Pulmonary rehabilitation (PR) is an important

Power of Outcome Measurements to Detect Clinically Significant Changes in Pulmonary Rehabilitation of Patients With COPD* Juan Pablo de Torres, MD; Victor Pinto-Plata, MD; Edward Ingenito, MD; Peter Bagley, MD, FCCP; Anthony Gray, MD, FCCP; Robert Berger, MD; and Bartolome Celli, MD, FCCP Study objectives: Several validated instruments are used to measure outcomes, such as exercise performance, dyspnea, and health-related quality of life after pulmonary rehabilitation (PR) in patients with COPD. However, no study has simultaneously compared the responsiveness of the most frequently used outcome measurements after PR. We designed this study to investigate the capacity of several of the most frequently used outcome measurements to detect changes after PR in a population of patients with severe COPD who qualified for lung volume reduction surgery. Design, patients, and interventions: We evaluated 37 patients with severe COPD (FEV 1 < 40%) before and after 6 to 8 weeks of outpatient PR. The following frequently used tools were evaluated: the 6-min walk distance (6MWD); functional dyspnea with the Medical Research Council (MRC) scale; baseline and transitional dyspnea index (BDI/TDI); resting and 6MWD visual analog scale (VAS); quality of life with a generic tool (the Short Form-36 [SF-36]); and two disease-specific tools, the Chronic Respiratory Disease Questionnaire (CRQ) and the St. George s Respiratory Questionnaire (SGRQ). Results: After PR, mean SD 6MWD increased in 33 of 37 patients (89%), from 285 97 to 343 92 m (p 0.009). Improvements were seen also in the MRC scale in 23 of 37 patients (62%; from 2.27 0.8 to 1.86 0.6; p 0.01); in CRQ dyspnea in 25 of 37 patients (67%; from 3.25 0.9 to 3.90 1.4; p 0.02); in CRQ mastery in 22 of 37 patients (60%; from 4.37 1.4 to 5.14 1.3; p 0.01); and in BDI/TDI functional in 24 of 37 patients (64%; from 1.4 0.8 to 0.7 1.1; p 0.002). There were smaller improvements in the SGRQ in 18 of 37 patients (48%) and in the SF-36 in 19 of 37 patients (51%), but they were not statistically significant. There were good correlations between the dyspnea components of all the tools. The 6MWD change did not correlate with the changes in the other outcomes. Clinically significant changes in the values for those outcome tools were detected in > 50% of patients for the BDI/TDI, 29% of patients for the MRC scale, in 37% of patients for the 6MWD, in 48% of patients for the VAS at peak exercise, in > 50% of patients for the CRQ, and in 40% of patients for the SGRQ. Conclusions: We conclude that the VAS peak exercise, BDI/TDI, and CRQ adequately reflect the beneficial effects of PR. The 6MWD evaluates a unique domain not related to quality of life. Due to their simplicity and sensitivity, VAS at peak exercise, 6MWD, and CRQ may be the best practical tools to evaluate responsiveness to PR. (CHEST 2002; 121:1092 1098) Key words: COPD; pulmonary rehabilitation; quality of life Abbreviations: ATS American Thoracic Society; BDI/TDI baseline and transitional dyspnea index; CRQ Chronic Respiratory Disease Questionnaire; HRQOL health-related quality of life; LVRS lung volume reduction surgery; MRC Medical Research Council; 6MWD 6-min walk distance; PR pulmonary rehabilitation; SF-36 Short-Form-36; SGRQ St. George s Respiratory Questionnaire; VAS visual analog scale *From the Pulmonary and Critical Care Division (Drs. de Torres, Pinto-Plata, and Celli), St. Elizabeth s Medical Center, Boston; Pulmonary and Critical Care Division (Dr. Ingenito), Brigham and Women s Hospital, Boston; Pulmonary and Critical Care Division (Dr. Bagley), Lahey Clinic, Boston; and Pulmonary and Critical Care Division (Dr. Gray), University of Massachusetts Medical Center, Boston; and Harvard School of Medicine (Dr. Berger), Boston, MA. Supported in part by the Overholt/Blue Cross/Blue Shield Emphysema Surgical Trial, The Thoracic Foundation, and Biovascular, Inc. Pulmonary rehabilitation (PR) is an important therapy in the management of patients with symptomatic COPD, because it improves the perception of dyspnea, exercise tolerance, and health- Manuscript received July 19, 2001; revision accepted October 23, 2001. Correspondence to: Bartolome Celli, MD, FCCP, St. Elizabeth s Medical Center, Pulmonary and Critical Care Division, 736 Cambridge St, Boston, MA 02135-2997; e-mail: bcelli@semc.org 1092 Clinical Investigations

related quality of life (HRQOL). 1 10 The effectiveness of PR has been evaluated using many different outcome tools. Functional dyspnea improvement has been documented using the Medical Research Council (MRC) scale 11 and the baseline and transitional dyspnea index (BDI/TDI), 12 whereas exercise dyspnea has been shown to improve using the visual analog scale (VAS) 13 and the Borg scale. 14 Increased exercise tolerance has been most frequently documented using the 6-min walk distance (6MWD). 15 HRQOL has been evaluated with disease-specific tools (the St. George s Respiratory Questionnaire [SGRQ] 16 and the Chronic Respiratory Disease Questionnaire [CRQ] 17 ) and also with more generic questionnaires, such as the Short Form-36 (SF-36). 18 Although all these tools are useful, they are time consuming and require training to be used and interpreted correctly. The health-care practitioner could be helped by well-validated information providing a guide to help select the simplest tools that adequately capture the changes induced by PR. To date, only two studies 19,20 have compared the power of the most frequently used outcome tools to detect different domains in cohorts of patients with COPD. However, the patients in those studies were not followed up prospectively, and the ability of the different tools to better detect changes induced by PR remains unknown. Only four studies 21 24 have prospectively compared different HRQOL questionnaires after PR. In one study, 21 the discriminative power of the questionnaires was not the primary outcome, and the information provided does not allow a conclusion about the value of either tool. One study 22 supports the validity of one general (SF-36) and one specific (CRQ) HRQOL questionnaire to detect changes after PR. The other two studies 23,24 showed that disease-specific HRQOL questionnaires are substitutive in detecting changes after PR. However, none of these studies simultaneously evaluated the change in dyspnea or the correlation between HRQOL and other outcomes such as the 6MWD. Therefore, it is entirely possible that one or two of the tools may encompass the information provided by the others and that time may be spent duplicating unnecessary information. We designed this study to investigate the capacity of several of the most frequently used outcome measurements to detect changes after PR, in a population of severe COPD patients who qualified for lung volume reduction surgery (LVRS). The tools evaluated included the following: dyspnea (VAS at rest and end of exercise, MRC scale, BDI/TDI), exercise performance (6MWD), and HRQOL (SF- 36, SGRQ, and CRQ). Study Population Materials and Methods The population consisted of the first 37 consecutive patients with severe COPD selected for LVRS at seven hospitals in Massachusetts as part of the Overholt/Blue Cross/Blue Shield Emphysema Surgical Trial. The human research committee at all institutions approved the study. All patients signed the informed consent. In summary, the study population consists of all the patients participating in an ongoing trial comparing LVRS and standard medical treatment. All patients completed 6 to 8 weeks of PR as a requisite to be randomized either to surgical or medical treatment. To be included in the study, the patients had to be 75 years old, with a medical history, physical examination, and radiologic imaging that confirmed the clinical diagnosis of emphysema. FEV 1 had to be 40%, total lung capacity had to be 125%, and residual volume had to be 175% of the predicted values. The patients were receiving maximal medical treatment defined as some form of -agonist or anticholinergic therapy, and could have received a trial of an additional agent: theophylline, oral prednisone (dose 15 mg/d), or inhaled steroids. All patients had to be able to complete the questionnaires and walk. PR Members of the Overholt/Blue Cross/Blue Shield Emphysema Surgical Trial coordinating committee trained the staff at all of the rehabilitation centers, including the adoption of a uniform exercise protocol and certification in outcome testing. Quality was ensured by a central coordinating officer. The PR program consisted of 24 sessions over 6 to 8 weeks. The sessions included warm-up and cool-down exercises, and lower-extremity exercise using a combination of treadmill, bicycle ergometer, free walking, or restorator to achieve 30 min of continuous exercise; upper-extremity exercise for a minimum of 15 min; diaphragmatic breathing and controlled coughing exercises; as well as a review of the proper use of inhalers. At-home exercise programs were prescribed, to be completed on days when the program was not attended. Patients were asked to keep a diary of activities, which was reviewed by the rehabilitation coordinator at each visit. Outcome Tools All the patients were assessed before and immediately after the PR program using the following protocol. Pulmonary function: Spirometry was performed according to American Thoracic Society (ATS) guidelines. FEV 1, FVC, and FEV 1 /FVC were calculated according to ATS criteria. 25 Dyspnea: Dyspnea was assessed using the following tools. For exercise, the VAS, 13 with anchors of no shortness of breath and extreme shortness of breath, was used. This test was evaluated at rest and at the end of 6MWD. Functional dyspnea was evaluated with two tools: the MRC scale modified by the ATS, 11 which consists of 5 increasing scores from 0 to 4 with anchors of no breathlessness except with strenuous exercise and breathless when dressing or undressing (there is no report of the change in MRC score that is associated with a clinically significant improvement or deterioration); and the BDI/TDI developed by Mahler et al, 12 which includes three domains (functional impairment, magnitude of task, and magnitude of effort) that are graded from 3to 3, where 1to 3 signifies deterioration, 0 signifies no change, and 1 to 3 signifies improvement (the change in 1 U has been thought to imply clinical significance). www.chestjournal.org CHEST / 121 / 4/ APRIL, 2002 1093

Exercise Endurance: This was evaluated using the modified 6MWD protocol. 15 In short, patients were instructed to walk at their fastest pace and to cover the longest possible distance over 6 min. Oxygen saturation was continuously monitored, supplementary oxygen was supplied to keep oxygen saturation 88%, and the equipment was carried by the technician. This could imply that different oxygen flows were used for the tests. Every 2 min, the patients were encouraged using the phrase you are doing well, keep on going. Once familiar with the instructions, and after two practice sessions, the two tests were completed in a 50-m corridor with a 30-min rest between tests. The longest walk was taken to represent the value for that visit. It has been shown that there is no significant difference after two tests. 26 The following variables were recorded: meters walked, heart and respiratory rates, BP, oxygen flow, and lowest oxygen saturation. HRQOL: HRQOL was evaluated using three questionnaires. Two are disease-specific questionnaires, the SGRQ 16 and the CRQ. The SGRQ consists of 50 items with 76 weighted responses and three component scores: symptoms, activities, and impacts (psychosocial dysfunction). A total score is calculated from all three components, with zero indicating no health impairment and 100 representing maximum impairment. A change in score of 4 U has been identified as clinically significant. The CRQ 17 is comprised of four scores: dyspnea, fatigue, emotional function, and mastery measured on a 7-point scale, with a score of 7 indicating no health impairment. The dyspnea component of the questionnaire is individualized to five activities that cause dyspnea and are assessed in order of importance and severity by the patient. A change of 0.5 U has been identified as the minimum clinically significant change. The other HRQOL tool was a generic one, the SF-36, 18,27 which consists of 36 questions that cover nine health concepts: physical function, social function, role function, role function emotional, mental health, bodily pain, vitality, general health perceptions, and health transitions. For all measures of the nine health components, scores were transformed linearly to scales of 0 to 100, with 0 indicating maximal impairment and 100 indicating the minimal impairment. The magnitude of change in the score of the SF-36 that reflects a clinically significant change has not been determined. Statistical Analysis Values are presented as mean SD unless otherwise stated. For all tools, the values obtained before and after PR were compared using a paired t test. The number of patients that improved in each outcome measurement tool was expressed in absolute number and as percentage of the total. Pearson s correlation coefficients (r) were used to examine the degree of association among dyspnea tools, dyspnea and 6MWD, HRQOL tools, 6MWD and HRQOL tools, and dyspnea tools and dyspnea components of the HRQOL tools. For all tests, a p 0.05 was considered significant. The statistics were completed using a standard statistical package (GBStat; Dynamic Microsystems; Silver Spring, MD). Results All 37 patients enrolled completed the 8 weeks of PR. Their demographic and physiologic characteristics are shown in Table 1. Both genders were represented. The patients had severe COPD and were severely hyperinflated. The pre-pr and post-pr values for the tools are Table 1 Demographic and Physiologic Data of the 37 Patients Participating in the Study* Variables Data Male/female gender, No. 18/19 Age, yr 63 6 (47 73) FEV 1,L 0.7 0.21 (0.43 1.21) FVC, L 2.11 0.7 (1.17 4.57) TLC, L 7.57 1.36 (5.25 11) RV, L 5.35 1.26 (3.38 8.3) Paco 2,mmHg 43.7 7 (32 60) Pimax, cm H 2 O 61 25 (20 107) Pemax, cm H 2 O 104 30 (65 180) *Data are presented as mean SD (range) unless otherwise indicated. TLC total lung capacity; RV residual volume; Pimax maximal inspiratory pressure; Pemax maximal expiratory pressure. shown in Table 2. There were statistical significant improvements in MRC scale (p 0.018), 6MWD (p 0.0096), VAS at peak exercise (p 0.0001), CRQ dyspnea (p 0.02), CRQ mastery (p 0.016), BDI/TDI magnitude of effort (p 0.0006), BDI/ TDI magnitude of task (p 0.0001), and BDI/TDI functional status (p 0.002). No statistical significant improvements were observed in CRQ emotional function, CRQ fatigue and in all of the domains of the SGRQ, and the SF-36. Clinically significant changes were seen in 50% of the patients (BDI/TDI, 1 U; CRQ, 0.5 U; and VAS peak exercise, 1 U). Fewer than 50% of Table 2 Outcome Measurement Tools* Measurement Tools Before PR After PR p Value MRC 2.27 0.8 1.86 0.6 0.01 VAS at rest, cm 1.1 1.5 0.9 1 NS 6MWD, m 285 97 343 92 0.009 VAS peak exercise, cm 6.2 2.5 5.01 2.8 0.0001 CRQ dyspnea 3.25 0.9 3.90 1.4 0.02 CRQ fatigue 4.05 1.2 4.52 1.1 NS CRQ emotion 4.85 1.5 5.10 1.1 NS CRQ mastery 4.37 1.4 5.14 1.3 0.01 BDI/TDI functional 1.4 0.8 0.7 1.1 0.002 BDI/TDI effort 1.6 0.7 0.9 1 0.0006 BDI/TDI task 1.7 0.6 0.8 1 0.0001 SGRQ symptoms 48 20 45 17 NS SGRQ activity 78 15 76 13 NS SGRQ impact 38 15 35 13 NS SGRQ total 52 13 48 12 NS SF-36 physical function 27 20 27 20 NS SF-36 bodily pain 87 18 87 20 NS SF-36 role function 20 34 24 33 NS SF-36 general health 51 22 49 23 NS SF-36 vitality 44 18 46 18 NS SF-36 social function 59 26 64 29 NS SF-36 role function emotion 60 46 63 36 NS SF-36 mental health 71 17 74 13 NS *Data are presented as mean SD. NS not significant. 1094 Clinical Investigations

the patients had clinically significant changes in SGRQ ( 4%), VAS at rest ( 1 U), 6MWD ( 54 m), and MRC scale ( 1). The difference in symptoms response to PR in the domain of disease-specific questionnaires (CRQ and SGRQ) is shown in Figure 1. The CRQ dyspnea tool was more sensitive to detect clinically significant changes (0.5 points) that the SGRQ (4 U), even though the latter did have a mean change of close to4u. A significant and strong correlation was seen among the dyspnea tools (MRC, VAS, CRQ dyspnea, and BDI/TDI) as well as among the HRQOL tools (CRQ, SGRQ, and SF-36) before PR. The 6MWD did not correlate with any of the other outcome tools. We also correlated the magnitude of variation ( ) Figure 1. Effect of PR on selected health status outcome. The number of patients and the degree of improvement were greater in the CRQ dyspnea domain (upper panel) than in the SGRQ symptoms domain (lower panel). Improvements in CRQ are reflected as positive changes, whereas, by design, improvements in SGRQ are reflected as negative changes. of the outcomes in 6MWD, MRC scale, VAS, CRQ dyspnea and mastery, and BDI/TDI magnitude of effort. The correlation between the changes is shown in Table 3. MRC correlated with VAS (r 0.35, p 0.05) and BDI/TDI magnitude of effort (r 0.37, p 0.05). CRQ dyspnea and mastery correlated with SGRQ impact (r 0.43 and r 0.61, respectively), and SGRQ total correlated with CRQ dyspnea (r 0.41, p 0.05), fatigue (r 0.44, p 0.05), and mastery (r 0.41, p 0.05). The change in the 6MWD did not correlate with any of the changes in the other outcome tools. Discussion This prospective study was designed to evaluate the responsiveness of the most commonly used outcome measurement tools for PR in patients with COPD, and had two major findings: (1) the clinically significant improvements in BDI/TDI, VAS at peak exercise, and the CRQ capture the beneficial effects of PR in most of the patients; and (2) the 6MWD in a unique way evaluates exercise capacity, and its change is not adequately reflected by the others tools. PR in patients with COPD provides benefits in the perception of dyspnea, exercise tolerance, and HRQOL. 1 10,28 However, the widespread availability of different outcomes measurement tools makes very difficult any direct comparison of many of these trials. Furthermore, the clinician or health-care professional involved in PR finds difficult and often times confusing the actual selection of tools to simply and effectively evaluate the effect of the programs. Four studies have compared the common tools used to assess HRQOL. 21 24 The investigators in the first study 21 showed that the CRQ was more sensitive than the SGRQ for detecting responses to PR. However, the main goal of the study was to stratify patients with the MRC scale and determine their response to PR, and not to determine the capacity of these tools to detect changes. Harper et al 22 showed that both generic (SF-36 and EuroQol) and diseasespecific (CRQ and SGRQ) questionnaires were reliable instruments over time. They also showed that generic questionnaires reflect better other health problems than disease-specific questionnaires and that the CRQ was more sensitive than the SGRQ. Singh et al 23 compared three disease-specific and two generic HRQOL questionnaires, showing that all disease-specific questionnaire findings improve after PR, but the CRQ was the most sensitive. Finally Griffiths et al 24 documented the changes observed in HRQOL 1 year after PR, but did not www.chestjournal.org CHEST / 121 / 4/ APRIL, 2002 1095

Table 3 Correlation Between the Changes in Outcome Measurements Tools Variables MRC VAS BDI/TDI 6MWD CRQ Dyspnea CRQ Mastery MRC 1 0.35* 0.37* 0.09 0.06 0.36* VAS 0.35* 1 0.28* 0.12 0 0.13 BDI/TDI 0.37* 0.28* 1 0.11 0.11 0.65* 6MWD 0.09 0.12 0.11 1 0.21 0.16 *p 0.05. evaluate the different responsiveness of the instruments used, perhaps due to the design of the study, which was to evaluate the effect of PR in utilization of health-care resources. There are two studies 19,20 that compared evaluation of dyspnea and HRQOL in a cross-sectional way in a cohort of patients with COPD. The first study 19 showed that clinical dyspnea ratings (MRC scale, baseline dyspnea index), activity of the SGRQ, and dyspnea of the CRQ performed identically in patients with COPD. The second study 20 also demonstrated that there were no substantial differences in the evaluatory properties of the CRQ and the SGRQ. However, the tools were not tested prospectively; therefore, the discriminative power of the tools to detect changes over time remains unknown. The unique feature of our study is that it compares several of the existing assessment tools using PR as the intervention in a cohort of patients with severe COPD. It is generally agreed that dyspnea is the most disabling symptom of these patients; therefore, improvement in this outcome is a very important goal of therapy. All of the studies that evaluated dyspnea as an outcome showed a significant improvement in perceived dyspnea in treated patients irrespective of the tool used. 1 9 We also observed an improvement in dyspnea that was reflected by all the dyspnea tools. Our findings are consistent with those of Hajiro et al, 19 who showed that these different tools performed similarly in a cross-sectional way. Because any one of these tools (VAS with exercise, MRC scale, dyspnea component of the CRQ, and BDI/TDI) can be used to detect the changes in dyspnea, their ease of administration could help define their use. The magnitude of change in the different tools is also important because that magnitude may indicate the clinical impact of the intervention. In this regard, the dyspnea scales behaved similarly, changing approximately by the same percentage from baseline. The MRC scale and VAS are simpler and easier to determine, and therefore could be recommended as first-line tools to evaluate dyspnea. The 6MWD is frequently used to assess the functional status and exercise capacity of patients with COPD because it is inexpensive, reliable, safe, and valid. 29 There are many studies 2,30 that have shown the beneficial effects of PR on exercise tolerance measured as the walked distance. The mean improvement for our patients was 65 m, a value that is compatible with previous studies 30 and higher than the 54 m thought to represent the minimal distance that has to improve for the COPD patients to notice a difference in their functional status. 31 The improvement in this severe group of patients does highlight the practical use of this test. However, only 29% of the patients improved 54 m in our group. It is possible that the 6MWD could have been represented by any of the other tools and thereby be not necessary in the routine evaluation of the effectiveness of PR. The results of our study suggest otherwise: the change in the distance walked did not correlate with the changes in any other outcome. We postulate that the 6MWD evaluates a unique domain. Because walking is one of the simplest and most common activities, it is an integrative useful test to assess the physical, psychological, and emotional capabilities of patients, and is not adequately assessed by any of the other tools here reported. In addition, the 6MWD has been shown to be a good predictor of survival in patients with COPD, 32 congestive heart failure, 33 and primary pulmonary hypertension. 34 Indeed, in this last group, it proved to have a better predictive power than symptomlimited cardiopulmonary exercise testing. Therefore, we believe the 6MWD should be recommended as a tool that provides independent information regarding a patient s functional status. Several studies 8,35 have shown that PR significantly improves the reported quality of life in patients with symptomatic COPD. However, few studies have directly compared the discriminative power of generic and disease-specific HRQOL tools as primary outcomes in a prospective way. 21 24 All these studies have shown that disease-specific questionnaires perform better than the generic ones. In addition, the studies 23,24 have also shown that the CRQ may be more sensitive than the SGRQ. There is only one study 20 that specifically compared the discriminative power of the different disease-specific quality-of-life questionnaires in a 1096 Clinical Investigations

cross-sectional design. The authors compared the discriminative power of the SGRQ, the CRQ, and the Breathing Problems Questionnaire. The SGRQ and the CRQ had similar discriminative power. Our results showed a correlation among the tools when tested cross-sectionally at baseline (dyspnea: VAS, BDI/TDI, MRC, and CRQ dyspnea; HRQOL: CRQ and SGRQ) but they differed when tested prospectively. The SF-36 questionnaire was less sensible to detect any change after PR. The CRQ questionnaire showed a statistically significant (p 0.05) and a clinically significant ( 0.5 points) improvement in three of its components, respectively. Furthermore, 50% of the patients showed improvement in three of the components (dyspnea, fatigue, and mastery). The mean change observed in the SGRQ almost reached the level of clinical significance ( 4 points), but it only changed in 50% of the patients. However, given the large distribution of changes in the response, it did not reach statistical significance. Perhaps with a larger number of patients, the value would have reached statistical significance. Taken as a whole, the results suggest that the CRQ is a more sensitive tool to assess the changes resulting from pulmonary rehabilitation. The differences in sensitivity may be due to the way in which the questionnaires were developed. Whereas the SGRQ and SF-36 are designed to provide a categorical response to many questions (yes-or-no format), the CRQ allows the symptoms to be expressed in a graduated scale. Because most patients with severe symptoms may only perceive subtle changes, a graded scale may better detect those changes. In addition, the CRQ allows the patient to grade the dyspnea in a group of self-selected activities, thereby magnifying those areas that may be missed in questionnaires that include a fixed set of questions. Our study findings are compatible with the article published by Singh et al, 23 who prospectively compared three different disease-specific questionnaires (SGRQ, CRQ, and the Breathing Problems Questionnaire) and two generic (global quality of life and an activity checklist), showing that all three diseasespecific questionnaires improved after PR, but the CRQ was the most sensitive. We found interesting that those parameters that improved in 50% of the patients were the same (MRC, 6MWD, CRQ dyspnea, fatigue and mastery, BDI/TDI, and VAS). These results indicate that these tools can evaluate the changes resulting from PR over time, and that they provide similar information. One possible criticism of our study is the strict selection criteria of our patients. They all had severe COPD (FEV 1 40% predicted) and were highly motivated because they were included in a surgical trial. However, the study purpose was better served by having a homogeneous population, highly compliant and with clearly defined goals. In addition, if changes were detected in a population with this severe degree of airflow limitation, we could speculate that the same outcomes could show even larger responses in patients with less severe disease. The study could also be criticized because of the relatively small number of patients. However, the total number is similar to the majority of published studies. 19 21 In addition, every patient recruited completed all aspects of the study, and there were no dropouts or missing values. In conclusion, we found that in highly motivated patients with severe COPD, the 6MWD, MRC scale, BDI/TDI, VAS exercise, and the CRQ adequately reflect the beneficial effects of PR. In addition, we have observed that the 6MWD evaluate a unique domain. The MRC scale, BDI/TDI, and CRQ dyspnea evaluate similar domains. We propose a simple testing scheme for patients undergoing PR that should include the 6MWD, the CRQ, and the VAS with exercise. Our study suggests that these are the best practical tools to evaluate dyspnea, exercise tolerance, and HRQOL, and their response to a PR program. References 1 Casaburi R. Exercise training in chronic obstructive lung disease. In: Casaburi R, Petty TL, ed. Principles and practice of pulmonary rehabilitation. Philadelphia, PA: W.B. Saunders, 1993; 361 382 2 Ries AL, Kaplan R, Linberg T, et al. Effects of pulmonary rehabilitation on physiologic and psychosocial outcomes in patients with chronic obstructive pulmonary disease. Ann Intern Med 1995; 122:823 832 3 Cockcroft A, Saunders M, Berry G. Randomized controlled trial of rehabilitation in chronic respiratory disability. Thorax 1981; 36:2003 2005 4 Sinclair D, Ingram C. Controlled trial of supervised exercise training in chronic bronchitis. Br Med J 1980; 1:519 521 5 Reardon J, Awad E, Normandin E, et al. The effect of comprehensive outpatient pulmonary rehabilitation on dyspnea. Chest 1994; 105:1046 1052 6O Donnell D, Webb H, McGuire M. Older patients with COPD: benefits of exercise training. Geriatrics 1993; 48: 59 62 7O Hara W, Lasachuk B, Matheson P, et al. Weight training and backpacking in chronic obstructive pulmonary disease. Respir Care 1984; 29:1202 1210 8 Goldstein R, Gork E, Stubbing D, et al. Randomized controlled trial of respiratory rehabilitation. Lancet 1994; 344: 1394 1397 9 Wijkstra P, Van der Mark T, Kraan J, et al. Effects of home rehabilitation on physical performance in patients with chronic obstructive pulmonary disease (COPD). Eur Respir J 1996; 9:104 110 10 Lacasse Y, Wong E, Guyatt G, et al. Health status measurement instruments in chronic obstructive pulmonary disease. Can Respir J 1997; 4:152 164 www.chestjournal.org CHEST / 121 / 4/ APRIL, 2002 1097

11 Brooks SM. Surveillance for respiratory hazards. ATS News 1982; 8:12 16 12 Mahler DA, Weinberg DH, Wells CK, et al. The measurement of dyspnea: contents, interobserver agreement, and physiologic correlates of two new clinical indexes. Chest 1984; 85:751 758 13 Stervus SS, Galunter EH. Ratio scales and category scales for a dozen perceptual continua. J Exp Psychol 1957; 54:377 411 14 Borg GAV. A category scale with ratio properties for intermodal and interindividual comparisons. In: Geissler HG, Petzold P, eds. Psychophysical judgment and the process of perception. Berlin, Germany: Veb Deuscher Verlag der Wissen Schaften, 1982; 25 34 15 Guyatt GH, Pugsley SQ, Sullivan MJ, et al. Effect on encouragement on walking test performance. Thorax 1984; 39:818 822 16 Jones PW, Quirk FH, Baveystock CM, et al. A self complete measure for chronic airflow limitation: the St. George s Respiratory Questionnaire. Am Rev Respir Dis 1992; 145: 1321 1327 17 Guyatt GH, Berman LB, Townsend M, et al. A measure of quality of life for clinical trials in chronic lung disease. Thorax 1987; 42:773 778 18 Mahler D, Mackowiak J. Evaluation of Short-Form 36-item questionnaire to measure health-related quality of life in patients with COPD. Chest 1995; 107:1585 1589 19 Hajiro K, Tsukino M, Ikeda A, et al. Analysis of clinical methods to evaluate dyspnea in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1998; 158:1185 1189 20 Hajiro T, Nishimura K, Tsukino M, et al. 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 21 Wedzicha J, Bestall J, Garrod R, et al. Randomized controlled trial of pulmonary rehabilitation in severe chronic obstructive pulmonary disease patients, stratified with the MRC dyspnea scale. Eur Respir J 1998 12:363 369 22 Harper R, Brazier JE, Waterhouse JC, et al. Comparison of outcome measures for patients with chronic obstructive pulmonary disease (COPD) in outpatient setting. Thorax 1997; 52:879 887 23 Singh SJ, Sodergan SC, Hyland ME, et al. A comparison of three disease specific and two generic health-status measures to evaluate the outcome of the pulmonary rehabilitation of COPD. Respir Med 2001; 95:71 77 24 Griffiths TL, Burr ML, Campbell IA, et al. Results at 1 year of outpatient multidisciplinary pulmonary rehabilitation: a randomized controlled trial. Lancet 2000; 355:362 368 25 American Thoracic Society. Standards for the diagnosis and care of patients with chronic pulmonary obstructive disease (COPD) and asthma. Am Rev Respir Dis 1987; 136:225 244 26 Roomi J, Johnson MM, Waters K, et al. Respiratory rehabilitation, exercise capacity and quality of life in chronic airways disease in old age. Age Ageing 1996; 25:12 16 27 Ware Jr JE, Sherbourne CD. The MOS-item short-form health survey (SF-36): I. Conceptual framework and item selection. Med Care 1992; 30:473 481 28 Lacasse Y, Wong E, Guyatt GH, et al. Meta-analysis of respiratory rehabilitation in chronic obstructive pulmonary disease. Lancet 1996; 348:1115 1119 29 Jones NL, Campbell EJM, Edwards RHT, et al. Clinical exercise testing. Philadelphia, PA: W.B. Saunders, 1975 30 ZuWallack R, Patel K, Reardon J, et al. Predictors of improvement in the 12 minute walking distance following a six-week outpatient pulmonary rehabilitation program. Chest 1991; 99:805 808 31 Redelmeier DA, Bayoumi AM, Goldstein RS, et al. Interpreting small differences in functional status: the six minute walk test in chronic lung disease. Am J Respir Crit Care Med 1997; 155:1278 1282 32 Gerardi DA, Lovett L, Benoit-Connors, et al. Variables related to increased mortality following out-patient pulmonary rehabilitation. Eur Respir J 1996; 9:431 435 33 Bitter V, Weiner D, Yusuf S, et al. Prediction of mortality and morbidity with six minute walk test in patients with left ventricular dysfunction. JAMA 1993; 270:1702 1707 34 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 35 Guell R, Casan P, Belda J, et al. Long term effects of outpatient rehabilitation of COPD. Chest 2000; 117:976 983 1098 Clinical Investigations