Impact of Dyspnea and Physiologic Function on General Health Status in Patients with Chronic Obstructive Pulmonary Disease* Donald A. Mahler, M.D., F.C.C.R;t Kathy Faryniarz, R.N., M.S.; Donald Tomlinson, B.S.; Gene L. Colice, M.D., F.C.C.R;t Arthur G. Robins, M.D.;+ Elaine M. Olmstead, B.A.;§ and Gerald L O~Connor; Ph.D., DSc.1I Study Objective: To examine the relationships among clinical dyspnea ratings, physiologic pulmonary function, and general health status in symptomatic patients with chronic obstructive pulmonary disease (COPO). Design: Observational data collected at a baseline state. Setting: Outpatient pulmonary disease clinics at a university hospital and two VA medical centers. Patients: One hundred ten male patients with COPO with no significant comorbidity were recruited. Measurements and Results: Clinical ratings of dyspnea were measured by the multidimensional baseline dyspnea index (BDI). Pulmonary function tests included forced vital capacity (FVC), forced expiratory volume in 1 s (FEV.), and maximal inspiratory mouth pressure (Plmax). General health status was assessed with the Medical Outcomes Study short-form survey, which consists of 20 questions that cover six health components. The mean age of the patients was 67±8 yr (±SO). The mean value for FVC was 2.84±0.84 L (68± 18 percent of predicted), for FEV. was 1.28 ± 0.59 L (44 ± 17 percent of predicted), and for Plmax was 59.0±25.0 cm HsO. The HDI score and PImax were signmcantly correlated with 6ve of the six components of general health status. Only three of the six components of general health were signi6cantly correlated with FEV. as percent predicted (r. value range, 0.30 to 0.44) and with
FVC as percent predicted (r. value range, 0.25 to 0.33). Statistical comparisons showed that the HOI score had signi6cantly higher correlations than FVC (percent predicted), FEV. (percent predicted), and Plmax values with physical functioning and role functioning. Multiple linear regression analysis showed that the HDI score was the only statistically signi6cant predictor of role functioning, mental health, and health perceptions for general health status, whereas both the HOI score and FEV. (percent predicted) were independent predictors of physical functioning and social functioning. Conclusion: Dyspnea ratings inOuence and predict general health status to a greater extent than do physiologic measurements in symptomatic patients with COPo. A shift in focus from the pathophysiology of disease to assessment and relief of symptoms may provide more meaningful benefits for the individual patient in terms of quality of life. This consideration requires that health-care providers use available measuring tools in clinical practice to quantify symptoms, as well as overall health status. (Chest 1992; 102:395-401)
What are the important outcomes in evaluating and caring for patients with chronic respiratory disease? Although improvement in general health status, or well-being, is the presumed goal of medical care for individual patients, the traditional approach has been to rely on pulmonary function tests (PFrs) to quantify the severity of respiratory disease and to assess response to therapy;1.2 however, dyspnea is the most common complaint for which patients with
COPD seek medical attention. More recent emphasis has focused on relief of symptoms, especially dyspnea, as a major goal of treatment.3-6 Both of these approaches assume that pulmonary function or dyspnea or both are related to and influence general health status in patients with chronic respiratory disease. 7-10 The purpose of this study was to examine the relationships among dyspnea ratings, PFrs, and general health status in symptomatic patients with chronic obstructive pulmonary disease (COPD). The hypothesis tested was that the symptom of dyspnea has a greater impact on reported health status than does respiratory physiology as measured by PFTs.
*From the Departments of Medicine and of Community and Family Medicine, Dartmouth Medical School, Lebanon, NH. t Associate Professor of Medicine. *Adjunct Assistant Professor of Medicine. §Senior Research Assistant in Clinical Medicine. IIAssistant Professor of Medicine and of Community and Family Medicine. Supported by a grant from the American Lung Association of New Hampshire. Manuscript received August 19; revision accepted November 22. Reprint requests: Dr. Mahler, PulfTWnllry and Critical Care Aledicine, One Medical Center Drive, l..£banon, New Hampshire 03756
=
=
BDI baseline dyspnea index; MHMH Mary Hitchcock Memorial Hospital; MOS Medical Outcomes Study; QWB quality of well-being scale; SIP sickness impact pro61e
=
=
=
MATERIALS AND METHODS
Studied Population A group of 110 male patients \\rith COPD (as defined by the American Thoracic Societyll) of varyin~ de~rees of severity and with no si~nificant comorbidity were entered into the study at three CHEST / 102 / 2 / AUGUS'T, 1992
395
hospital outpatient clinics. Criteria for inclusion were as follows: diagnosis of symptomatic COPD; FEV/FVC ratio less than 70 percent; and dyspnea on exertion. Criteria for exclusion included: legally blind or deaf; cancer; uncontrolled diabetes, hypertension, or psychiatric illness; class 2 or greater on New York Heart Association's criteria fi>r chest pain; musculoskeletal or neurologic disease with functional impairment; liver failure or cirrhosis; inability to return for follow-up every 6 mo; any unstable gastrointestinal condition; and current alcohol or other substance abuse problem. Each patient provided signed written consent, and the study's protocol was approved by the institutional review board at each participating institution.
Forced vital capacity (FVC) and forced expiratory volume in 1 s (FEV 1) were measured in the seated position using the M100B testing system (Gould Inc) or a portable spirometer (Spirometries model 2451). Both systems met criteria for standardization and were calibrated by standard methods prior to testing. Direct comparison of the two instruments showed highly reliable values. Measures were selected from the best of five efforts for FVC and FEV1. Predicted normal values were taken from Morris and colleagues. 17 Maximal inspiratory pressure (PImax) was measured at functional residual capacity using a pressure gauge connected to a cylinder and mouthpiece. Hi A minimum of five efforts was performed until a reproducible maximal value was obtained.
Data CoUection
Statistical Analysis
Dyspnea, general health status, and physiologic function were assessed in each patient on the same day during a clinically stable period in the outpatient clinic. This sequence of collecting data was selected so that scores for dyspnea and general health status would be obtained prior to results of physiologic tests. The clinical rating of dyspnea was graded by a nurse-clinician or cardiopulmonary technician using the baseline dyspnea index (BDI), a multidimensional instrument for measuring breathlessness based on three components evoking dyspnea: functional impairment; magnitude of task; and magnitude of effort (appendix A).12 Previous studies have demonstrated the validity and reliability of the BDI in patients with chronic respiratory disease.3.5.6.12-14 A nurse or technician selected from one of five grades of dyspnea based on a brief history for each of the three components of the BDl. This process usually took 3 to 4 min. General health status was assessed with the Medical Outcomes Study's (MOS) short-form survey, which consists of 20 questions that cover six health concepts: physical functioning; role functioning; social functioning; mental health; health perceptions; and pain (appendix B).15 Validity and reliability of the short-fonn general health survey has been established in 11,186 adult English-speaking patients, including those with COPD, who received medical care from a total of 526 health-care providers. 16 For all measures of the six health components, scores were transformed linearly to scales of 0 to 100, with 0 and 100 assigned to the lowest and highest possible scores, respectively.I s
Spearman's rank-order correlation coefficients (r.) were determined to examine the degree ofcorrelation among the clinical rating of dyspnea, values for the six components of the MOS general health survey, and results of physiologic testing (FVC, FEV1, and PI max). In order to compare two correlation coefficients, the values of the correlation coefficients were transformed to provide a normal distribution. 19 Z scores were then calculated from the transformed values, and corresponding two-tailed p values were calculated for continuous variables. A p value of less than 0.05 was (.'onsidered significant. Values are presented as the mean ± SD. Multiple linear regression analysis was performed to evaluate the individual contribution of the BDI score and physiologic parameters (independent variables) as predictors of health status (dependent variable) while controlling for the effects of the other variables, including age, years of education, and enrollment site. 2n The T statistic from the multiple linear regression output \\'as calculated; it represents the t-test of the null hypothesis that there is no relationship between the independent and dependent variables. 2tl RESULTS
Between December 1988 and June 1989, patients were recruited at Mary Hitchcock Memorial Hospital (MHMH) in Hanover, NH (n = 51); VA Medical Center
Table I-Age, BDI, Components ofMOS General Health Survey, and Physiologic Function Testsfor Total Group of 110 Subjects and by Site· Variable Age, yr BDlt Functional impairment Magnitude of task MaJtnitude of effort Focal score MOS health components; Physical functioning Role functioning Social functioning Mental health Health perceptions Pain Physiologic function tests FVC, percent of predicted FEV I' percent of predicted Plmax, cm H 2O
Combined (n = 110)
VA-MAN (n=21)
VA-WRJ (n=38)
MHMH (n=51)
67±8
66±7
67±7
67±8
1.6± 1.1 1.6±0.8 1.6±O.8 4.8±2.4
1.2± 1.2 1.7±0.9 1.5±O.8 4.4±2.7
1.6± 1.2 1.6±0.7 1.6±O.7 4.7±2.4
1.8± 1.1 1.7±0.8 1.6±0.8 5.1 ±2.4
32.7±26.4 36.6±43.6 70.6±32.3 67.6± 19.7 31.2±21.6 38.9±32.9
29.4±27.8 26.2±43.6 65.7±26.9 70.7± 18.2 27.0± 24.1 44.0±35.3
29.2±24.9 30.3±42.4 68.9±35.3 63.6± 19.9 28.6±23.3 40.1 ±34.2
36.8±26.9 45.6±43.5 73.7±32.1 69.3±20.1 34.9± 18.9 35.8±31.3
68.0± 17.8 43.7± 17.3 59.0±25.0
6O.9± 17.2 37.3± 14.8 52.8±22.7
61.1 ± 15.8 41.2± 16.3 58.2±28.7
76.1 ± 16.2 48.2± 18.1 62.2±22.7
*Table data are means±SD. VA-MAN, VA-Manchester, NH; VA-WRJ, VA-White River Junction, Vf; and MHMH, Mary Hitchcock Memorial Hospital. tHigher BDI scores indicate less dyspnea. tHigher MOS functional s(.'Ores indicate better health, except pain, where higher score indicates more pain. 396
Impact of Dyspnea and Physiologic Function in COPO (Mahler et al)
in White River Junction, \IT (n = 38); and the VA Medical Center in Manchester, NH (n = 21). The mean age was 67 ± 8 yr (± SD) (range, 41 to 82 yr), mean height was 174±7 cm (ran~e, 157 to 193 cm), and mean weight was 77 ± 19 kg (range, 43 to 181 kg). Eighty-one percent of the subjects were ex-smokers, and 19 percent were current smokers. Medications currently prescribed for COPD included the following: inhaled or oral (3-adrenergic agonists, 83 percent; theophylline, 70 percent; inhaled corticosteroids, 31 percent; inhaled ipratropium bromide, 28 percent; oral corticosteroids, 19 percent; and inhaled cromolyn, 3 percent. Five percent of patients used intermittent supplemental oxygen. Values for the BDI, health status components, and physiologic function tests for the total group and for subgroups by site are listed in Table 1. These results represent a spectrum from mild to severe COPD based on PFTs. For the 110 patients the mean FVC was 68 ± 18 percent (± SD) of the predicted value, and mean FEV. was 44 ± 17 percent of the predicted value. Levels of dyspnea and general health varied widely in the patients. The BDI scores were significantly correlated with FVC as percent predicted (rs = 0.36; p
FVC as percent predicted (r. value range, 0.25 to 0.33). Statistical comparisons showed that the HOI score had signi6cantly higher correlations than FVC (percent predicted), FEV. (percent predicted), and Plmax values with physical functioning and role functioning. Multiple linear regression analysis showed that the HDI score was the only statistically signi6cant predictor of role functioning, mental health, and health perceptions for general health status, whereas both the HOI score and FEV. (percent predicted) were independent predictors of physical functioning and social functioning. Conclusion: Dyspnea ratings inOuence and predict general health status to a greater extent than do physiologic measurements in symptomatic patients with COPo. A shift in focus from the pathophysiology of disease to assessment and relief of symptoms may provide more meaningful benefits for the individual patient in terms of quality of life. This consideration requires that health-care providers use available measuring tools in clinical practice to quantify symptoms, as well as overall health status. (Chest 1992; 102:395-401)
What are the important outcomes in evaluating and caring for patients with chronic respiratory disease? Although improvement in general health status, or well-being, is the presumed goal of medical care for individual patients, the traditional approach has been to rely on pulmonary function tests (PFrs) to quantify the severity of respiratory disease and to assess response to therapy;1.2 however, dyspnea is the most common complaint for which patients with
COPD seek medical attention. More recent emphasis has focused on relief of symptoms, especially dyspnea, as a major goal of treatment.3-6 Both of these approaches assume that pulmonary function or dyspnea or both are related to and influence general health status in patients with chronic respiratory disease. 7-10 The purpose of this study was to examine the relationships among dyspnea ratings, PFrs, and general health status in symptomatic patients with chronic obstructive pulmonary disease (COPD). The hypothesis tested was that the symptom of dyspnea has a greater impact on reported health status than does respiratory physiology as measured by PFTs.
*From the Departments of Medicine and of Community and Family Medicine, Dartmouth Medical School, Lebanon, NH. t Associate Professor of Medicine. *Adjunct Assistant Professor of Medicine. §Senior Research Assistant in Clinical Medicine. IIAssistant Professor of Medicine and of Community and Family Medicine. Supported by a grant from the American Lung Association of New Hampshire. Manuscript received August 19; revision accepted November 22. Reprint requests: Dr. Mahler, PulfTWnllry and Critical Care Aledicine, One Medical Center Drive, l..£banon, New Hampshire 03756
=
=
BDI baseline dyspnea index; MHMH Mary Hitchcock Memorial Hospital; MOS Medical Outcomes Study; QWB quality of well-being scale; SIP sickness impact pro61e
=
=
=
MATERIALS AND METHODS
Studied Population A group of 110 male patients \\rith COPD (as defined by the American Thoracic Societyll) of varyin~ de~rees of severity and with no si~nificant comorbidity were entered into the study at three CHEST / 102 / 2 / AUGUS'T, 1992
395
hospital outpatient clinics. Criteria for inclusion were as follows: diagnosis of symptomatic COPD; FEV/FVC ratio less than 70 percent; and dyspnea on exertion. Criteria for exclusion included: legally blind or deaf; cancer; uncontrolled diabetes, hypertension, or psychiatric illness; class 2 or greater on New York Heart Association's criteria fi>r chest pain; musculoskeletal or neurologic disease with functional impairment; liver failure or cirrhosis; inability to return for follow-up every 6 mo; any unstable gastrointestinal condition; and current alcohol or other substance abuse problem. Each patient provided signed written consent, and the study's protocol was approved by the institutional review board at each participating institution.
Forced vital capacity (FVC) and forced expiratory volume in 1 s (FEV 1) were measured in the seated position using the M100B testing system (Gould Inc) or a portable spirometer (Spirometries model 2451). Both systems met criteria for standardization and were calibrated by standard methods prior to testing. Direct comparison of the two instruments showed highly reliable values. Measures were selected from the best of five efforts for FVC and FEV1. Predicted normal values were taken from Morris and colleagues. 17 Maximal inspiratory pressure (PImax) was measured at functional residual capacity using a pressure gauge connected to a cylinder and mouthpiece. Hi A minimum of five efforts was performed until a reproducible maximal value was obtained.
Data CoUection
Statistical Analysis
Dyspnea, general health status, and physiologic function were assessed in each patient on the same day during a clinically stable period in the outpatient clinic. This sequence of collecting data was selected so that scores for dyspnea and general health status would be obtained prior to results of physiologic tests. The clinical rating of dyspnea was graded by a nurse-clinician or cardiopulmonary technician using the baseline dyspnea index (BDI), a multidimensional instrument for measuring breathlessness based on three components evoking dyspnea: functional impairment; magnitude of task; and magnitude of effort (appendix A).12 Previous studies have demonstrated the validity and reliability of the BDI in patients with chronic respiratory disease.3.5.6.12-14 A nurse or technician selected from one of five grades of dyspnea based on a brief history for each of the three components of the BDl. This process usually took 3 to 4 min. General health status was assessed with the Medical Outcomes Study's (MOS) short-form survey, which consists of 20 questions that cover six health concepts: physical functioning; role functioning; social functioning; mental health; health perceptions; and pain (appendix B).15 Validity and reliability of the short-fonn general health survey has been established in 11,186 adult English-speaking patients, including those with COPD, who received medical care from a total of 526 health-care providers. 16 For all measures of the six health components, scores were transformed linearly to scales of 0 to 100, with 0 and 100 assigned to the lowest and highest possible scores, respectively.I s
Spearman's rank-order correlation coefficients (r.) were determined to examine the degree ofcorrelation among the clinical rating of dyspnea, values for the six components of the MOS general health survey, and results of physiologic testing (FVC, FEV1, and PI max). In order to compare two correlation coefficients, the values of the correlation coefficients were transformed to provide a normal distribution. 19 Z scores were then calculated from the transformed values, and corresponding two-tailed p values were calculated for continuous variables. A p value of less than 0.05 was (.'onsidered significant. Values are presented as the mean ± SD. Multiple linear regression analysis was performed to evaluate the individual contribution of the BDI score and physiologic parameters (independent variables) as predictors of health status (dependent variable) while controlling for the effects of the other variables, including age, years of education, and enrollment site. 2n The T statistic from the multiple linear regression output \\'as calculated; it represents the t-test of the null hypothesis that there is no relationship between the independent and dependent variables. 2tl RESULTS
Between December 1988 and June 1989, patients were recruited at Mary Hitchcock Memorial Hospital (MHMH) in Hanover, NH (n = 51); VA Medical Center
Table I-Age, BDI, Components ofMOS General Health Survey, and Physiologic Function Testsfor Total Group of 110 Subjects and by Site· Variable Age, yr BDlt Functional impairment Magnitude of task MaJtnitude of effort Focal score MOS health components; Physical functioning Role functioning Social functioning Mental health Health perceptions Pain Physiologic function tests FVC, percent of predicted FEV I' percent of predicted Plmax, cm H 2O
Combined (n = 110)
VA-MAN (n=21)
VA-WRJ (n=38)
MHMH (n=51)
67±8
66±7
67±7
67±8
1.6± 1.1 1.6±0.8 1.6±O.8 4.8±2.4
1.2± 1.2 1.7±0.9 1.5±O.8 4.4±2.7
1.6± 1.2 1.6±0.7 1.6±O.7 4.7±2.4
1.8± 1.1 1.7±0.8 1.6±0.8 5.1 ±2.4
32.7±26.4 36.6±43.6 70.6±32.3 67.6± 19.7 31.2±21.6 38.9±32.9
29.4±27.8 26.2±43.6 65.7±26.9 70.7± 18.2 27.0± 24.1 44.0±35.3
29.2±24.9 30.3±42.4 68.9±35.3 63.6± 19.9 28.6±23.3 40.1 ±34.2
36.8±26.9 45.6±43.5 73.7±32.1 69.3±20.1 34.9± 18.9 35.8±31.3
68.0± 17.8 43.7± 17.3 59.0±25.0
6O.9± 17.2 37.3± 14.8 52.8±22.7
61.1 ± 15.8 41.2± 16.3 58.2±28.7
76.1 ± 16.2 48.2± 18.1 62.2±22.7
*Table data are means±SD. VA-MAN, VA-Manchester, NH; VA-WRJ, VA-White River Junction, Vf; and MHMH, Mary Hitchcock Memorial Hospital. tHigher BDI scores indicate less dyspnea. tHigher MOS functional s(.'Ores indicate better health, except pain, where higher score indicates more pain. 396
Impact of Dyspnea and Physiologic Function in COPO (Mahler et al)
in White River Junction, \IT (n = 38); and the VA Medical Center in Manchester, NH (n = 21). The mean age was 67 ± 8 yr (± SD) (range, 41 to 82 yr), mean height was 174±7 cm (ran~e, 157 to 193 cm), and mean weight was 77 ± 19 kg (range, 43 to 181 kg). Eighty-one percent of the subjects were ex-smokers, and 19 percent were current smokers. Medications currently prescribed for COPD included the following: inhaled or oral (3-adrenergic agonists, 83 percent; theophylline, 70 percent; inhaled corticosteroids, 31 percent; inhaled ipratropium bromide, 28 percent; oral corticosteroids, 19 percent; and inhaled cromolyn, 3 percent. Five percent of patients used intermittent supplemental oxygen. Values for the BDI, health status components, and physiologic function tests for the total group and for subgroups by site are listed in Table 1. These results represent a spectrum from mild to severe COPD based on PFTs. For the 110 patients the mean FVC was 68 ± 18 percent (± SD) of the predicted value, and mean FEV. was 44 ± 17 percent of the predicted value. Levels of dyspnea and general health varied widely in the patients. The BDI scores were significantly correlated with FVC as percent predicted (rs = 0.36; p
