death certificate.
Statistical Analysis
Data for continuous variables are presented as means ±SD. Comparison among the three countries was completed with the use of one-way analysis of variance. The differences between survivors and nonsurvivors in pulmonary-function variables and other pertinent characteristics were established with the use of t-tests for independent samples. To evaluate the capacity of the BODE index to predict the risk of death, we performed Cox proportional-hazards regression analyses.27 We estimated the hazard ratio, 95 percent confidence interval, and P value for the BODE score, before and after adjustment for coexisting conditions as measured by the Charlson index. We repeated these analyses using the BODE index as the predictor of interest in dummy-variable form, using the first quartile as the reference group. These analyses yielded estimates of risk similar to those obtained from analyses using the BODE score as a continuous variable. Thus, we focus our presentation on the predictive characteristics of the BODE index and present only bivariate results for survival according to quartiles of the BODE index in a Kaplan–Meier analysis. The statistical significance was evaluated with the use of the log-rank test. We also performed bivariate analysis on the stage of COPD according to the validated staging system of the American Thoracic Society.3
In the Cox regression analysis, we assessed the reliability of the model with the body-mass index, degree of airflow obstruction and dyspnea, and exercise capacity score as the predictor of the time to death by computing bootstrap estimates using the full sample for the hazard ratio and its 95 percent confidence interval (according to the percentile method). This approach has the advantage of not requiring that the data be split into subgroups and is more precise than alternative methods, such as cross-validation.28
Finally, in order to determine how much more precise the BODE index is than the FEV1 alone, we computed the C statistics29 for a model containing FEV1 or the BODE score as the sole independent variable. We compared the survival times and estimated the probabilities of death up to 52 months. In these analyses, the C statistic is a mathematical function of the sensitivity and specificity of the BODE index in classifying patients by means of the Cox model as either dying or surviving. The null value for the C statistic is 0.5, with a maximum of 1.0 (with higher values being better).29
Results
The validation cohort consisted primarily of elderly patients (Table 3 and Table 4) with all degrees of severity of COPD. The FEV1 was slightly lower among patients in the United States than among those in Venezuela or Spain. The U.S. patients also had more functional impairment, more severe dyspnea, and more coexisting conditions. The 27 patients (4 percent) lost to follow-up were evenly distributed according to the severity of COPD and did not differ significantly from the rest of the cohort with respect to any measured characteristic. There were 162 deaths (26 percent) over a median follow-up of 28 months (range, 4 to 68). The majority of patients (61 percent) died of respiratory insufficiency, 14 percent died of myocardial infarction, 12 percent of lung cancer, and the rest of miscellaneous causes. The BODE score was lower among survivors than among those who died from any cause (3.7±2.2 vs. 5.9±2.6, P<0.005). The score was also lower among survivors than among those who died of respiratory causes, and the difference between the scores was larger (3.6±2.2 vs. 6.7±2.3, P<0.001).
Table 3. Characteristics of the Patients in the Validation Cohort.
Table 4. Characteristics of the Validation Cohort According to Country.
Table 5 shows the BODE index as a predictor of death from any cause after correction for coexisting conditions. There were significantly more deaths in the United States (32 percent) than in Spain (15 percent) or Venezuela (13 percent) (P<0.001). However, when the analysis was done separately for each country, the predictive power of the BODE index was similar; therefore, the data are presented together. Table 5 shows that the BODE index was also a predictor of death from respiratory causes after correction for coexisting conditions (hazard ratio, 1.63; 95 percent confidence interval, 1.48 to 1.80; P<0.001). The Kaplan–Meier analysis of survival (Figure 1A) shows that each quartile increase in the BODE score was associated with increased mortality (P<0.001). Thus, the highest quartile (a BODE score of 7 to 10) was associated with a mortality rate of 80 percent at 52 months. These same data are shown in Figure 1B in relation to the severity of COPD according to the staging system of the American Thoracic Society. The C statistic for the ability of the BODE index to predict the risk of death was 0.74, as compared with a value of 0.65 with the use of FEV1 alone (expressed as a percentage of the predicted value). The computation of 2000 bootstrap samples for these data and estimation of the hazard ratios for death indicated that for each one-point increment in the BODE score the hazard ratio for death from any cause was 1.34 (95 percent confidence interval, 1.26 to 1.42) and the hazard ratio for death from a respiratory cause was 1.62 (95 percent confidence interval, 1.48 to 1.77).
Table 5. Risk of Death from Any Cause and from Respiratory Failure, Pneumonia, or Pulmonary Embolism.
Figure 1. Kaplan–Meier Survival Curves for the Four Quartiles of the Body-Mass Index, Degree of Airflow Obstruction and Dyspnea, and Exercise Capacity Index (Panel A) and the Three Stages of Severity of Chronic Obstructive Pulmonary Disease as Defined by the American Thoracic Society (Panel B).
In Panel A, quartile 1 is a score of 0 to 2, quartile 2 is a score of 3 to 4, quartile 3 a score of 5 to 6, and quartile 4 a score of 7 to 10. Survival differed significantly among the four groups (P<0.001 by the log-rank test). In Panel B, stage I is defined by a forced expiratory volume in one second (FEV1) that is more than 50 percent of the predicted value, stage II by an FEV1 that is 36 to 50 percent of the predicted value, and stage III by an FEV1 that is no more than 35 percent of the predicted value. Survival differed significantly among the three groups (P<0.001 by the log-rank test).
Discussion
We devised a simple grading system for COPD — the BODE index — and validated its use by showing that it is a better predictor of the risk of death from any cause and from respiratory causes than is the FEV1 alone. We believe that the BODE index is useful because it includes one domain that quantifies the degree of pulmonary impairment (FEV1), one that captures the patient's perception of symptoms (the MMRC dyspnea scale), and two independent domains (the distance walked in six minutes and the body-mass index) that express the systemic consequences of COPD. The FEV1 is essential for the diagnosis and quantification of the respiratory impairment resulting from COPD.1,3,4 In addition, the rate of decline in FEV1 is a good marker of disease progression and mortality.18,19 However, the FEV1 does not adequately reflect all the systemic manifestations of the disease. For example, the FEV1 correlates weakly with the degree of dyspnea,20 and the change in FEV1 does not reflect the rate of decline in patients' health.30 More important, prospective observational studies of patients with COPD have found that the degree of dyspnea8 and health-status scores31 are more accurate predictors of the risk of death than is the FEV1. Thus, although the FEV1 is important to obtain and essential in the staging of disease in any patient with COPD, other variables provide useful information that can improve the comprehensibility of the evaluation of patients with COPD. Each variable should correlate independently with the prognosis of COPD, should be easily measurable, and should serve as a surrogate for other potentially important variables.
In the BODE index, we included two descriptors of systemic involvement in COPD: the body-mass index and the distance walked in six minutes. Both are simply obtained and independently predict the risk of death.7,9,10 It is likely that they share some common underlying physiological determinants, but the distance walked in six minutes contains a degree of sensitivity not provided by the body-mass index. The six-minute–walk test is simple to perform and has been standardized.12 Its use as a clinical tool has gained acceptance, since it is a good predictor of the risk of death among patients with other chronic diseases, including congestive heart failure23 and pulmonary hypertension.24 Indeed, the distance walked in six minutes has been accepted as a good outcome measure after interventions such as pulmonary rehabilitation.32 The body-mass index was also an independent predictor of the risk of death and was therefore included in the BODE index. We evaluated the independent prognostic power of body-mass index in our cohort using different thresholds and found that values below 21 were associated with an increased risk of death, an observation similar to that reported by Landbo and coworkers in a large population study.10
The Global Initiative for Chronic Obstructive Lung Disease and the American Thoracic Society recommend that a patient's perception of dyspnea be included in any new staging system for COPD.1,3 Dyspnea represents the most disabling symptom of COPD; the degree of dyspnea provides information regarding the patient's perception of illness and can be measured. The MMRC dyspnea scale is simple to administer and correlates with other dyspnea scales20 and with scores of health status.21 Furthermore, in a large cohort of prospectively followed patients with COPD, which used the threshold values included in the BODE index, the score on the MMRC dyspnea scale was a better predictor of the risk of death than was the FEV1.8
The BODE index combines the four variables by means of a simple scale. We also explored whether weighting the variables included in the index improved the predictive power of the BODE index. Interestingly, it failed to do so, most likely because each variable included has already proved to be a good predictor of the outcome of COPD.
Our study had some limitations. First, relatively few women were recruited, even though enrollment was independent of sex. It probably reflects the problem of the underdiagnosis of COPD in women. Second, there were differences among the three countries. For example, patients in the United States had a higher mortality rate, more severe dyspnea, more functional limitations, and more coexisting conditions than patients in Venezuela or Spain, even though the severity of airflow obstruction was relatively similar among the patients as a whole. The reasons for these differences are unknown, because there have been no systematic comparisons of the regional manifestations of COPD. In all three countries, the BODE index was the best predictor of survival, an observation that renders our findings widely applicable.
Three studies have reported the effects of the grouping of variables to express the various domains affected by COPD.33,34,35 These studies did not include variables now known to be important predictors of outcome, such as the body-mass index. However, as we found in our study, they showed that the FEV1, the degree of dyspnea, and exercise performance provide independent information regarding the degree of compromise in patients with COPD.
Besides its excellent predictive power with regard to outcome, the BODE index is simple to calculate and requires no special equipment. This makes it a practical tool of potentially widespread applicability. Although the BODE index is a predictor of the risk of death, we do not know whether it will be a useful indicator of the outcome in clinical trials, the degree of utilization of health care resources, or the clinical response to therapy.
We are indebted to Dr. Gordon L. Snider, whose guidance, comments, and criticisms were fundamental to the final manuscript.
Source Information
From the COPD Center at St. Elizabeth's Medical Center, Tufts University School of Medicine, Boston (B.R.C., V.P.P.); Bay Pines Veterans Affairs Medical Center, Bay Pines, Fla. (C.G.C.); Hospital Miguel Servet, Zaragoza, Spain (J.M.M.); Hospital Nuestra Senora de La Candelaria, Tenerife, Spain (C.C.); Hospital Universitario de Caracas and Hospital Jose I. Baldo, Caracas, Venezuela (M.M.O., R.A.M.); and Boston University School of Public Health, Boston (H.J.C.).
Address reprint requests to Dr. Celli at Pulmonary and Critical Care Medicine, St. Elizabeth's Medical Center, 736 Cambridge St., Boston, MA 02135, or at bcelli@copdnet.org.
References
Pauwels RA, Buist AS, Calverley PM, Jenkins CR, Hurd SS. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: NHLBI/WHO Global Initiative for Chronic Obstructive Lung Disease (GOLD) Workshop summary. Am J Respir Crit Care Med 2001;163:1256-1276.
Murray CJL, Lopez AD. Mortality by cause for eight regions of the world: Global Burden of Disease Study. Lancet 1997;349:1269-1276.
Definitions, epidemiology, pathophysiology, diagnosis, and staging. Am J Respir Crit Care Med 1995;152:Suppl:S78-S83.
Siafakas NM, Vermeire P, Pride NB, et al. Optimal assessment and management of chronic obstructive pulmonary disease (COPD). Eur Respir J 1995;8:1398-1420.
Nocturnal Oxygen Therapy Trial Group. Continuous or nocturnal oxygen therapy in hypoxemic chronic obstructive pulmonary disease: a clinical trial. Ann Intern Med 1980;93:391-398.
Intermittent positive pressure breathing therapy of chronic obstructive pulmonary disease: a clinical trial. Ann Intern Med 1983;99:612-620.
Gerardi DA, Lovett L, Benoit-Connors ML, Reardon JZ, ZuWallack RL. Variables related to increased mortality following out-patient pulmonary rehabilitation. Eur Respir J 1996;9:431-435.
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;121:1434-1440.
Schols AM, Slangen J, Volovics L, Wouters EF. Weight loss is a reversible factor in the prognosis of chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1998;157:1791-1797.
Landbo C, Prescott E, Lange P, Vestbo J, Almdal TP. Prognostic value of nutritional status in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1999;160:1856-1861.
American Thoracic Society Statement. Lung function testing: selection of reference values and interpretative strategies. Am Rev Respir Dis 1991;144:1202-1218.
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;166:111-117.
Mahler D, Wells C. Evaluation of clinical methods for rating dyspnea. Chest 1988;93:580-586.
Charlson M, Szatrowski T, Peterson J, Gold J. Validation of a combined comorbidity index. J Clin Epidemiol 1994;47:1245-1251.
Ferrer M, Alonso J, Morera J, et al. Chronic obstructive pulmonary disease stage and health-related quality of life. Ann Intern Med 1997;127:1072-1079.
Dewan NA, Rafique S, Kanwar B, et al. Acute exacerbation of COPD: factors associated with poor treatment outcome. Chest 2000;117:662-671.
Friedman M, Serby CW, Menjoge SS, Wilson JD, Hilleman DE, Witek TJ Jr. Pharmacoeconomic evaluation of a combination of ipratropium plus albuterol compared with ipratropium alone and albuterol alone in COPD. Chest 1999;115:635-641.
Anthonisen NR, Wright EC, Hodgkin JE. Prognosis in chronic obstructive pulmonary disease. Am Rev Respir Dis 1986;133:14-20.
Burrows B. Predictors of loss of lung function and mortality in obstructive lung diseases. Eur Respir Rev 1991;1:340-5.
Mahler DA, Weinberg DH, Wells CK, Feinstein AR. 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, Ikeda A, Koyama H, Izumi T. 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;157:785-790.
Szekely LA, Oelberg DA, Wright C, et al. Preoperative predictors of operative morbidity and mortality in COPD patients undergoing bilateral lung volume reduction surgery. Chest 1997;111:550-558.
Shah M, Hasselblad V, Gheorgiadis M, et al. Prognostic usefulness of the six-minute walk in patients with advanced congestive heart failure secondary to ischemic and nonischemic cardiomyopathy. Am J Cardiol 2001;88:987-993.
Miyamoto S, Nagaya N, Satoh T, et al. Clinical correlates and prognostic significance of six-minute walk test in patients with primary pulmonary hypertension: comparison with cardiopulmonary exercise testing. Am J Respir Crit Care Med 2000;161:487-492.
Redelmeier DA, Bayoumi AM, Goldstein RS, Guyatt GH. Interpreting small differences in functional status: the Six Minute Walk test in chronic lung disease patients. Am J Respir Crit Care Med 1997;155:1278-1282.
Decramer M, Gosselink R, Troosters T, Verschueren M, Evers G. Muscle weakness is related to utilization of health care resources in COPD patients. Eur Respir J 1997;10:417-423.
Cox DR. Regression models and life-tables. J R Stat Soc 1972;34:187-220.
Harrell FE Jr, Lee KL, Mark DB. Multivariate prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med 1996;15:361-387.
Nam B-H, D'Agostino R. Discrimination index, the area under the ROC curve. In: Huber-Carol C, Balakrishnan N, Nikulin MS, Mesbah M, eds. Goodness-of-fit tests and model validity. Boston: Birkh?user, 2002:273-7.
Burge PS, Calverley PM, Jones PW, Spencer S, Anderson JA, Maslen TK. Randomised, double blind, placebo controlled study of fluticasone propionate in patients with moderate to severe chronic obstructive pulmonary disease: the ISOLDE trial. BMJ 2000;320:1297-1303.
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;166:680-685.
Pulmonary rehabilitation -- 1999. Am J Respir Crit Care Med 1999;159:1666-1682.
Ries AL, Kaplan RM, Blumberg E. Use of factor analysis to consolidate multiple outcome measures in chronic obstructive pulmonary disease. J Clin Epidemiol 1991;44:497-503.
Mahler DA, Harver A. A factor analysis of dyspnea ratings, respiratory muscle strength, and lung function in patients with chronic obstructive pulmonary disease. Am Rev Respir Dis 1992;145:467-470.
Wegner RE, Jorres RA, Kirsten DK, Magnussen H. Factor analysis of exercise capacity, dyspnoea ratings and lung function in patients with severe COPD. Eur Respir J 1994;7:725-729.
Related Letters:
Predicting Outcomes in Chronic Obstructive Pulmonary Disease
Garcia-Pachon E., Padilla-Nava I., Kostianev S. S., Iluchev D. H., Sankaranarayanan V., Ziedalski T., Gould M. K., Ussetti P., Laporta R., Celli B. R.(Bartolome R. Celli, M.D.,)
Background Chronic obstructive pulmonary disease (COPD) is characterized by an incompletely reversible limitation in airflow. A physiological variable — the forced expiratory volume in one second (FEV1) — is often used to grade the severity of COPD. However, patients with COPD have systemic manifestations that are not reflected by the FEV1. We hypothesized that a multidimensional grading system that assessed the respiratory and systemic expressions of COPD would better categorize and predict outcome in these patients.
Methods We first evaluated 207 patients and found that four factors predicted the risk of death in this cohort: the body-mass index (B), the degree of airflow obstruction (O) and dyspnea (D), and exercise capacity (E), measured by the six-minute–walk test. We used these variables to construct the BODE index, a multidimensional 10-point scale in which higher scores indicate a higher risk of death. We then prospectively validated the index in a cohort of 625 patients, with death from any cause and from respiratory causes as the outcome variables.
Results There were 25 deaths among the first 207 patients and 162 deaths (26 percent) in the validation cohort. Sixty-one percent of the deaths in the validation cohort were due to respiratory insufficiency, 14 percent to myocardial infarction, 12 percent to lung cancer, and 13 percent to other causes. Patients with higher BODE scores were at higher risk for death; the hazard ratio for death from any cause per one-point increase in the BODE score was 1.34 (95 percent confidence interval, 1.26 to 1.42; P<0.001), and the hazard ratio for death from respiratory causes was 1.62 (95 percent confidence interval, 1.48 to 1.77; P<0.001). The C statistic for the ability of the BODE index to predict the risk of death was larger than that for the FEV1 (0.74 vs. 0.65).
Conclusions The BODE index, a simple multidimensional grading system, is better than the FEV1 at predicting the risk of death from any cause and from respiratory causes among patients with COPD.
Chronic obstructive pulmonary disease (COPD), a common disease characterized by a poorly reversible limitation in airflow,1 is predicted to be the third most frequent cause of death in the world by 2020.2 The risk of death in patients with COPD is often graded with the use of a single physiological variable, the forced expiratory volume in one second (FEV1).1,3,4 However, other risk factors, such as the presence of hypoxemia or hypercapnia,5,6 a short distance walked in a fixed time,7 a high degree of functional breathlessness,8 and a low body-mass index (the weight in kilograms divided by the square of the height in meters),9,10 are also associated with an increased risk of death. We hypothesized that a multidimensional grading system that assessed the respiratory, perceptive, and systemic aspects of COPD would better categorize the illness and predict the outcome than does the FEV1 alone. We used data from an initial cohort of 207 patients to identify four factors that predicted the risk of death: the body-mass index (B), the degree of airflow obstruction (O) and functional dyspnea (D), and exercise capacity (E) as assessed by the six-minute–walk test. We then integrated these variables into a multidimensional index — the BODE index — and validated the index in a second cohort of 625 patients, with death from any cause and death from respiratory causes as the outcome variables.
Methods
Between January 1997 and June 2002, a total of 859 outpatients with a wide range in the severity of COPD were recruited from clinics in the United States, Spain, and Venezuela. The study was approved by the human-research review board at each site, and all patients provided written informed consent. COPD was defined by a history of smoking that exceeded 20 pack-years and a ratio of FEV1 to forced vital capacity (FVC) of less than 0.7 measured 20 minutes after the administration of albuterol.1 All patients were in clinically stable condition and receiving appropriate therapy. Patients who were receiving inhaled oxygen had to have been taking a stable dose for at least six months before study entry. The exclusion criteria were an illness other than COPD that was likely to result in death within three years; asthma, defined as an increase in the FEV1 of more than 15 percent above the base-line value or of 200 ml after the administration of a bronchodilator; an inability to take the lung-function and six-minute–walk tests; a myocardial infarction within the preceding four months; unstable angina; or congestive heart failure (New York Heart Association class III or IV).
Variables Selected for the Bode Index
We determined the following variables in the first 207 patients who were recruited between 1995 and 1997: age; sex; pack-years of smoking; FVC; FEV1, measured in liters and as a percentage of the predicted value according to the guidelines of the American Thoracic Society11; the best of two six-minute–walk tests performed at least 30 minutes apart12; the degree of dyspnea, measured with the use of the modified Medical Research Council (MMRC) dyspnea scale13; the body-mass index9,10; the functional residual capacity and inspiratory capacity11; the hematocrit; and the albumin level. The validated Charlson index was used to determine the degree of comorbidity. This index has been shown to predict mortality.14 The differences in these values between survivors and nonsurvivors are shown in Table 1.
Table 1. Characteristics of the First 207 Patients, According to Whether They Survived.
Each of these possible explanatory variables was independently evaluated to determine its association with one-year mortality in a stepwise forward logistic-regression analysis. A subgroup of four variables had the strongest association — the body-mass index, FEV1 as a percentage of the predicted value, score on the MMRC dyspnea scale, and the distance walked in six minutes (generalized r2=0.21, P<0.001) — and these were included in the BODE index (Table 2). All these variables predict important outcomes, are easily measured, and may change over time. We chose the post-bronchodilator FEV1 as a percent of the predicted value, classified according to the three stages identified by the American Thoracic Society, because it can be used to predict health status,15 the rate of exacerbation of COPD,16 the pharmacoeconomic costs of the disease,17 and the risk of death.18,19 We chose the MMRC dyspnea scale because it predicts the likelihood of survival among patients with COPD8 and correlates well with other scales and health-status scores.20,21 We chose the six-minute–walk test because it predicts the risk of death in patients with COPD,7 patients who have undergone lung-reduction surgery,22 patients with cardiomyopathy,23 and those with pulmonary hypertension.24 In addition, the test has been standardized,12 the clinically significant thresholds have been determined,25 and it can be used to predict resource utilization.26 Finally, there is an inverse relation between body-mass index and survival9,10 that is not linear but that has an inflection point, which was 21 in our cohort and in another study.10
Table 2. Variables and Point Values Used for the Computation of the Body-Mass Index, Degree of Airflow Obstruction and Dyspnea, and Exercise Capacity (BODE) Index.
Validation of the Bode Index
The BODE index was validated prospectively in two ways in a different cohort of 625 patients who were recruited between January 1997 and January 2003. First, we used the empirical model: for each threshold value of FEV1, distance walked in six minutes, and score on the MMRC dyspnea scale shown in Table 2, the patients received points ranging from 0 (lowest value) to 3 (maximal value). For body-mass index the values were 0 or 1, because of the unique relation between body-mass index and survival described above. The points for each variable were added, so that the BODE index ranged from 0 to 10 points, with higher scores indicating a greater risk of death. In an exploratory analysis, the various components of the BODE index were assigned different weights, with no corresponding increase in predictive value.
Study Protocol
In the cohort, patients were evaluated with the use of the BODE index within six weeks after enrollment and were seen every three to six months for at least two years or until death. The patient and family were contacted if the patient failed to return for appointments. Death from any cause and from specific respiratory causes was recorded. The cause of death was determined by the investigators at each site after reviewing the medical record and death certificate.
Statistical Analysis
Data for continuous variables are presented as means ±SD. Comparison among the three countries was completed with the use of one-way analysis of variance. The differences between survivors and nonsurvivors in pulmonary-function variables and other pertinent characteristics were established with the use of t-tests for independent samples. To evaluate the capacity of the BODE index to predict the risk of death, we performed Cox proportional-hazards regression analyses.27 We estimated the hazard ratio, 95 percent confidence interval, and P value for the BODE score, before and after adjustment for coexisting conditions as measured by the Charlson index. We repeated these analyses using the BODE index as the predictor of interest in dummy-variable form, using the first quartile as the reference group. These analyses yielded estimates of risk similar to those obtained from analyses using the BODE score as a continuous variable. Thus, we focus our presentation on the predictive characteristics of the BODE index and present only bivariate results for survival according to quartiles of the BODE index in a Kaplan–Meier analysis. The statistical significance was evaluated with the use of the log-rank test. We also performed bivariate analysis on the stage of COPD according to the validated staging system of the American Thoracic Society.3
In the Cox regression analysis, we assessed the reliability of the model with the body-mass index, degree of airflow obstruction and dyspnea, and exercise capacity score as the predictor of the time to death by computing bootstrap estimates using the full sample for the hazard ratio and its 95 percent confidence interval (according to the percentile method). This approach has the advantage of not requiring that the data be split into subgroups and is more precise than alternative methods, such as cross-validation.28
Finally, in order to determine how much more precise the BODE index is than the FEV1 alone, we computed the C statistics29 for a model containing FEV1 or the BODE score as the sole independent variable. We compared the survival times and estimated the probabilities of death up to 52 months. In these analyses, the C statistic is a mathematical function of the sensitivity and specificity of the BODE index in classifying patients by means of the Cox model as either dying or surviving. The null value for the C statistic is 0.5, with a maximum of 1.0 (with higher values being better).29
Results
The validation cohort consisted primarily of elderly patients (Table 3 and Table 4) with all degrees of severity of COPD. The FEV1 was slightly lower among patients in the United States than among those in Venezuela or Spain. The U.S. patients also had more functional impairment, more severe dyspnea, and more coexisting conditions. The 27 patients (4 percent) lost to follow-up were evenly distributed according to the severity of COPD and did not differ significantly from the rest of the cohort with respect to any measured characteristic. There were 162 deaths (26 percent) over a median follow-up of 28 months (range, 4 to 68). The majority of patients (61 percent) died of respiratory insufficiency, 14 percent died of myocardial infarction, 12 percent of lung cancer, and the rest of miscellaneous causes. The BODE score was lower among survivors than among those who died from any cause (3.7±2.2 vs. 5.9±2.6, P<0.005). The score was also lower among survivors than among those who died of respiratory causes, and the difference between the scores was larger (3.6±2.2 vs. 6.7±2.3, P<0.001).
Table 3. Characteristics of the Patients in the Validation Cohort.
Table 4. Characteristics of the Validation Cohort According to Country.
Table 5 shows the BODE index as a predictor of death from any cause after correction for coexisting conditions. There were significantly more deaths in the United States (32 percent) than in Spain (15 percent) or Venezuela (13 percent) (P<0.001). However, when the analysis was done separately for each country, the predictive power of the BODE index was similar; therefore, the data are presented together. Table 5 shows that the BODE index was also a predictor of death from respiratory causes after correction for coexisting conditions (hazard ratio, 1.63; 95 percent confidence interval, 1.48 to 1.80; P<0.001). The Kaplan–Meier analysis of survival (Figure 1A) shows that each quartile increase in the BODE score was associated with increased mortality (P<0.001). Thus, the highest quartile (a BODE score of 7 to 10) was associated with a mortality rate of 80 percent at 52 months. These same data are shown in Figure 1B in relation to the severity of COPD according to the staging system of the American Thoracic Society. The C statistic for the ability of the BODE index to predict the risk of death was 0.74, as compared with a value of 0.65 with the use of FEV1 alone (expressed as a percentage of the predicted value). The computation of 2000 bootstrap samples for these data and estimation of the hazard ratios for death indicated that for each one-point increment in the BODE score the hazard ratio for death from any cause was 1.34 (95 percent confidence interval, 1.26 to 1.42) and the hazard ratio for death from a respiratory cause was 1.62 (95 percent confidence interval, 1.48 to 1.77).
Table 5. Risk of Death from Any Cause and from Respiratory Failure, Pneumonia, or Pulmonary Embolism.
Figure 1. Kaplan–Meier Survival Curves for the Four Quartiles of the Body-Mass Index, Degree of Airflow Obstruction and Dyspnea, and Exercise Capacity Index (Panel A) and the Three Stages of Severity of Chronic Obstructive Pulmonary Disease as Defined by the American Thoracic Society (Panel B).
In Panel A, quartile 1 is a score of 0 to 2, quartile 2 is a score of 3 to 4, quartile 3 a score of 5 to 6, and quartile 4 a score of 7 to 10. Survival differed significantly among the four groups (P<0.001 by the log-rank test). In Panel B, stage I is defined by a forced expiratory volume in one second (FEV1) that is more than 50 percent of the predicted value, stage II by an FEV1 that is 36 to 50 percent of the predicted value, and stage III by an FEV1 that is no more than 35 percent of the predicted value. Survival differed significantly among the three groups (P<0.001 by the log-rank test).
Discussion
We devised a simple grading system for COPD — the BODE index — and validated its use by showing that it is a better predictor of the risk of death from any cause and from respiratory causes than is the FEV1 alone. We believe that the BODE index is useful because it includes one domain that quantifies the degree of pulmonary impairment (FEV1), one that captures the patient's perception of symptoms (the MMRC dyspnea scale), and two independent domains (the distance walked in six minutes and the body-mass index) that express the systemic consequences of COPD. The FEV1 is essential for the diagnosis and quantification of the respiratory impairment resulting from COPD.1,3,4 In addition, the rate of decline in FEV1 is a good marker of disease progression and mortality.18,19 However, the FEV1 does not adequately reflect all the systemic manifestations of the disease. For example, the FEV1 correlates weakly with the degree of dyspnea,20 and the change in FEV1 does not reflect the rate of decline in patients' health.30 More important, prospective observational studies of patients with COPD have found that the degree of dyspnea8 and health-status scores31 are more accurate predictors of the risk of death than is the FEV1. Thus, although the FEV1 is important to obtain and essential in the staging of disease in any patient with COPD, other variables provide useful information that can improve the comprehensibility of the evaluation of patients with COPD. Each variable should correlate independently with the prognosis of COPD, should be easily measurable, and should serve as a surrogate for other potentially important variables.
In the BODE index, we included two descriptors of systemic involvement in COPD: the body-mass index and the distance walked in six minutes. Both are simply obtained and independently predict the risk of death.7,9,10 It is likely that they share some common underlying physiological determinants, but the distance walked in six minutes contains a degree of sensitivity not provided by the body-mass index. The six-minute–walk test is simple to perform and has been standardized.12 Its use as a clinical tool has gained acceptance, since it is a good predictor of the risk of death among patients with other chronic diseases, including congestive heart failure23 and pulmonary hypertension.24 Indeed, the distance walked in six minutes has been accepted as a good outcome measure after interventions such as pulmonary rehabilitation.32 The body-mass index was also an independent predictor of the risk of death and was therefore included in the BODE index. We evaluated the independent prognostic power of body-mass index in our cohort using different thresholds and found that values below 21 were associated with an increased risk of death, an observation similar to that reported by Landbo and coworkers in a large population study.10
The Global Initiative for Chronic Obstructive Lung Disease and the American Thoracic Society recommend that a patient's perception of dyspnea be included in any new staging system for COPD.1,3 Dyspnea represents the most disabling symptom of COPD; the degree of dyspnea provides information regarding the patient's perception of illness and can be measured. The MMRC dyspnea scale is simple to administer and correlates with other dyspnea scales20 and with scores of health status.21 Furthermore, in a large cohort of prospectively followed patients with COPD, which used the threshold values included in the BODE index, the score on the MMRC dyspnea scale was a better predictor of the risk of death than was the FEV1.8
The BODE index combines the four variables by means of a simple scale. We also explored whether weighting the variables included in the index improved the predictive power of the BODE index. Interestingly, it failed to do so, most likely because each variable included has already proved to be a good predictor of the outcome of COPD.
Our study had some limitations. First, relatively few women were recruited, even though enrollment was independent of sex. It probably reflects the problem of the underdiagnosis of COPD in women. Second, there were differences among the three countries. For example, patients in the United States had a higher mortality rate, more severe dyspnea, more functional limitations, and more coexisting conditions than patients in Venezuela or Spain, even though the severity of airflow obstruction was relatively similar among the patients as a whole. The reasons for these differences are unknown, because there have been no systematic comparisons of the regional manifestations of COPD. In all three countries, the BODE index was the best predictor of survival, an observation that renders our findings widely applicable.
Three studies have reported the effects of the grouping of variables to express the various domains affected by COPD.33,34,35 These studies did not include variables now known to be important predictors of outcome, such as the body-mass index. However, as we found in our study, they showed that the FEV1, the degree of dyspnea, and exercise performance provide independent information regarding the degree of compromise in patients with COPD.
Besides its excellent predictive power with regard to outcome, the BODE index is simple to calculate and requires no special equipment. This makes it a practical tool of potentially widespread applicability. Although the BODE index is a predictor of the risk of death, we do not know whether it will be a useful indicator of the outcome in clinical trials, the degree of utilization of health care resources, or the clinical response to therapy.
We are indebted to Dr. Gordon L. Snider, whose guidance, comments, and criticisms were fundamental to the final manuscript.
Source Information
From the COPD Center at St. Elizabeth's Medical Center, Tufts University School of Medicine, Boston (B.R.C., V.P.P.); Bay Pines Veterans Affairs Medical Center, Bay Pines, Fla. (C.G.C.); Hospital Miguel Servet, Zaragoza, Spain (J.M.M.); Hospital Nuestra Senora de La Candelaria, Tenerife, Spain (C.C.); Hospital Universitario de Caracas and Hospital Jose I. Baldo, Caracas, Venezuela (M.M.O., R.A.M.); and Boston University School of Public Health, Boston (H.J.C.).
Address reprint requests to Dr. Celli at Pulmonary and Critical Care Medicine, St. Elizabeth's Medical Center, 736 Cambridge St., Boston, MA 02135, or at bcelli@copdnet.org.
References
Pauwels RA, Buist AS, Calverley PM, Jenkins CR, Hurd SS. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: NHLBI/WHO Global Initiative for Chronic Obstructive Lung Disease (GOLD) Workshop summary. Am J Respir Crit Care Med 2001;163:1256-1276.
Murray CJL, Lopez AD. Mortality by cause for eight regions of the world: Global Burden of Disease Study. Lancet 1997;349:1269-1276.
Definitions, epidemiology, pathophysiology, diagnosis, and staging. Am J Respir Crit Care Med 1995;152:Suppl:S78-S83.
Siafakas NM, Vermeire P, Pride NB, et al. Optimal assessment and management of chronic obstructive pulmonary disease (COPD). Eur Respir J 1995;8:1398-1420.
Nocturnal Oxygen Therapy Trial Group. Continuous or nocturnal oxygen therapy in hypoxemic chronic obstructive pulmonary disease: a clinical trial. Ann Intern Med 1980;93:391-398.
Intermittent positive pressure breathing therapy of chronic obstructive pulmonary disease: a clinical trial. Ann Intern Med 1983;99:612-620.
Gerardi DA, Lovett L, Benoit-Connors ML, Reardon JZ, ZuWallack RL. Variables related to increased mortality following out-patient pulmonary rehabilitation. Eur Respir J 1996;9:431-435.
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;121:1434-1440.
Schols AM, Slangen J, Volovics L, Wouters EF. Weight loss is a reversible factor in the prognosis of chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1998;157:1791-1797.
Landbo C, Prescott E, Lange P, Vestbo J, Almdal TP. Prognostic value of nutritional status in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1999;160:1856-1861.
American Thoracic Society Statement. Lung function testing: selection of reference values and interpretative strategies. Am Rev Respir Dis 1991;144:1202-1218.
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;166:111-117.
Mahler D, Wells C. Evaluation of clinical methods for rating dyspnea. Chest 1988;93:580-586.
Charlson M, Szatrowski T, Peterson J, Gold J. Validation of a combined comorbidity index. J Clin Epidemiol 1994;47:1245-1251.
Ferrer M, Alonso J, Morera J, et al. Chronic obstructive pulmonary disease stage and health-related quality of life. Ann Intern Med 1997;127:1072-1079.
Dewan NA, Rafique S, Kanwar B, et al. Acute exacerbation of COPD: factors associated with poor treatment outcome. Chest 2000;117:662-671.
Friedman M, Serby CW, Menjoge SS, Wilson JD, Hilleman DE, Witek TJ Jr. Pharmacoeconomic evaluation of a combination of ipratropium plus albuterol compared with ipratropium alone and albuterol alone in COPD. Chest 1999;115:635-641.
Anthonisen NR, Wright EC, Hodgkin JE. Prognosis in chronic obstructive pulmonary disease. Am Rev Respir Dis 1986;133:14-20.
Burrows B. Predictors of loss of lung function and mortality in obstructive lung diseases. Eur Respir Rev 1991;1:340-5.
Mahler DA, Weinberg DH, Wells CK, Feinstein AR. 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, Ikeda A, Koyama H, Izumi T. 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;157:785-790.
Szekely LA, Oelberg DA, Wright C, et al. Preoperative predictors of operative morbidity and mortality in COPD patients undergoing bilateral lung volume reduction surgery. Chest 1997;111:550-558.
Shah M, Hasselblad V, Gheorgiadis M, et al. Prognostic usefulness of the six-minute walk in patients with advanced congestive heart failure secondary to ischemic and nonischemic cardiomyopathy. Am J Cardiol 2001;88:987-993.
Miyamoto S, Nagaya N, Satoh T, et al. Clinical correlates and prognostic significance of six-minute walk test in patients with primary pulmonary hypertension: comparison with cardiopulmonary exercise testing. Am J Respir Crit Care Med 2000;161:487-492.
Redelmeier DA, Bayoumi AM, Goldstein RS, Guyatt GH. Interpreting small differences in functional status: the Six Minute Walk test in chronic lung disease patients. Am J Respir Crit Care Med 1997;155:1278-1282.
Decramer M, Gosselink R, Troosters T, Verschueren M, Evers G. Muscle weakness is related to utilization of health care resources in COPD patients. Eur Respir J 1997;10:417-423.
Cox DR. Regression models and life-tables. J R Stat Soc 1972;34:187-220.
Harrell FE Jr, Lee KL, Mark DB. Multivariate prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med 1996;15:361-387.
Nam B-H, D'Agostino R. Discrimination index, the area under the ROC curve. In: Huber-Carol C, Balakrishnan N, Nikulin MS, Mesbah M, eds. Goodness-of-fit tests and model validity. Boston: Birkh?user, 2002:273-7.
Burge PS, Calverley PM, Jones PW, Spencer S, Anderson JA, Maslen TK. Randomised, double blind, placebo controlled study of fluticasone propionate in patients with moderate to severe chronic obstructive pulmonary disease: the ISOLDE trial. BMJ 2000;320:1297-1303.
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;166:680-685.
Pulmonary rehabilitation -- 1999. Am J Respir Crit Care Med 1999;159:1666-1682.
Ries AL, Kaplan RM, Blumberg E. Use of factor analysis to consolidate multiple outcome measures in chronic obstructive pulmonary disease. J Clin Epidemiol 1991;44:497-503.
Mahler DA, Harver A. A factor analysis of dyspnea ratings, respiratory muscle strength, and lung function in patients with chronic obstructive pulmonary disease. Am Rev Respir Dis 1992;145:467-470.
Wegner RE, Jorres RA, Kirsten DK, Magnussen H. Factor analysis of exercise capacity, dyspnoea ratings and lung function in patients with severe COPD. Eur Respir J 1994;7:725-729.
Related Letters:
Predicting Outcomes in Chronic Obstructive Pulmonary Disease
Garcia-Pachon E., Padilla-Nava I., Kostianev S. S., Iluchev D. H., Sankaranarayanan V., Ziedalski T., Gould M. K., Ussetti P., Laporta R., Celli B. R.(Bartolome R. Celli, M.D.,)