ARTICLE IN PRESS. Katia Foglio a, Luca Bianchi a, Gisella Bruletti b, Roberto Porta a, Michele Vitacca a, Bruno Balbi c, Nicolino Ambrosino d,

Respiratory Medicine (07) 101, 1961 1970 Seven-year time course of lung function, symptoms, health-related quality of life, and exercise tolerance in COPD patients undergoing pulmonary rehabilitation programs Katia Foglio a, Luca Bianchi a, Gisella Bruletti b, Roberto Porta a, Michele Vitacca a, Bruno Balbi c, Nicolino Ambrosino d, a Pulmonary Rehabilitation and Lung Function Unit, Scientific Institute of Gussago, Italy b Psychology Unit, Scientific Institute of Gussago, Italy c Pulmonary Rehabilitation and Lung Function Unit, Scientific Institute of Veruno, Fondazione S. Maugeri IRCCS, Italy d Pulmonary Unit, Cardio-Thoracic Department, University Hospital, Pisa, Italy Received 23 February 07; accepted 10 April 07 Available online 24 May 07 KEYWORDS Dyspnoea; Health-related quality of life; Exercise tolerance Summary Aim: To evaluate the long-term course of outcome indexes in patients with chronic obstructive pulmonary disease (COPD) undergoing repeated pulmonary rehabilitation programs (PRP). Design: Prospective, observational study. Setting: Pulmonary Rehabilitation Center. Patients: Forty-eight COPD patients (M 33, age 59.678.9 years, forced expiratory volume in 1 s (FEV 1 )58716% predicted, DLCO 71717% predicted.) undergoing 5 Day-Hospital based PRPs in a period of 7.270.8 years. Measurements: Lung function, exercise capacity (incremental cycloergometry, test-6- minute walking test (6MWD)), dyspnoea (Baseline-BDI and Transitional-TDI Dyspnoea Index and Medical Research Council score MRC), health-related quality of life (HRQL) (St. George Respiratory Questionnaire (SGRQ), and the derived BODE index were assessed pre and post each PRP. Results: During follow-up, patients showed a 18722 (mean7sem) ml/year FEV 1 decline (95%CI: 24.4 to 11.6; po0.001). Exercise tolerance and BDI remained stable over time whereas SGRQ improved (DSGRQ total score: 9.6714%, po0.001). BODE index significantly worsened (from 1.2771.14 to 1.9871.64; po0.001), being this change mainly attributable to worsening in FEV 1. Each PRP elicited significant improvement in Corresponding author. U.O. Pneumologia, Dipartimento Cardio-Toracico, Azienda Ospedaliero-Universitaria Pisana, Via Paradisa 2, Cisanello 56124, Pisa, Italy. Tel.: +39 050996786; fax: +39 050996779. E-mail address: n.ambrosino@ao-pisa.toscana.it (N. Ambrosino). 0954-6111/$ - see front matter & 07 Elsevier Ltd. All rights reserved. doi:10.1016/j.rmed.07.04.007

1962 K. Foglio et al. exercise capacity, dyspnoea, SGRQ and BODE score. Post-PRP improvements in 6MWD, MRC and TDI were higher after the first three than after the last two PRPs (po0.001), whereas the greatest gain in SGRQ was observed after PRP1 and then it was lower (po0.03) but stable in the following periods of observation. Conclusion: Despite progressive lost in effectiveness of repeated PRP, COPD patients undergoing those programs do not show any significant worsening in exercise tolerance, dyspnoea and HRQL along a period of 7 years. & 07 Elsevier Ltd. All rights reserved. Introduction Chronic obstructive pulmonary disease (COPD) is a leading cause of death, a major medical and an increasing economic problem. 1 The disease results in a progressive worsening in lung function, dyspnoea, exercise capacity and healthrelated quality of life (HRQL). 2,3 Pulmonary rehabilitation has been well established as a means of enhancing standard therapy to alleviate symptoms and optimise function independent of the stage of disease. 1,4 Intensive, multidisciplinary, outpatient pulmonary rehabilitation programs (PRP) are an effective intervention in the short- and the long-term: benefits including improved exercise tolerance, symptoms, and HRQL with decrease in health care expenditures. 4,5 The benefits appear to decline after 6 12 months following the formal program, benefits in HRQL being better preserved than exercise performance, 6 and may still be identified up to 2 years after the intervention. 7 Few data are available on long-time course of outcome measures in COPD patients. Therefore, we have observed the 7-year course of pulmonary function, symptoms, exercise tolerance and HRQL in COPD patients undergoing repeated PRPs. Materials and methods The study was conducted according to the declaration of Helsinki. Informed consent for personal data treatment for routinary and scientific use was obtained by every patient at admission into our department and approved by the ethical committee of Fondazione S. Maugeri. Patients The study population included all COPD patients consecutively admitted for the first time to the Scientific Institute of Gussago, Salvatore Maugeri Foundation, for Day-Hospitalbased PRP from January 1996 to June 1998. During this period 65 COPD patients were admitted. At the end of the observational period (June 05) 48 patients had performed five PRPs. Out of the 17 patients lost during the observational period, 5 patients died due to acute respiratory failure (1 patient), lung cancer (2 patients), rupture of abdominal aortic aneurysm (1 patient) and prostatic cancer (1 patient); 12 further patients were lost at follow-up. Time elapsed from PRP1 to PRP5 was 7.270.8 years, whereas the average interval between each PRP was 1.470.5 years. The Scientific Institute of Gussago is referral rehabilitation and chronic care centre for a large geographic area in Northern Italy. After a first evaluation and performance of a PRP, patients are usually followed at outpatient clinics and repeat PRP at interval of 12 18 months according to the logistic needs of patients and hospital. Two-year follow-up data of some of these patients have been published previously. 6,7 Diagnosis of COPD was made in the presence of history of smoking (410 pack-year), a post-bronchodilator forced expiratory volume in 1 s (FEV 1 ) o80% of predicted and a FEV 1 /forced vital capacity (FVC) ratio o0.7. Patients were retrospectively stratified according to Global Initiative for Chronic Obstructive Lung Disease (GOLD) 1 stages of severity. At admission to PRP, all patients had stopped smoking for 2 10 years, were in stable conditions, as assessed by absence of worsening in symptoms, stability in blood gas values (no respiratory acidosis), and were free from exacerbations as defined by a change in the patient s baseline dyspnoea, cough and/or sputum beyond day-to-day reliability sufficient to warrant a change in management, in the 4 weeks prior to PRP. Patients with a post-bronchodilator (0 mcg of salbutamol) change in FEV 1 415% or ml were excluded from the study as well as patients with other organ failure, cancer or inability to cooperate. All patients received regular treatment with inhaled bronchodilators and inhaled steroids according to current guidelines for their disease stage. In case of exacerbation of the disease, patients could change regular treatment or add oral steroids; changes in regular treatment were, however, recorded at the moment of the next follow-up visit or admission to PRP. Pulmonary rehabilitation program Our PRP was a Day-Hospital-based multidisciplinary 8-week PRP which has been described elsewhere 6,7 including the optimisation of drug therapy, three 3-h sessions per week for 8 10 weeks, with: (i) supervised incremental exercise until achieving 30 min continuous cycling at 50 70% of the maximal load achieved on an incremental cycloergometer exercise test carried out at admission 8 ; (ii) abdominal, upper and lower limb muscle activities lifting progressively increasing light weights (300 500 gm), shoulder and full arm circling 9 ; (iii) patient and family education; (iv) nutritional programs and psychosocial counselling, when appropriate. A multidisciplinary team consisting of chest physicians, nurses, physical therapists, dietician and psychologist offered care. If desired, patients could rest in dedicated beds and have meals at the hospital facilities.

Seven-year time course in COPD patients undergoing pulmonary rehabilitation programs 1963 After discharge from each PRP, the patients were referred as outpatients for control visits every 6 months. They were encouraged to perform activities of daily life (ADL) but no structured exercise programs were prescribed. Also patients could contact the center for extra visits at any time, if needed. Outcome measurements Outcome measures were assessed prior to (T1, T3, T5, T7, T9) and at the end of each PRP (T2, T4, T6, T8, T10). Lung function Spirometry and lung volumes were assessed according to the American Thoracic Society (ATS) guidelines. 10 Dynamic lung volumes were calculated by means of a mass flow sensor (model VMax 22C Sensor Medics, Yorba Linda, CA, USA) whereas static lung volumes were measured by means of a constant volume body plethysmograph (model VMax 62 Sensor Medics, Yorba Linda, CA, USA). The predicted values according to Quanjer 11 were used. Arterial blood gases were measured by means of automated analysers (8 Ciba Corning, Medfield, MA, USA; RapidLab 865, Bayer Corporation East Walpole, MA, USA) on samples from the radial artery while the patients in the sitting position were breathing room air for at least 1 h. Maximal inspiratory (MIP) and expiratory (MEP) pressures were measured at the level of Functional Residual Capacity (FRC), 12 using a respiratory module system (Sensor Medics, Yorba Linda, CA, USA). The predicted values according to Bruschi et al. 13 were used. Diffusing capacity for carbon monoxide (DLCO) was measured by the single-breath technique using a dedicated module system (Sensor Medics, Yorba Linda, CA, USA) according to ATS standards 14 and utilising normal reference values derived from those of Crapo and colleagues. 15 Exercise capacity Symptom limited incremental exercise tests were performed on an electrically braked cycloergometer (Ergometris 800S, Sensormedics, Yorba Linda, CA, USA) using the standard 1-min incremental cycle exercise protocol as previously described. 6,7 Functional and metabolic data were determined at rest and during exercise by means of a computerised system (2900Z, Sensormedics, Yorba Linda, CA, USA). At rest and at 10-W intervals, patients were asked their perceived breathlessness and leg fatigue by pointing a number or phrase on a 10-point modified Borg scale 16 set in large type on a sheet in front of them. Exercise capacity was evaluated also by means of the 6-minute walking test (6MWD) according to standard guidelines. 17 All measurements were performed and recorded under the supervision of nurses not involved in the study. Dyspnoea Italian versions of the Medical Research Council (MRC) scale 18 and of the Baseline (BDI) and the Transitional Dyspnoea Index (TDI) 19 were used to assess chronic exertion dyspnoea, prior to each PRP and pre post-prp changes in dyspnoea, respectively. Quality of life The HRQL was evaluated by means of a validated Italian translation of the St. George s Respiratory Questionnaire (SGRQ).,21 The BODE Index, which integrates body mass index, airflow limitation (FEV 1 ), MRC score and 6MWD was also calculated retrospectively. 22 Exacerbations and hospitalisation The procedure of recording exacerbations and hospitalisation has been described elsewhere. 7 For the purpose of this study, we defined exacerbations as episodes not requiring hospitalisation but requiring a change of usual medication and prescription of systemic steroids and/or antibiotics by the general practitioner or the respiratory physician. Hospital admissions were decided by the patients general practitioners. The number of hospitalisations and exacerbations not requiring hospitalisation in the last year (follow back) prior to PRP-1 were recorded. These data were obtained asking the patients at follow-up visits, from the hospital registers, and interviewing relatives or the general practitioner. After discharge from the PRP1, patients were asked to keep a record of hospital admissions and exacerbations. Furthermore, data were collected from hospital registers, interviewing relatives, or the general practitioner. Finally, mortality rate was recorded at the follow-up visit. Statistical analysis Results are expressed as mean71 standard error of the mean (SEM) or deviation (SD) when specified. All analyses were performed using a specific software (SPSS release 12.0, SPSS Inc. Chicago, USA). Differences between paired groups of data were evaluated by T-test or Wilcoxon signed rank test, when appropriate. Time course of variables was evaluated by univariate analysis of variance (ANOVA) for repeated measure options with Bonferroni s adjustment for normally distributed variables. Time course of non-parametric variables (mainly dyspnoea scores and SGRQ) were analysed by means of Friedman s test, followed by multiple Wilcoxon signed ranks tests. According to the post-prp change in exercise performance (6MWD) and in SGRQ, the patients were divided ex post facto into two groups: improvers (either an increase in 6MWD by at least 54 meters 23 ; or a reduction in SGRQ total score by at least 4% 24 and non-improvers. w 2 analysis was performed for evaluation of differences in frequencies of non-parametric variables. Coefficients of correlation were calculated using Pearson s and Spearman s tests for parametric and non-parametric variables, respectively. A p valueo0.05 or less was considered as significant. Results Time course of pre-prp outcome measures Lung and respiratory muscle function Functional characteristics of the patients in study at each PRP admission are shown in Table 1. Patients were retrospectively classified according to GOLD 1 stages I IV taking into account FEV 1 value measured at T0. Out of 48 patients,

1964 K. Foglio et al. Table 1 Anthropometric, demographic and functional data of all patients that completed five PRPs (48 patients). Stage T1 T3 T5 T7 T9 ANOVA p Sex, M/F 33/15 Age 59.678.1 Body mass index 26.974.4 2774.6 27.274.6 27.174.3 27.374.6 0.31 BODE score, total 1.2771.14 1.2171.22 1.471.2 1.5871.47 1.9871.64 o0.001 score FVC, % predicted 79.8714.4 80.6714.7 81.1715.7 82715 81.3715.4 0.56 VC, % predicted 84715 86715 84716 84717 84715 0.30 FEV 1, ml 167560 16017555 15517541 15507576 14907559 o0.001 FEV 1, % predicted 57.7716.3 57.6716 56.4715.2 56.3716.3 54.6716.1 y 0.002 RV, % predicted 147738 145737 149738 145741 150737 0.08 TLC, % predicted 111717 111717 110717 111717 110721 0.91 MIP, % predicted 65.9716.6 65.6717.3 65.6718 66718.2 64.5718.2 0.78 MEP % predicted 68.3717 68.8717.5 68.2718.3 64.5716.9 60.1713.5 o0.001 DLCO, % predicted 70.9716.6 68.58716 70717.3 66.9716.1 63.4717.4 o0.001 PaO 2, mmhg 76.378 76.978.4 75.479.1 74.678.1 74.379.7 0.034 PaCO 2, mmhg 38.874.3 38.673.5 38.373.8 38.973.9 3973.9 0.59 ph 7.70.2 7.4170.2 7.4170.2 7.4170.2 7.4170.2 0.1 Pair-wise comparison with Bonferroni s adjustment: po0.001 versus T1. y Pair-wise comparison with Bonferroni s adjustment: p ¼ 0.027 versus T1. 6 were in GOLD stage I (12.5%), 27 in stage II (56.3%), 14 in stage III (29.2%) and 1 in GOLD stage IV (2.1%). A slight but statistically significant decline in FEV 1 was observed (ANOVA for repeated measures po0.001; mean difference T1 T9: 129.6722.8 (SEM) ml (95%CI: 64 195 ml; pair-wise comparison with Bonferroni s method po0.001), that is a 18722 ml year decline (95%CI: 24.4 to 11.6), that is about the same than in normal population. MIP did not change over time whereas MEP remained stable at the first three observations (T1 T3) and then resulted to be significantly (ANOVA po0.001) reduced over the last two observations (mean difference T1 T9: 8.2%, 95%CI: 2.2 14.2%; p ¼ 0.002), as well as DLCO (% predicted). (ANOVA po0.001; mean difference T1 T9: 7.5%; 95%CI: 3.1 11.9%; po0.001). Exercise tolerance As shown in Fig. 1 (upper-left and upper-right panels), exercise tolerance assessed at admission to each PRP did not change throughout the study period, independent of the method of evaluation, whether the 6MWD or incremental cyclo-ergometry. Subjective sensations Table 2 shows the time course of dyspnoea at admission to each PRP as assessed by the BDI and MRC. Despite some fluctuations of perceived dyspnoea over the study period (Friedmans s test p ¼ 0.05 and 0.009 for BDI and MRC, respectively), no statistically significant difference was observed in either scores at T9 when compared to T1 (Wilcoxon signed ranks tests T1 T9: p ¼ 0.31 and 0.337 for BDI and MRC, respectively) (Fig. 1, lower-left panel). Health-related quality of life Table 3 shows the time course of the HRQL assessed at admission of each PRP by means of the SGRQ, as total score and as separate domains of symptoms, activity and impact. Total score (Fig. 1, lower-right panel) significantly improved throughout the study period (Friedman s test po0.001); this improvement was equally distributed among the three domains of symptoms (Friedman s test p ¼ 0.005), activity (Friedman s test p ¼ 0.005) and impact (Friedman s test p ¼ 0.004). The greatest improvement in total score was observed between T1 and T3, whereas later on HRQL remained stable (Table 3). As also shown in Table 3, BODE index significantly worsened by 29% from T1 to T9 (1.6771.42 2.1571.66; Friedman s test p ¼ 0.004), being this change mainly attributable to the slight worsening in FEV 1 (from 1.0271 to 1.3171.03; Friedman s test p ¼ 0.002). Changes induced by PRPs Exercise tolerance As expected 7 and as shown in Fig. 1, each PRP was able to elicit a statistically significant improvement in 6MWD which, however, was less than reported as the smallest difference associated with a noticeable clinical difference for COPD patients 23 (Fig. 1, upper-left panel). Peak workload attained at exercise test improved significantly after PRP1 through PRP4, whereas it remained unchanged after PRP5 (Fig. 1, upper-right panel). The benefit in exercise tolerance gained after each PRP was lost at the following PRP admission; moreover, the extent of improvement in exercise capacity reduced significantly over time (ANOVA p ¼ 0.009 and po 0.001 for 6MWD and peak workload, respectively). In particular, the extent of improvement in peak workload was greater (p ¼ 0.001) after PRP1 and PRP3 when compared to

Seven-year time course in COPD patients undergoing pulmonary rehabilitation programs 1965 6 MWD (m) 590 570 550 530 510 490 470 450 430 410 390 370 350 T0 T1 T2 T3 T4 T5 T6 T7 T8 T9 Peak Workload (watts) 1 110 100 90 80 70 60 T0 T1 T2 T3 T4 T5 T6 T7 T8 T9 MRC Scale Score 2.2 2.1 2.0 1.9 1.8 1.7 1.6 1.5 1.4 1.3 1.2 1.1 1.0 0.9 0.8 0.7 0.6 0.5 T0 T1 T2 T3 T4 T5 T6 T7 T8 T9 SGRQ Tot (%) 60 55 50 45 35 30 25 T0 T1 T2 T3 T4 T5 T6 T7 T8 T9 Figure 1 Time course of exercise tolerance as assessed by 6MWD (upper-left panel) and cycloergometry peak workload (upper-right panel), dyspnoea as assessed by MRC scale score (lower-left panel) and HRQL (by SGRQ) lower-right panel) in the 48 patients who completed five consecutive PRPs. Lines and marks represent mean values at the different time points of the observational period. Vertical bars represent SD. po0.001 after versus before each PRP. y po0.001 versus T0. after PRP5. When considering the whole study period, the total number of improvers in 6MWD significantly decreased over time being 16 out of 48 at T2 (33%), 15 at T4 (31%), 16 at T6 (33%), 6 at T8 (12.5%): (w 2 p ¼ 0.01 versus T2) and 7 at T10 (14.6%), (w 2 p ¼ 0.03 versus T2). Dyspnoea All PRPs, but not PRP5, were followed by a significant (Wilcoxon s test po0.001) improvement of perceived dyspnoea, as assessed either by TDI or MRC score (Fig. 1, lower-left panel). Time course of changes induced by each PRP was different between the two scores (Table 2). The extent of dyspnoea relief, as assessed by TDI, induced by each PRP significantly decreased over time (Friedman s test p ¼ 0.009); in contrast, change in MRC was greatest after PRP1 and remained fairly stable after PRP2, PRP3 and PRP4, finally being least after PRP5 (Friedman s test p ¼ 0.01) (Table 2). Health-related quality of life SGRQ total score improved significantly after each PRP (Wilcoxon s test po 0.001 for all after PRP evaluations) (Fig. 1, lower-right panel), but to a greater extent following PRP1 (Table 3) than after the other post-prp evaluations. Unlike exercise tolerance (6MWD), frequency of patients attaining a clinically significant change in SGRQ did not change across the study period, being 6 out of 48 (12.5%), 4 (8.3%), 6 (12.5%), 6 (12.5%) and 7 (14.6%) at T2, T4, T6, T8 and T10, respectively. BODE total score significantly improved after each PRP (from 1.6771.42 to 1.2771.14 after PRP1, signed ranks po0.001; from 1.5271.32 to 1.2171.22 after PRP2 po0.001; from 1.7571.31 to 1.71.2 after PRP3 po0.001; from 1.9871.59 to 1.5871.47 after PRP4 po0.001 and from 2.1571.66 to 1.9871.64 after PRP5 p ¼ 0.011). This change was due to improvement in MRC score (po0.0001 after PRP1, PRP2, PRP3 and PRP4; p ¼ 0.06 after PRP5). Exacerbations and hospitalisations Frequency and distribution of exacerbations in the 1-year period prior to and in the years of the observational period are shown in Fig. 2. When compared to the year before the first PRP (2.5470.9), mean number of exacerbations/ patient observed in the following years significantly decreased (1.0270.86 po0.001, 0.8170.73 po0.001, 0.7970.77 po0.001, 0.9670.072 po0.001 after PRP1, PRP2, PRP3, PRP4, respectively). Also the number of

1966 Table 2 Time course of dyspnoea as assessed by BDI/TDI and MRC and changes induced by each PRP. T1 Mean diff. T2 T1 T3 Mean diff. T4 T3 T5 Mean diff. T6 T5 T7 Mean diff. T8 T7 T9 T10 T9 p among pre-rehab times p among postrehab changes BDI 6.871.3 7.371.5 6.671.3 6.771.3 6.571.4 0.05 y TDI 5.872.2 4.872 4.271.8 3.471.7 2.871.7 0.009 y MRC 1.570.6 0.670.5 1.470.6 0.370.5 1.670.6 0.470.5 1.870.6 0.470.5 1.670.7 0.270.4 z 0.009 y 0.01 y TDI refers to post pre-prp mean difference. y Friedman s Test Wilcoxon signed ranked test p ¼ 0.001 versus change T2 T1, T4 T3, and T6 T5, p ¼ 0.044 versus change T8 T7. z Wilcoxon signed ranked test p ¼ 0.001 versus change T2 T1, p ¼ 0.033 versus change T6 T5 and p ¼ 0.012 versus change T8 T7. Table 3 Time course of St. George Respiratory Questionnaire score and BODE index at pre-rehabilitation times and as post-rehabilitation changes ARTICLE IN PRESS T1 T2 T1 T3 T4 T3 T5 T6 T5 T7 T8 T7 T9 T10 T9 p among pre-rehab times p among post-rehab changes SGRQ, total score 45.07713.59 9.33710.27 y 37.36714.88 4.8277.66 36.22712.67 4.3176.51 35.07714.7 4.6478.77 35.714.59 5.2479.09 o0.001 0.067 Symptoms, % 52.16718.78 2.96713.89 47.13718.61 4.84710.15 42.137.06 4.53712.94 41.167.36 5.73716.41.84722.15 9.18717.26 0.005 0.198 Activity, % 58.62714.58 10.44712.84 50.67718. 11.51713.35 50.42714.92 8.31711.02 49.31718.35 2.1178.61 49.82715.91 3.13711.82 0.005 o0.001 Impact, % 34.38715.75 4.91711.12 26.93715.85 4.2478.51 26.56714.23 4.7179.02 24.84715.69 4.67710.25 25.58717.09 4.5878.53 0.004 0.023 BODE total 1.6771.42 0.470.57 1.5271.32 0.3170.51 1.7571.31 0.3570.48 1.9871.6 0.470.54 2.1571.66 0.1770.43 0.004 0.105 score BMI score 0.170.31 070 0.0870.28 070 0.0870.280 070 0.0870.280 070 0.0870.280 070 0.92 0.98 FEV 1 score 1.0271 070 0.9670.99 070 1.0870.96 070 1.171 070 1.3171.03 070 0.002 0.99 MRC score 0.570.65 0.3570.53 0.4870.54 0.3170.51 0.5870.54 0.3570.48 0.7570.6 0.3770.49 0.5870.65 0.170.37 0.01 0.26 6MWT score 0.0470.2 0.0470.2 070 070 070 070 0.0670.24 0.2170.14 0.1770.52 0.0670.24 0.011 0.82 Friedman s test. y Wilcoxon signed ranked test p ¼ 0.02 versus change T4 T3, p ¼ 0.002 versus change T6 T5, p ¼ 0.008 versus change T8 T7 and p ¼ 0.026 versus change T10 T9. K. Foglio et al.

Seven-year time course in COPD patients undergoing pulmonary rehabilitation programs 1967 cases 48 44 36 32 28 24 16 12 8 4 0 1-year pre PRP 0 1 2 3 4 number cases 48 44 36 32 28 24 16 12 8 4 0 PRP1-PRP2 0 1 2 3 4 number cases 48 44 36 32 28 24 16 12 8 4 0 PRP2-PRP3 0 1 2 3 4 number cases 48 44 36 32 28 24 16 12 8 4 0 PRP3-PRP4 0 1 2 3 4 number cases 48 44 36 32 28 24 16 12 8 4 0 PRP4-PRP5 0 1 2 3 4 number Figure 2 Frequency distribution of hospitalisations (solid bars) and exacerbations (white bars) during 1 year prior to and after PRP1, PRP2, PRP3 and PRP4. patients free from exacerbations increased significantly (w 2 po0.001) after each PRP when compared to the year before the observational period (1 patient), being 14, 17, 19 and 13 patients after PRP1, PRP2, PRP3 and PRP4, respectively. Frequency and distribution of hospitalisations in the 1- year period prior to and in the years of the observational period are also shown in Fig. 2. The mean number of hospitalisations per patient in the year before the observational period was 0.6870.09, which was significantly greater than after PRP1 (0.0870.04 po0.001), PRP2 (0.0870.05 po0.001), PRP3 (0.0270.02 po0.001) and PRP4 (0.1570.05 po0.001). Moreover, when compared to the 1-year prior to observational period (/48 patients), the number of patients free from hospitalisation was greater after PRP1 (44/48 w 2 po0.001), PRP2 (45/48 w 2 po0.0001), PRP3 (47/48 w 2 po0.001) and PRP4 (/48 w 2 po0.001) (Fig. 2). Finally, no correlation was found between decline in lung function and any of the outcome measures (before and after

1968 K. Foglio et al. PRP) as well as with rate of hospitalisations and exacerbations experienced by the patients throughout the study period. Discussion There are few studies of the long-term time course of outcome measures in COPD patients. This study shows that along a period of 7 years, COPD patients undergoing repeated PRPs do not show any significant worsening in exercise tolerance, dyspnoea, and health related quality of life. PRP appear progressively less effective in improving such outcome measures. Evidence-based support for pulmonary rehabilitation in the management of COPD patients has grown tremendously in the last decade, and this comprehensive intervention has been unequivocally demonstrated to reduce dyspnoea, increase exercise performance, and improve HRQL. Furthermore, an emerging literature is beginning to reveal its effectiveness in reducing health care costs. 4,5 Lung function An interesting result of our study is the reduced rate of decline in lung function as assessed by FEV 1. In fact our patients showed a mean 18 ml year decline in FEV 1, a value, lower than reported in some studies 25 but not in others, 26 that is about the same than in normal population. The reason for this observation is not clear. No short-term improvement in lung function is reported after PRP. 4 We can speculate that, smoking cessation, 2,27 programmed visits, and related optimisation of medical therapy might have contributed to this result. The possibility that these patients who attended five consecutive PRPs were just those whose lung function loss was the least, cannot be definitely ruled out. That would also give a plausible explanation for the small number of deaths observed throughout the study period, although an indirect relationship between mortality and exercise capacity has been suggested. 28,29 Nevertheless another explanation is possible: it has been recently shown that regular physical activity may reduce lung function decline and risk of developing COPD among active smokers. 30 Another possible explanation for this observation may be related to the reduction in number of exacerbations in the years following the PRP which confirms the results of other studies 4,7,28 (Fig. 2), although no significant correlation was found in this study. It has been reported that acute exacerbations of COPD may be important in decline of lung function. 31,32 Recently, Donaldson et al. 33 have shown that moderate to severe COPD patients who suffered from frequent exacerbations experienced a significantly greater year decline in FEV 1 of ml than patients who had infrequent exacerbations in whom the year FEV 1 decline was 32 ml. Seemungal et al. 34 demonstrated that recovery from a decline in lung function after COPD exacerbation may be very slow: at 35th day, the peak expiratory flow rate in 75% of cases had returned to baseline level, but at 91st day, the PEF in 7% of cases had not returned yet to baseline and in 3% of cases a further exacerbation had occurred before recovery was completed. Therefore, it may be possible that the observed low FEV 1 decline rate is related to reduction in exacerbations observed in our patients. An original result of our study is that MIP did not change over time whereas MEP remained stable at the first three observations and then resulted to be significantly reduced over the last two observations. The long-term course of respiratory muscle function of COPD patients has not been reported before. The lack of changes in MIP is not surprising as our PRP did not include any specific respiratory muscle training. 35 Exercise tolerance Exercise tolerance did not change throughout the study period, independent of the evaluation whether the 6MWD or incremental cycloergometry. We used incremental cycloergometry and 6MWD as indexes of exercise capacity. As reported previously these two methods to assess exercise capacity explore aspects somehow different. However, strong relationships between the two methods are reported. 36 It has been shown that in moderate-to-very severe COPD patients (mean FEV 1 ¼ 37% predicted) exercise capacity deteriorated over time at a rate defined as no less rapid than decline in airflow limitation. 37 The lack of decline in exercise capacity observed in our patients is important. It has been shown that 6MWD as well as peak VO 2 predict mortality. 38,39 Dyspnoea In agreement with stability in exercise capacity, our patients did not suffer from progressive worsening in subjective sensations of dyspnoea. This is an important issue. It has been suggested that the categorisation of patients with COPD on the basis of the level of dyspnoea was more discriminating than staging of disease severity on the basis of airway obstruction with respect to the 5-year survival. Health-related quality of life SGRQ total score significantly improved throughout the study period. The greatest improvement in total score being observed between T1 and T3, whereas later on HRQL remained stable. Other authors have found evidence of a measurable and progressive deterioration in all the components of the SGRQ of a population of COPD patients. The rate of deterioration was correlated with FEV 1 decline. 41 At difference with exercise capacity, in our study HRQL improved in the first year and thereafter remained stable. In other words, at least in this population of COPD patients undergoing several PRPs there is a dissociation between exercise capacity and HRQL. This observation has been reported also in other studies 6,7,42 and probably reflects the comprehensive nature of the PRP and the fact that HRQL depends on more than just exercise ability. Although it has been shown that hospitalisations and exacerbations are important determinants of health status in COPD patients, 43,44 whether the maintenance of good HRQL observed in patients of our study is related to reduced hospitalisations and exacerbations is still speculative.

Seven-year time course in COPD patients undergoing pulmonary rehabilitation programs 1969 BODE index, significantly worsened from T1 to T9, being this change mainly attributable to normal-like decline in FEV 1. It has been shown that the BODE index 22 captures the beneficial effects induced by pulmonary rehabilitation in COPD patients, 29 suggesting that the post-rehabilitation response in BODE may play a role in their long-term survival. In that study, 29 the BODE improved by 19% after a PRP and returned to baseline after 2 years. The BODE worsened in patients not submitted to PRP by 4% at 12 months and 18% at 2 years. As a unique result, our study is the first to describe the long-time course of this index in these patients, the overall worsening at 7 years being 29%. Being the observed worsening of BODE index mainly related to the slight decline in FEV 1, we cannot speculate on its clinical meaning in our patients, given that a physiological worsening in BODE index in normals is not reported. Changes induced by PRPs As expected, 4,7 each PRP was able to elicit significant improvement in 6MWD and in peak cycloergometry workload. Nonetheless, the benefit in exercise tolerance gained after each program was lost at the following PRP admission; moreover, the extent of improvement generated by each PRP of both 6MWD and peak work rate reduced significantly over time. Each PRP was followed by an improvement in perceived dyspnoea. SGRQ total score improved significantly after each PRP but to a greater extent following the first PRP than after the other post-prp evaluations. In summary, PRP appear progressively less effective in improving outcome measures. Limitations of the study The lack of a control group not performing PRP might be criticised. Nevertheless, given the demonstrated effectiveness of pulmonary rehabilitation, 4 it would be unethical not to perform it in a control group. In conclusion, this study shows that along a period of 7 years COPD patients undergoing repeated PRPs do not show any significant worsening in exercise tolerance, dyspnoea, and health-related quality of life. PRP appear progressively less effective in improving such outcome measures. Conflict of interest statement. None of the authors have a conflict of interest to declare in relation to this work. References 1. Global initiative for chronic obstructive lung disease. Workshop report, global strategy for diagnosis, management, and prevention of COPD. Update September 05. Bethesda, MD: National Institutes of Health, National Heart, Lung and Blood Institute; 05. Available from: /www.goldcopd.coms. 2. Downs SH, Brandli O, Zellweger JP, et al. Accelerated decline in lung function in smoking women with airway obstruction: SAPALDIA 2 cohort study. Respir Res 05;26:45. 3. Pitta F, Troosters T, Spruit MA, Probst VS, Decramer M, Gosselink R. Characteristics of physical activities in daily life in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 05;171:972 7. 4. Nici L, Donner C, Wouters E, et al. on behalf of the ATS/ERS Pulmonary Rehabilitation Writing Committee. American Thoracic Society/European Respiratory Society Statement on Pulmonary Rehabilitation. Am J Respir Crit Care Med 06;173:1390 413. 5. Griffiths TL, Bourbeau J. The economics of pulmonary rehabilitation and self-management education for patients with chronic obstructive pulmonary disease. In: Donner CF, Ambrosino N, Goldstein RS, editors. Pulmonary rehabilitation. London: Arnold Pub; 05. p. 164 72. 6. Foglio K, Bianchi L, Bruletti G, Battista L, Pagani M, Ambrosino N. Long-term effectiveness of pulmonary rehabilitation in patients with chronic airway obstruction (CAO). Eur Respir J 1999;13:125 32. 7. Foglio K, Bianchi L, Ambrosino N. Is it really useful to repeat outpatient pulmonary rehabilitation programs in patients with chronic airway obstruction? A 2-year controlled study. Chest 01;119:1696 704. 8. Maltais F, LeBlanc P, Jobin J, et al. Intensity of training and physiologic adaptation in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1997;155: 555 61. 9. Clark CJ, Cochrane L, Mackay E. Low intensity peripheral muscle conditioning improves exercise tolerance and breathlessness in COPD. Eur Respir J 1996;9:2590 6. 10. Celli BR, MacNee W, Agusti A, et al. Standards for the diagnosis and treatment of patients with COPD: a summary of the ATS/ ERS position paper. Eur Respir J 04;23:932 46. 11. Quanjer PH. Working party on Standardization of lung function test. Bull Eur Physiopathol Respir 1983;19(Suppl. 5):7 10. 12. Black L, Hyatt R. Maximal airway pressures: normal values and relationship to age and sex. Am Rev Respir Dis 1969;99: 696 702. 13. Bruschi C, Cerveri I, Zoia MC, et al. Reference values of maximal respiratory mouth pressures: a population-based study. Am Rev Respir Dis 1992;146:790 3. 14. American Thoracic Society. Single-breath carbon monoxide diffusing capacity (transfer factor): recommendations for a standard technique, 1995 upadate. Am J Respir Crit Care Med 1995;152:2185 98. 15. Crapo RO, Morris AH. Standardized single breath normal values for carbon monoxide diffusing capacity. Am Rev Respir Dis 1981;123:185 9. 16. Borg GAV. Psychophysical basis of perceived exertion. Med Sci Sports Exerc 1992;14:377 81. 17. American Thoracic Society- ATS Statement. Guidelines for the six-minute walk test. Am J Respir Crit Care Med 02;166: 111 7. 18. Fletcher CM. Standardised questionnaire on respiratory symptoms: a statement prepared and approved by the MRC committee on the aetiology of Chronic Bronchitis (MRC breathlessness score). Br Med J 1960;2:1665. 19. Mahler DA, Guyatt GH, Jones PW. Clinical measurements of dyspnea. In: Mahler DA, editor. Dyspnea. New York: Marcel Dekker; 1998. p. 149 98.. Jones PW, Quirk FH, Baveystock CM, et al. A self-complete measure of health status for chronic airflow limitation. Am Rev Respir Dis 1992;145:1321 7. 21. Carone M, Bertolotti G, Zotti AM, et al. Validation of the St. George s respiratory questionnaire Italian version. Eur Respir J 1997;10(Suppl. 25):457S. 22. 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 04;350: 1005 12. 23. Redelmeier DA, Bayoumi AM, Goldstein RS, et al. 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 82. 24. Carone M, Jones PW. Impact of health status ( quality of life ) issues in chronic lung disease. In: Donner CF, Ambrosino N,

1970 K. Foglio et al. Goldstein RS, editors. Pulmonary rehabilitation. London: Arnold Pub; 05. p. 164 72. 25. Burge PS, Calverley PMA, Jones PW, et al. Randomised, double blind, placebo controlled study of fluticasone propionate in patients with moderate to severe chronic obstructive pulmonary disease: the ISOLDE trial. Br Med J 00;3:1297 303. 26. Calverley PMA, Anderson JA, Celli B, et al., for the TORCH investigators. Salmeterol and fluticasone propionate and survival in chronic obstructive pulmonary disease. N Engl J Med 07;356:775 89. 27. Doll R, Peto R, Boreham J, et al. Mortality in relation to smoking: 50 years observations on male British doctors. Br Med J 04;328:1519 27. 28. Garcia-Aymerich J, Lange P, Benet M, Schnohr P, Anto JM. Regular physical activity reduces hospital admission and mortality in chronic obstructive pulmonary disease: a population based cohort study. Thorax 06;61:772 8. 29. Cote CG, Celli BR. Pulmonary rehabilitation and the BODE index in COPD. Eur Respir J 05;26:630 6. 30. Garcia-Aymerich J, Lange P, Benet M, Schnohr P, Anto JM. Regular physical activity modifies smoking-related lung function decline and reduces risk of Chronic Obstructive Pulmonary Disease. A population-based cohort study. Am J Respir Crit Care Med 07;175:458 63. 31. Vitacca M, Bianchi L, Ziliani M, Sarvà M, Ambrosino N. Effects of acute on chronic respiratory failure on hypercapnia and threemonth survival. Chest 05;128:19 15. 32. Vitacca M, Bianchi L, Barbano L, Ambrosino N. Lung and respiratory muscle function at discharge from a respiratory intensive care unit. Monaldi Arch Chest Dis 05;63:142 8. 33. Donaldson GC, Seemungal TAR, Bhowmik A, et al. Relationship between exacerbation frequency and lung function decline in chronic obstructive pulmonary disease. Thorax 02;57: 847 52. 34. Seemungal TAR, Donaldson GC, Bhowmik A, et al. Time course and recovery of exacerbations in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 00;161:1608 13. 35. Weiner P, McConnell A. Respiratory muscle training in chronic obstructive pulmonary disease: inspiratory, expiratory or both? Curr Opin Pulm Med 05;11:1 4. 36. Turner SE, Eastwood PR, Cecins NM, et al. Physiologic responses to incremental and self-paced exercise in COPD: a comparison of three tests. Chest 04;126:766 73. 37. Oga T, Nishimura K, Tsukino M, et al. Exercise capacity deterioration in patients with COPD. Longitudinal evaluation over 5 years. Chest 05;128:62 9. 38. Oga T, Nishimura K, Tsukino M, et al. Analysis of the factors related to mortality in chronic obstructive pulmonary disease: role of exercise capacity and health status. Am J Respir Crit Care Med 03;167:544 9. 39. Pinto-Plata VM, Cote C, Cabral H, et al. The 6-min walk distance: change over time and value as a predictor of survival in severe COPD. Eur Respir J 04;23:28 33.. Nishimura K, Izumi T, Tsukino M, et al. Dyspnea is a better predictor of 5-year survival than airway obstruction in patients with COPD. Chest 02;121:1434. 41. Spencer S, Calverley PM, Sherwood Burge P, et al. Inhaled steroids in obstructive lung disease. Health status deterioration in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 01;163:122 8. 42. Reardon J, Patel K, ZuWallack RL. Improvement in quality of life is unrelated to improvement in exercise endurance after outpatient pulmonary rehabilitation. J Cardiopulm. Rehabil 1993;13:51 4. 43. Osman LM, Godden DJ, Friend JAR, et al. Quality of life and hospital re-admission in patients with chronic obstructive pulmonary disease. Thorax 1997;52:67 71. 44. Seemungal TAR, Donaldson GC, Paul EA, et al. Effect of exacerbation on quality of life in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1998;157:1418 22.