Treatment of COPD is largely palliative, directed

Transcription

1 Evaluation of Health-Related Quality of Life in Low-Income Patients With COPD Receiving Long-term Oxygen Therapy* Cláudia Adriana Sant Anna Ferreira, PT; Rafael Stelmach, MD; Maria Ignez Zanetti Feltrin, PT; Wilson Jacob Filho, MD; Toshio Chiba, MD; and Alberto Cukier, MD, FCCP Study objective: To assess health-related quality of life (HRQL) in a low-income population of patients with hypoxemia and COPD receiving long-term oxygen therapy (LTOT). Design: Cross-sectional study. Setting: Large, tertiary care, university teaching hospital. Patients or participants: Thirty-six patients with COPD requiring LTOT (mean age, 63.5 years; mean FEV 1, 32.1% of predicted; PaO 2, 50.2 mm Hg) and 33 control subjects with COPD but no severe hypoxemia (mean age, 63.1 years; FEV 1, 35.7%; PaO 2, 66.5 mm Hg). Interventions: Patients underwent pulmonary function testing to assess physiologic function and the degree of respiratory impairment. A baseline dyspnea index (BDI) was used to determine levels of dyspnea, and a 6-min walk test was performed to evaluate physical performance and exercise capacity. The St. George Respiratory Questionnaire (SGRQ) and the Medical Outcomes Study Short-Form 36-item questionnaire (SF-36) were used to assess health status and HRQL. Measurements and results: The scores on the SGRQ and SF-36 indicated severe impairment. Patients receiving LTOT showed a trend toward worse scores on most dimensions of the SGRQ and SF-36, but differences between groups were only statistically significant for the physical functioning and social functioning dimensions of the SF-36. Dyspnea, as measured by the BDI, significantly correlated with all questionnaire domains except the SF-36 pain index. Conclusions: The HRQL of these low-income patients with COPD was markedly impaired, with more pronounced impairment in those receiving LTOT. The severity of dyspnea was a significant predictor of various components of quality of life in these patients. (CHEST 2003; 123: ) Key words: COPD; long-term oxygen therapy; quality of life Abbreviations: ADL activities of daily living; BDI baseline dyspnea index; HRQL health-related quality of life; LTOT long-term oxygen therapy; Pemax maximum expiratory pressure; Pimax maximum inspiratory pressure; QoL quality of life; SF-36 Medical Outcomes Study Short-Form 36-item questionnaire; SGRQ St. George Respiratory Questionnaire Treatment of COPD is largely palliative, directed toward symptom control and the reduction of acute exacerbations. The impact of the disease and its treatment on the health-related quality of life (HRQL) of these patients has been studied in recent years. Compared with the general population, patients with severe COPD have significant impairment in their quality of life (QoL). 1 5 QoL is lower in patients with severe COPD, but even those with *From the Division of Respiratory Diseases, Heart Institute (InCor)/Hospital das Clínicas and the Division of General Internal Medicine, University of Sao Paulo School of Medicine, Brazil. Manuscript received January 2, 2002; revision accepted July 16, Correspondence to: Alberto Cukier, MD, FCCP, Al. Jau, 263 apto 101, , Sao Paulo, Brazil; pnealberto@incor.usp.br mild disease have HRQL levels lower than those seen in the general population. 2 This impairment manifests not only as a loss of physical mobility, but also as emotional and sleep disturbances. 6 Long-term oxygen therapy (LTOT) became a standard treatment for patients with COPD and severe hypoxemia, mainly due to reports of increased life expectancy in such patients. 7 9 Several studies have shown the deterioration in the QoL of patients with COPD and hypoxemia. 3,4,10,11 However, few studies have addressed the impact of LTOT on QoL, and available results are conflicting. Okubadejo et al 12 detected no change in the QoL of patients with COPD after 6 months of LTOT. Other investigators have reported similar results. 13,14 Conversely, there are some reports of improved QoL after LTOT. 7, Clinical Investigations

2 The aim of this study was to use the Portuguese (Brazil) version of the St. George Respiratory Questionnaire (SGRQ) and the Medical Outcomes Study Short-Form 36-item questionnaire (SF-36) to evaluate HRQL in a low-income population of patients with hypoxemia and COPD receiving LTOT. These results were compared with those from patients with a similar degree of airflow limitation but less severe hypoxemia not requiring LTOT, who served as a control group. Setting Materials and Methods This study was conducted in a large, tertiary care, university teaching hospital in Sao Paulo, Brazil. The patients assisted at our institution belong to the least-favored segment of the population. The average monthly family income of the patients receiving LTOT was approximately $350. None of the patients included in this study would have been able to afford LTOT. Patient Population Study subjects included patients with COPD and severe hypoxemia from an outpatient chest clinic at our institution, who had been receiving LTOT from cylinders for years (mean SD) [LTOT group]. All patients in the LTOT group had a diagnosis of COPD, a FEV L, and a Pao 2 55 mm Hg when stable, or a Pao 2 60 mm Hg with evidence of cor pulmonale or polycythemia. Exclusion criteria were radiologic evidence of any significant abnormality not attributable to COPD, an inability to understand or complete QoL questionnaires, and an inability to perform the 6-min walking test. Of 91 adult patients receiving LTOT, 48 patients had a diagnosis of COPD. Of these, seven patients died during the recruitment period, three patients were hospitalized, and two patients were excluded from the study due to low adherence to the LTOT regimen (assessed by cylinder refill). The remaining 36 patients with COPD receiving LTOT were enrolled in the study (LTOT group). Their mean age was years, and 21 patients were male. Thirty-three patients (25 men) with COPD but less severe hypoxemia comprised the control group. The mean age of the control subjects was years, and none of these patients were receiving LTOT. Assessment Procedures Parameters assessed included physiologic function and functional respiratory impairment (spirometry, arterial blood gas analysis), level of dyspnea (baseline dyspnea index [BDI]), physical performance (6-min walking distance), and HRQL (QoL questionnaires). The validated Portuguese (Brazilian) versions of the SGRQ 16 and the SF were used to measure QoL. All assessments in both groups were made by the same person who administered the questionnaires because some of the patients were illiterate. The disease-specific SGRQ contains 76 items that produce three component scores related to symptoms, activity, and impact, as well as a total score. The symptoms score reflects the frequency and severity of respiratory manifestations, the activity score relates to activities that cause or are limited by breathlessness, and the impact score reflects aspects of social functioning and psychosocial disturbances that result from respiratory diseases. Scores on the SGRQ range from 0 (no disturbance of HRQL) to 100. The SF-36 is a general questionnaire that measures HRQL across eight domains: physical functioning, role physical, role emotional (role limitations due to emotional problems), social functioning, general health perceptions, mental health, bodily pain, and vitality. Higher scores on the SF-36 (range 0 to 100) indicate better QoL. Spirometry measurements were obtained with the patient in a clinically stable condition. Results of blood gas analysis performed for diagnostic or therapeutic purposes up to 6 months before study enrollment were obtained from patient medical records. The maximum inspiratory pressure (Pimax) and maximum expiratory pressure (Pemax) were determined from the functional residual capacity. Dyspnea was assessed using the BDI, 18 and the exercise capacity was evaluated by a 6-min walking test conducted in a 25-m hospital corridor. Patients with an oxygen saturation of 90% at rest received supplemental oxygen. The Ethics Committee at our institution approved the study, and all patients provided written informed consent to participate. Statistical Analysis All data are expressed as means (SD). The Student t test 19 was used to compare measures of physiologic function, dyspnea, and physical performance between groups, whereas the Wilcoxon rank sum test 19 was considered more suitable for comparing QoL variables. Pearson correlations 19 were used to assess the strength of linear relationships between variables. Results were considered statistically significant at p Results Sixty-nine patients with COPD completed the study. Table 1 summarizes measures of physiologic function, dyspnea, and physical performance (exercise capacity). The groups were matched according to flow limitation, as measured by FEV 1. Compared to the control group, the LTOT group had a lower FVC (p 0.001), more severe hypoxemia (p 0.001), Table 1 Measures of Physiologic Function, Dyspnea, and Physical Performance* Variables LTOT (n 36) Control (n 33) p Value Physiologic function FVC, % predicted 54.8 (20.5) 71.1 (19.6) FEV 1, % predicted 32.1 (14.4) 35.7 (13.9) NS Pao 2, mm Hg 50.2 (9.2) 66.5 (11.4) Paco 2, mm Hg 50.3 (8.7) 40.8 (6.4) Pimax, cm H 2 O 40.2 (19.2) 48.0 (20.1) NS Pemax, cm H 2 O 82.6 (29.3) 98.0 (25.3) BDI Focal score (0 12) 3.8 (2.6) 5.3 (3.1) Physical performance 6-min walking test, m 283 (143) 377 (116) *Values are presented as mean (SD). NS not significant. CHEST / 123 / 1/ JANUARY,

3 and hypercapnia (p 0.001). The LTOT patients also had significantly worse dyspnea and exercise capacity scores. The Pimax was similar in both groups, but the LTOT group had a significantly lower Pemax. Tables 2, 3 list the mean scores for the SF-36 and SGRQ in both groups. Overall, patients receiving LTOT seemed to have a worse HRQL, but the differences between groups were only statistically significant for two of the SF-36 dimensions, physical functioning and social functioning. Table 4 shows correlations between the component scores on the SF-36 and SGRQ and physiologic variables, dyspnea ratings, and exercise capacity. Dyspnea, as measured by the BDI, significantly correlated with all domains except the SF-36 pain index. Discussion Our results indicate that patients with moderateto-severe COPD have a marked impairment in HRQL. Patients receiving LTOT appeared to have worse scores on almost all dimensions of the SF-36 and SGRQ questionnaires, although differences were statistically significant only for physical functioning and social functioning dimensions of the SF-36. There are no prior reports about the performance of the Portuguese version of the SGRQ and little data regarding the Portuguese version of the SF since these questionnaires were validated in Brazil. Therefore, it is important to compare our results from these questionnaires with previously published data. Several authors have examined the profile of the effect of COPD on the SGRQ. 2,12,21 24 In a Spanish cohort, Ferrer et al 2 showed that the effect of COPD on QoL increased as disease severity increased. Mean SGRQ total scores for patients with American Thoracic Society 25 stage I, II, and III disease were 34, 42, and 55, respectively. On all Components Table 2 SF-36 Scores* LTOT (n 36) Control (n 33) p Value Physical functioning 16.9 (23.3) 30.3 (29.6) Role physical 9.7 (26.9) 24.2 (38.7) NS Role emotional 51.7 (44.6) 66.5 (40.8) NS Social functioning 29.1 (35.1) 55.2 (40.6) Mental health index 57.9 (25.8) 64.7 (27.3) NS Pain index 56.9 (32.4) 68.1 (28.9) NS Vitality 45.5 (26.6) 57.6 (26.6) NS General health perceptions 54.4 (21.7) 63.7 (27.7) NS *Values are presented as mean (SD). See Table 1 for expansion of abbreviation. Components Table 3 SGRQ Scores* LTOT (n 36) Control (n 33) p Value Symptoms 60.9 (21.4) 66.6 (21.4) NS Activities 76.2 (22.4) 71.8 (23.3) NS Impact 57.1 (22.1) 53.6 (22.8) NS Total 63.5 (20.4) 61.2 (20.2) NS *Values are presented as mean (SD). See Table 1 for expansion of abbreviation. sections of the SGRQ, scores were moderately to strongly associated with FEV 1 categories. Dyspnea also increased in a statistically significant manner as FEV 1 worsened. In contrast, Hajiro et al 23 did not find a relationship between American Thoracic Society disease stage and SGRQ scores, because of a lack of difference between scores of patients with stage II and III disease. However, the same group 24 recently reported a relationship between disease stage based on British Thoracic Society guidelines and SGRQ scores. In both studies, a significant association was detected between the level of dyspnea and SGRQ scores. Ketelaars et al 26 reported the results of 126 Dutch patients with COPD (mean age, 65 years; mean percentage of predicted FEV 1, 39%) whose mean SGRQ component scores were 64 for symptoms, 68 for activity, and 46 for impact. Significant correlations were found between the FEV 1, performance on the 12-min walking test, coping strategies, and the HRQL components of activity and impact. Few studies have evaluated COPD patients with the SF-36 questionnaire. 23,27,28 Crockett et al 28 compared the results of the Nottingham Health Profile and the SF-36 questionnaires obtained at baseline in 60 Australian patients with severe COPD being assessed for LTOT. They showed serious impairment in the QoL of these patients due to large decrements in physical health and smaller decrements in emotional and mental health. Mahler and Mackowiak 27 administered the SF-36 questionnaire to 50 patients with COPD (mean FEV 1, 48% predicted) and showed that dyspnea ratings were significant predictors of seven of the components of the SF-36. The same tendency was recognized in the study by Hajiro et al, 23 the sole previous study to have used both the SGRQ and SF-36 to assess COPD patients. Hajiro and colleagues 23 found that scores on all scales of the SF-36 except for mental health and bodily pain were significantly lower in patients with moderate-to-severe dyspnea than in patients with mild dyspnea. The importance of dyspnea as the major limiting factor of COPD was 138 Clinical Investigations

4 Table 4 Correlations Between HRQL Component Scores and Measures of Physiologic Function, Dyspnea, and Physical Performance in 69 Patients Components FEV 1 Pao 2 Pimax BDI 6-min Walking Test SGRQ Symptoms * Activities Impact Total * SF-36 Physical functioning * Role physical * Role emotional * Social functioning * Mental health index Pain index Vitality General health perceptions 0.25* *p p p confirmed by our documentation of a relationship between BDI scores and scores on all domains of the SGRQ and seven components of the SF-36. Studies that have used different assessment tools have produced similar results. 1,3,4 Overall, these studies suggest that while the functional status and health perceptions of patients with COPD may not be related solely to observable pulmonary impairment, their HRQL appears to be related to the degree of pulmonary disease. Moreover, most results emphasize the importance that patients place on dyspnea, the leading symptom of this disease. Our results in moderately to severely impaired patients are in accordance with these data. Our patients mean scores on individual SGRQ and SF-36 components were comparable to those reported for similarly compromised patients 2,12,21,22,26,28 and worse than those obtained in patients with milder disease. 2,27 The only exceptions applied to comparisons with scores obtained in two Japanese studies. 23,24 Hajiro and colleagues 23 reported better scores for both the SGRQ and SF-36 in their COPD patients, suggesting some cultural differences in responding to the questions between Oriental and occidental populations. The influence of hypoxemia and LTOT on the QoL of patients with COPD has been addressed in several reports. In the Nocturnal Oxygen Therapy Trial, hypoxemic patients demonstrated significant impairment in most domains of a generic questionnaire compared to a control group. A relationship was found between neuropsychological impairment and QoL scores. However, no relationship was found between Pao 2 and QoL scores, 4,11 in contrast to findings in other studies. 2,21,29 31 Okubadejo et al 12,21,22 reported different aspects related to hypoxemic patients with COPD in three sequential articles. In their first study, 21 they used both general and disease-specific questionnaires to investigate the relationship between blood gas data and QoL in a group of 41 patients with severe COPD and a mean Pao 2 of 56 mm Hg. They observed a relationship between hypoxemia and SGRQ scores, in that the Pao 2 was associated with higher, thus, worse QoL scores. The SGRQ symptoms and impact scores correlated with levels of hypoxemia, but the activity score did not. In the second study, Okubadejo et al 12 evaluated a group of patients who were selected to start LTOT and another group of COPD patients with less severe hypoxemia. All patients were followed up for 6 months. Compared with the control group, patients receiving LTOT had a nonsignificant improvement in their QoL. The investigators concluded that LTOT does not have a major beneficial effect on QoL, but that it also does not reduce health status. The aim of their third study 22 was to assess relationships between activities of daily living (ADL), QoL, mood state, and airway obstruction in patients with severe hypoxemic COPD using LTOT and in patients with severe COPD who did not meet the criteria for LTOT. This study showed that independence in ADL is related to the degree of airflow limitation, health status, and mood. The patients using LTOT were significantly less independent in their ADL than the control group. Our patients using LTOT had worse scores on almost all dimensions of the SF-36 and SGRQ questionnaires, although differences were statistically significant only for physical functioning and social functioning dimensions of the SF-36. We CHEST / 123 / 1/ JANUARY,

5 observed weak relationships between a few HRQL domains and FEV 1 or Pao 2. Compared to the control group (matched in terms of FEV 1 ), the hypoxemic patients were more disabled and symptomatic, as suggested by worse BDI scores and shorter distances achieved on the 6-min walking test. A possible implication of these results is that LTOT does not improve dyspnea. However, we are unable to draw such a conclusion due to the crosssectional design of this study. Previously published studies have addressed the influence of the administration of oxygen on the perception of dyspnea, with conflicting results. Besides reports suggesting that hypoxemic and breathless patients benefit symptomatically from supplemental oxygen, 32,33 there are other studies that have not confirmed such an effect. 34,35 Still other authors have concluded that oxygen prescribed for symptomatic relief of dyspnea may benefit only select patients with COPD. 36,37 We matched the groups in our study according to flow limitation, as measured by the FEV 1. However, compared to the control group, the LTOT group had a significantly lower FVC. One possible explanation for this finding is that a concomitant restrictive lung disease might have been present in the LTOT group. Although a definitive conclusion is not possible, as we did not measure lung volumes, this possibility seems unlikely based on clinical and radiologic data. A more probable explanation, which could also explain the significantly worse dyspnea and exercise capacity scores, is that the lungs of the LTOT patients were hyperinflated. LTOT also can influence emotional states, which are determinants of QoL. As we did not use specific measures of neuropsychological impairment, we cannot address how our low-income population coped with the situation. Published data suggest that patients tend not to accept oxygen therapy for several reasons, including the restricted mobility and noise disturbance imposed by the oxygen delivery equipment, 10 the negative expectation linked to the use of a device for extended periods to maintain health, 22 and leisure or work limitations as patients become housebound or nearly housebound. 10,11,22 In conclusion, patients with COPD have important HRQL impairment that is multifactorial in origin. As the disease progresses, the limitation worsens and very few interventions influence this evolution. After hospitalization, oxygen therapy represents the largest share of the total direct medical cost of respiratory care for patients with COPD receiving LTOT. 38 In Brazil, LTOT is mainly carried out by institutions reimbursed by the government, at a monthly cost per patient of approximately $150. The worldwide debate over the cost-benefit and affordability of LTOT 38,39 is even more pertinent in economic emergent countries. Costbenefit analysis needs to establish which interventions optimize the health of the system as a whole, so that the scarce resources devoted to health care may be optimally distributed. Although the best methods for judging the economic value of health-care interventions are controversial, 40 determining the influence of any intervention via measures which go beyond clinical and mortality end points will be increasingly mandatory. Studies comparing LTOT patients with various incomes and in different settings could help clarify this issue. ACKNOWLEDGMENT: The authors thank Carine Favalli Redigolo for performing statistical analysis. References 1 Alonso J, Antó JM, Gonzáles M, et al. Measurement of general health status of non-oxygen-dependent chronic obstructive pulmonary disease patients. Med Care 1992; 30: Ferrer M, Alonso J, Morera J, et al. Chronic obstructive pulmonary disease stage and health-related quality of life: The Quality of Life of Chronic Obstructive Pulmonary Disease Study. Ann Intern Med 1997; 127: Monsó E, Fiz JM, Izquierdo J, et al. Quality of life in severe chronic obstructive pulmonary disease: correlation with lung and muscle function. Respir Med 1998; 92: Prigatano GP, Wright EC, Levin D. Quality of life and its predictors in patients with chronic obstructive pulmonary disease. Arch Intern Med 1984; 144: Jones PW, Quirk FH, Baveystock CM, et al. A self-complete measure of health status for chronic airflow limitation: the St. George s Respiratory Questionnaire. Am Rev Respir Dis 1992; 145: Jones PW. Quality of life measurement for patients with diseases of the airways. Thorax 1991; 46: Continuous or nocturnal oxygen therapy in hypoxemic chronic obstructive lung disease: a clinical trial; Nocturnal Oxygen Therapy Trial. Ann Intern Med 1980; 93: MRC Working Party. Long-term domiciliar oxygen therapy in chronic hypoxic cor pulmonale complicating chronic bronchitis and emphysema. Lancet 1981; 1: Weitzenblum E, Sautegeau A, Ehrhart M, et al. Long-term oxygen therapy can reverse the progression of pulmonary hypertension in patients with chronic obstructive pulmonary disease. Am Rev Respir Dis 1985; 131: Wedzicha JA. Effects of long-term oxygen therapy on neuropsychiatric function and quality of life. Respir Care 2000; 45: McSweeny J, Grant I, Heaton RK, et al. Life quality of patients with chronic obstructive pulmonary disease. Arch Intern Med 1982; 142: Okubadejo AA, Paul EA, Jones PW, et al. Does long-term oxygen therapy affect quality of life in patients with chronic obstructive pulmonary disease and severe hypoxaemia? Eur Respir J 1996; 9: Heaton RK, Grant I, McSweeney AJ, et al. Psychologic effects of continuous and nocturnal oxygen therapy in hypoxemic chronic obstructive pulmonary disease. Arch Intern Med 1983; 143: Lahdensuo A, Ojanen M, Ahonen A, et al. Psychosocial effects of continuous oxygen therapy in hypoxemic chronic 140 Clinical Investigations

6 obstructive pulmonary disease patients. Eur Respir J 1989; 2: Dilworth JP, Higgs CMB, Jones PA, et al. Acceptability of oxygen concentrators: the patient s view. Br J Gen Pract 1990; 40: Sousa TC, Jardim JR, Jones P. Validação do Questionário do Hospital Saint. George na doença respiratória (SGRQ) em pacientes portadores de doença pulmonar obstrutiva crônica no Brasil. J Pneumol 2000; 26: Ciconelli RM, Ferraz MB, Santos W, et al. Tradução para a Língua Portuguesa e validação do Questionário genérico da avaliação de qualidade de vida SF-36 (Brasil SF-36). Rev Bras Reumatol 1999; 39: 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: Rosner B. Fundamentals of biostatistics. Panta, CO: Wads Worth, 1995; Neto JFR, Ferraz MB, Cendoroglo S, et al. Quality of life at the initiation of maintenance dialysis treatment: a comparison between the SF-36 and KDQ questionnaires. Qual Life Res 2000; 9: Okubadejo AA, Jones PW, Wedzicha JA. Quality of life in patients with chronic obstructive pulmonary disease and severe hypoxaemia. Thorax 1996; 51: Okubadejo AA, O Shea L, Jones PW, et al. Home assessment of activities of daily living in patients with severe chronic obstructive pulmonary disease on long-term oxygen therapy. Eur Respir J 1997; 10: Hajiro T, Nishimura K, Tsukino M, et al. A comparison of the level of dyspnea vs disease severity in indicating the healthrelated quality of life of patients with COPD. Chest 1999; 116: Hajiro T, Nishimura K, Tsukino M, et al. Stages of disease severity and factors that affect the health status of patients with chronic obstructive pulmonary disease. Respir Med 2000; 94: Standards for the diagnosis and care of patients with chronic obstructive pulmonary disease. American Thoracic Society. Am J Respir Crit Care Med 1995; 152 (5 pt 2):S77 S Ketelaars CAJ, Schlösser MAG, Mostert R, et al. Determinants of health-related quality of life in patients with chronic obstructive pulmonary disease. Thorax 1996; 51: Mahler DA, Mackowiak JI. Evaluation of the Short-Form 36-item questionnaire to measure health-related quality of life in patients with COPD. Chest 1995; 107: Crockett AJ, Cranston JM, Moss JR, et al. The MOS SF-36 health survey questionnaire in severe chronic airflow limitation: comparison with the Nottingham Health Profile. Qual Life Res 1996; 5: Stavem K, Erikssen J, Boe J. Health-related quality of life is associated with arterial Po 2 in chronic obstructive pulmonary disease. Respir Med 2000; 94: Grant I, Prigatano GP, Heaton RK, et al. Progressive neuropsychologic impairment and hypoxemia: relationship in chronic obstructive pulmonary disease. Arch Gen Psychiatry 1987; 44: Janssens JP, Rochat T, Frey JG, et al. Health-related quality of life in patients under long-term oxygen therapy: a homebased descriptive study. Respir Med 1997; 91: Swinburn CR, Mould H, Stone TN, et al. Symptomatic benefit of supplemental oxygen in hypoxemic patients with chronic lung disease. Am Rev Respir Dis 1991; 143: Woodcock AA, Gross ER, Geddes DM. Oxygen relieves breathlessness in pink puffers. Lancet 1981; 1: McKeon JL, Murree-Allen K, Saunders NA. Effects of breathing supplemental oxygen before progressive exercise in patients with chronic obstructive lung disease. Thorax 1988; 43: Rhind CB, Prince KL, Scott W, et al. Symptomatic oxygen therapy in hypoxic chronic bronchitis [abstract]. Thorax 1986; 41: Evans TW, Waterhouse JC, Carter A, et al. Short burst oxygen treatment for breathlessness in chronic obstructive airways disease. Thorax 1986; 41: Killen JWW, Corris PA. A pragmatic assessment of the placement of oxygen when given for exercise induced dyspnoea. Thorax 2000; 55: Fleury NP, Lanoe JL, Fleury B, et al. The cost of treating COPD patients with long-term oxygen therapy in a French population. Chest 1996; 110: Dunne PJ. The demographics and economics of long-term oxygen therapy. Respir Care 2000; 45: Torgerson D, Raftery J. Measuring outcomes in economic evaluations. BMJ 1999; 318: CHEST / 123 / 1/ JANUARY,