The long-term survival of patients with chronic. Functional Status and Survival Following Pulmonary Rehabilitation*

Transcription

1 Functional Status and Survival Following Pulmonary Rehabilitation* Janet B. Bowen, BA, CRTT; John J. Votto, DO, FCCP; Roger S. Thrall, PhD; Margaret Campbell Haggerty, APRN; Rebecca Stockdale-Woolley, APRN; Tapas Bandyopadhyay, MD; and Richard L. ZuWallack, MD Background: Functional exercise capacity has been shown to be a strong predictor of survival following pulmonary rehabilitation. This study evaluated whether questionnaire-rated functional status is also predictive of survival. Patients and methods: Following pulmonary rehabilitation, patients with advanced chronic lung disease were evaluated for survival, 6-min walk distance, and questionnaire-rated functional status. The latter was measured using the pulmonary functional status scale, which has subscores of functional activities, psychological status, and dyspnea. Information on survival was available on 149 patients. Results: The mean age was 69 years, and 45% of patients were male. Eighty-nine percent had a diagnosis of COPD, and their FEV 1 was 37 18% of predicted. Ninety-one (61%) were married. The 3-year survival for the group was 85%. Age, gender, body mass index, and primary diagnosis were not related to survival. Variables strongly associated with increased survival following pulmonary rehabilitation included a higher postrehabilitation Functional Activities score, a longer postrehabilitation 6-min walk distance, and being married (vs widowed, single, or divorced). Disease severity variables associated with survival included an initial referral to outpatient pulmonary rehabilitation, no supplemental oxygen requirement, and a higher percentpredicted FEV 1. Conclusion: Indicators of functional status are strong predictors of survival in patients with advanced lung disease. (CHEST 2000; 118: ) Key words: pulmonary; questionnaire; rehabilitation; status; survival Abbreviations: BMI body mass index; PFSS pulmonary functional status scale The long-term survival of patients with chronic respiratory disease who are referred for pulmonary rehabilitation is generally poor. For example, the 6-year survival of 119 patients with COPD in a randomized trial of outpatient pulmonary rehabilitation was 61%. 1 In another study of individuals given pulmonary rehabilitation, the 3-year survival following outpatient treatment was 80%. 2 The increased mortality reflects the severity of the underlying respiratory disease and the high likelihood of significant comorbidity in this patient population. *From Hospital for Special Care (Ms. Bowen and Drs. Votto and Thrall), New Britain, CT; Norwalk Hospital (Ms. Haggerty), Norwalk, CT; St. Raphael Hospital (Ms. Stockdale-Woolley), New Haven, CT; University of Connecticut Health Center (Dr. Bandyopadhyay), Farmington, CT; and St. Francis Hospital & Medical Center (Dr. ZuWallack), Hartford, CT This work was supported by funds from each author s respective institution. Manuscript received November 15, 1999; revision accepted April 11, Correspondence to: Janet Bowen, BA, CRTT, Department of Research, Hospital for Special Care, 2150 Corbin Ave, New Britain, CT 06053; bowenj@hfsc.org Life expectancy in patients with COPD, who make up the majority of referrals to pulmonary rehabilitation, has been shown to be related to disease severity factors, such as the severity of airflow obstruction, 3,4 arterial hypoxemia, 5 and the degree of secondary pulmonary hypertension. 6 More recently, Gerardi and colleagues 2 demonstrated that an increased postrehabilitation 12-min walk test distance was a strong predictor of survival in individuals with predominantly moderate-to-severe COPD. This measure of functional exercise capacity was a better predictor of survival status than the Pao 2, Paco 2, FEV 1, and nutritional status, and was even predictive of nonrespiratory deaths. The timed walk distance performance correlates moderately with pulmonary function, 7 and is probably strongly influenced by nonrespiratory factors, such as cardiovascular and peripheral muscular conditioning and body composition. Since the latter comorbid states also probably influence survival, the importance of the walk test as a predictor of survival is understandable. The present study was undertaken to determine CHEST / 118 / 3/ SEPTEMBER,

2 whether another functional status indicator, questionnaire-measured functional status, is also a predictor of survival outcome in patients completing pulmonary rehabilitation. Patients Materials and Methods Survival following pulmonary rehabilitation was studied in patients with symptomatic lung disease who were referred to a participating Connecticut Pulmonary Rehabilitation Consortium program from 1993 to Ten programs in the consortium gave outpatient pulmonary rehabilitation; 1 program also gave inpatient pulmonary rehabilitation. One hundred sixty-four patients had been included in a short-term study evaluating the effect of pulmonary rehabilitation on questionnaire-measured functional status. 8 As part of this study, baseline demographic and severity data were recorded, and the 6-min walk test and the pulmonary functional status scale (PFSS) were measured before and shortly after pulmonary rehabilitation. Baseline information included age, gender, body mass index (BMI), medications, supplemental oxygen requirement, and marital status. Patients were considered not married if they were widowed, divorced, single, or separated from their spouse. Although it was not required for the study, an FEV 1 measurement done as part of the rehabilitation workup or within the preceding year was also recorded, if available. An arterial blood gas determination was not part of the study. The second of two 6-min walk attempts performed on the same day was recorded. For this test, the patient was instructed to walk as far as possible up and down a corridor or around the perimeter of a large room in 6 min. The PFSS (Version 1) 9 was used to assess functional status. This 56-item, self-complete questionnaire that takes about 15 min to complete has subscores of functional activities, dyspnea, and psychological status. The 36-item functional activities section, which was used in this study, has six domains: self-care (3 questions), mobility (4 questions), household tasks (3 questions), grocery shopping and meal preparation (5 questions), daily activities (14 questions), and relationships (7 questions). Although the reliability of this questionnaire has not been extensively tested in pulmonary rehabilitation patients, its functional activities score has been demonstrated to correlate very strongly (r 0.76) with the 6-min walk distance. 8 In 1998, approximately 4 years following the conclusion of pulmonary rehabilitation, the staff at each center assessed the survival status of their patients who were included in the original study. For patients whose survival status was unknown to the staff, this assessment included reviewing hospital and medical office records, contacting primary-care physicians, and contacting the patient or family. For patients who died, the cause of death was categorized into either primarily respiratory or nonrespiratory. The latter included cardiac, cancer, renal, and neurologic categories. The period of time from pulmonary rehabilitation to assessment or death was recorded for the analysis. Since this survival analysis did not alter patient care and was a retrospective study gathering data, Institutional Review Board approval was not required, and patient consent was not necessary. Statistical Analysis Continuous variables are expressed as means SDs. Univariate and multivariate regression analyses of survival data were performed using a Cox proportional hazards ratio model with right-censored survival data. 10 When necessary, categorical explanatory variables (sex, oxygen requirement, married/unmarried status) were coded as 1 or 0 for the analysis. Estimated percent change in the hazard for each 1-U decrease in the PFSS functional activities subscore was determined by subtracting 1.0 from the risk ratio and multiplying by Kaplan-Meier survival function estimate curves were created for descriptive purposes using SAS Lifetest. 12 Pearson correlation coefficients were used to compare PFSS scores and 6-min walk distances. Paired t tests were used to determine the significance of prerehabilitation to postrehabilitation changes in PFSS scores and 6-min walk distances. A p 0.05 was considered significant. Patients Results Of the 164 patients originally evaluated in the Connecticut Pulmonary Rehabilitation Consortium study from 1993 to 1994, follow-up information on survival status was obtainable in 149 patients (91%). Primary respiratory diagnoses on referral to pulmonary rehabilitation included COPD (133 patients, 89%), chronic asthma (12 patients), chest-wall disease (3 patients), and pulmonary fibrosis (1 patient). One hundred thirteen patients (76%) were given outpatient pulmonary rehabilitation, and 36 patients (24%) were given inpatient therapy. Characteristics at entry into pulmonary rehabilitation for outpatient and inpatient rehabilitation patients and the entire group are given in Table 1. As indicated, patients referred for outpatient rehabilitation had less airway obstruction, were less likely to need supplemental oxygen therapy, had longer 6-min walk distances, and had better questionnairerated functional status than patients referred for inpatient therapy. Ninety-one patients were married, and 58 were not married. Of the latter, 35 were widowed, 12 were divorced, and 11 were single. Married and unmarried patients were not different with respect to age, FEV 1, oxygen requirement, and postrehabilitation functional activities subscore. Married patients, however, were more often men (58% vs 24%, p 0.001), and had better performance on the postrehabilitation 6-min walk distance (1, feet vs feet, p 0.02) than unmarried patients. Functional status indicators, which were evaluated before and immediately after pulmonary rehabilitation, are given in Table 2. The 6-min walk distance and all three PFSS subscores improved with rehabilitation (all, p ). There was a highly significant correlation between the total group postrehabilitation 6-min walk distance and the functional activities subscore (r 0.66, p ). Survival Following Pulmonary Rehabilitation A Kaplan-Meier survival curve for the group is given in Figure 1. Forty patients (27%) died over a 698 Clinical Investigations

3 Table 1 Patient Characteristics* Variables Outpatient (n 113) Inpatient (n 36) Total Group (n 149) Age, yr Male patients, % Using supplemental oxygen, % BMI, kg/m FEV 1, percent predicted Postrehabilitation 6-min walk distance, feet Prerehabilitation PFSS functional activities subscore, U *Data are presented as mean SD unless otherwise indicated. Different from outpatient mean, p Different from outpatient mean, p follow-up period of months. Survival was 95% at 1 year after rehabilitation, 92% at 2 years, 85% at 3 years, and 73% at 4 years following pulmonary rehabilitation. The cause of death was primarily due to the primary respiratory disease in 19 patients and to nonrespiratory disease in 21 patients. The latter included deaths due to cardiac disease (12 patients), cancer (8 patients), and neurologic disease (1 patient). Factors Related to Survival Gender, BMI, and primary respiratory diagnosis were not related to survival. Older patients tended to have lower survival rates than did younger patients (p 0.08). Table 3 lists variables related to survival determined from univariate Cox proportional hazards ratio analyses. The two functional status indicators, the 6-min walk distance and the functional activities subscore, were the strongest predictors of postrehabilitation survival: individuals with higher functional activities scores or longer walk distances lived longer. These variables were stronger predictors of survival than either the FEV 1 or the need for supplemental oxygen therapy. For both measures, the postrehabilitation values were stronger predictors than the corresponding prerehabilitation values. The strongest predictor of survival following pulmonary rehabilitation was the PFSS functional activities subscore: individuals with higher levels of functional performance by this questionnaire survived longer. For every 1-U increase in this scale, the risk of dying over this time period decreased by approximately 12% (risk ratio 0.88, p 0.001). The relationship between functional performance and survival is depicted in Figure 2, which plots survival in patients above and below the mean score, which was 17 U. The PFSS psychological status and dyspnea subscores were not related to survival. None of the PFSS subscore prerehabilitation to postrehabilitation changes were related to survival. A longer postrehabilitation 6-min walk distance was also a strong predictor of survival: for every 100-feet U increase in the postrehabilitation 6-min walk test distance, estimated survival increased by approximately 11% (risk ratio 0.89, p 0.002). The relationship between walk test performance and survival is depicted in Figure 3. In this graph, survival is given for three subgroups: those with walk distances 1,000 feet (n 70), those with distances between 500 feet and 1,000 feet (n 48), and those with distances 500 feet (n 31). Patients with greater improvement in walk distances resulting from pulmonary rehabilitation also had longer esti- Table 2 Pulmonary Rehabilitation Outcome Measures (n 149)* Variables Prerehabilitation Postrehabilitation Changes 6-min walk distance, feet PFSS functional activities, U PFSS dyspnea, U PFSS psychological status, U *Data are presented as mean SD. Changes resulting from pulmonary rehabilitation in a group of patients given pulmonary rehabilitation by the Connecticut Pulmonary Rehabilitation Consortium centers between The prerehabilitation to postrehabilitation changes in the 6-min walk distance and all PFSS subscores were significant (p ). CHEST / 118 / 3/ SEPTEMBER,

4 Patients without a supplemental oxygen requirement lived longer than those who required oxygen for all or part of the day (risk ratio 0.40, p 0.005). Survival for patients not requiring oxygen was 97% at 12 months, 95% at 24 months, 92% at 36 months, and 80% at 48 months. In contrast, survival for oxygen-prescribed patients was 93%, 87%, 73%, and 61%, respectively. Marital status was also a significant predictor of survival: married individuals had longer survival than individuals who were not married (either widowed, divorced, single, or living separately; risk ratio 0.43, p 0.009). This relationship is depicted in Figure 4. The survival advantage of married individuals remained even when walk distance was included in the analysis as a cofactor (risk ratio 0.52, p 0.04). Discussion Figure 1. Kaplan-Meier survival function estimate curve for the entire group (n 149). Overall survival following pulmonary rehabilitation was 95% at 1 year, 92% at 2 years, 85% at 3 years, and 73% at 4 years. mated survival, although this was not as strong a predictor of survival as the actual distance at the end of therapy. When respiratory (n 19) and nonrespiratory (n 21) deaths were considered separately in survival analyses, both the PFSS functional activities subscore and the 6-min walk test continued to be significant predictors of mortality (p and p 0.03, respectively). Patients given outpatient pulmonary rehabilitation had better survival than those given inpatient rehabilitation (risk ratio 0.37, p 0.002). However, when either the postrehabilitation 6-min walk test or functional activities subscore was added as a covariate to the analysis, the rehabilitation setting lost significance as a predictor. The purpose of this study was to evaluate the relationship of functional status to subsequent survival following pulmonary rehabilitation. Survival data from 10 centers in the Connecticut Pulmonary Rehabilitation Consortium were analyzed. Although these programs varied in their approaches to rehabilitation, they all standardized their outcome assessment, which consisted of the 6-min walk test and the PFSS. The variability in the pulmonary rehabilitation programs and in their patient populations probably makes the results of this analysis more applicable to postpulmonary rehabilitation in general. Although any individual with chronic pulmonary disease who remained symptomatic despite standard medical therapy was considered a potential candidate for pulmonary rehabilitation, 89% had COPD. The 85% 3-year survival of the Consortium patients is similar to the 80% reported by Gerardi et al 2 in a group of 158 patients who attended one outpatient pulmonary rehabilitation center. The reported 1-year to 4-year survival rates also appear similar to those depicted in a Kaplan-Meier survival curve by Table 3 Factors Related to Increased Survival Following Pulmonary Rehabilitation* Variables 2 p Value Higher PFSS functional activities subscore (postrehabilitation) Longer 6-min walk distance (postrehabilitation), feet Outpatient (vs inpatient) pulmonary rehabilitation Higher PFSS functional activities subscore (prerehabilitation) Not requiring long-term oxygen therapy Being married (vs widowed, single, or divorced) Longer 6-min walk distance (prerehabilitation), feet Greater improvement in the 6-min walk distance Higher FEV 1, % predicted *Variables related to survival using univariate Cox proportional hazards ratio analyses. Wald 2 statistic for testing the null hypothesis that B Clinical Investigations

5 Figure 2. The relationship between the PFSS functional activities subscore and survival following pulmonary rehabilitation. Estimated survival is presented for two subgroups: those above (n 75) and below (n 74) the mean postrehabilitation PFSS functional activities subscores of 17. The relationship between functional status and survival was highly significant (p 0.001). Ries et al, 1 in their controlled trial of outpatient pulmonary rehabilitation. This survival is considerably better than the 51% 2-year survival reported by Connors et al 13 for patients discharged from an acute-care hospital following an exacerbation of COPD. While it is tempting to attribute this difference in survival to the rehabilitation intervention, the large, controlled trial of pulmonary rehabilitation by Ries et al 1 was not able to show a significant effect on survival. Furthermore, the patients admitted to the hospital with exacerbations of COPD in the above study had more advanced respiratory disease and were in clinically less stable condition than those referred to pulmonary rehabilitation. We found that both functional status indicators, the PFSS functional activities subscore and the 6-min walk distance, were strongly related to survival following pulmonary rehabilitation. Thus, patients with less limitation in activities of daily living or greater functional exercise capacity lived longer than did those with poorer performance on these tests. Both measures were, in fact, better predictors of survival than the traditional, direct measures of respiratory disease severity, the FEV 1 and the need for supplemental oxygen. Less than 50% of the deaths following pulmonary rehabilitation were due primarily to respiratory disease, attesting to the importance of serious comorbidity in this population of patients. Despite the varied causes of death, however, questionnaire-rated functional status and the 6-min walk distance still predicted survival. Figure 3. The relationship between the postrehabilitation 6-min walk distance and survival following pulmonary rehabilitation. Estimated survival is presented for three subgroups based on postrehabilitation 6-min walk performance: those with walk distances 1,000 feet (n 70), those with walk distances between 500 feet and 1,000 feet (n 48), and those with walk distances 500 feet (n 31). For every 100-foot increase in exercise performance, estimated survival was increased by approximately 11%. The relationship between postrehabilitation walk distance and survival was highly significant (p 0.002). To our knowledge, the importance of questionnaire-measured functional status as a marker of survival following pulmonary rehabilitation has not been previously reported. The relevance of this measurement of activities of daily living as a predictor of survival is understandable given its strong correlation with the timed walk distance in our study. The timed walk distance has already been demonstrated to be a strong predictor of survival. 2 A reduction in functional status has been demonstrated by Connors and colleagues 13 to predict survival following hospitalization for exacerbation of COPD. In their study, patients with greater interference in activities of daily living had poorer survival. Interestingly, postrehabilitation performance was a better predictor of survival than prerehabilitation performance for both measures of function. The better predictive ability of the postrehabilitation 12-min walk distance on subsequent survival has been noted previously. 2 A possible explanation of this effect is that patients enter pulmonary rehabilitation with varying levels of motivation, pacing skills, peripheral muscle and cardiovascular deconditioning, and functional capacity utilization. 14 Pulmonary rehabilitation may reduce the variability from these potentially confounding variables, leaving behind truer measures of function in the immediate postrehabilitation period. CHEST / 118 / 3/ SEPTEMBER,

6 Figure 4. The relationship between the marital status and survival following pulmonary rehabilitation. Unmarried indicates widowed, divorced, or single. The relationship between marital status and survival was highly significant (p 0.009). Other factors related to longer survival included the referral to outpatient rehabilitation (vs inpatient rehabilitation), no need for supplemental oxygen therapy, being married (vs being widowed, divorced, or single), having a greater improvement in the 6-min walk distance, and a higher percent-predicted FEV 1 at baseline. Despite its prognostic importance in large series of COPD patients, 3 the FEV 1 was the weakest predictor of survival in this study. This is partially explained by the relatively narrow range of FEV 1 values in patients referred to pulmonary rehabilitation, thus limiting this measure as a predictor of survival. However, it also makes it clear that factors outside of the primary disease process strongly influence survival. Patients given inpatient pulmonary rehabilitation had a far worse prognosis than did those in outpatient programs. This is undoubtedly due to selection bias, with the sickest patients given inpatient therapy. Although arterial oxygen tension was not measured as part of the study, a surrogate marker of oxygenation, the supplemental oxygen requirement, followed the functional status variables in strength as a predictor of survival. At 3 years following pulmonary rehabilitation, 92% of the patients without oxygen requirement but only 73% of the patients using oxygen were alive. The latter statistic is in contrast to reported survival of oxygen-requiring patients from a large series in Europe, which was 70%, 50%, and 43% at 1, 2, and 3 years, respectively. 15 The difference in survival is probably explained by the fact that the European patients were followed up after having been admitted to the hospital for respiratory insufficiency and severe hypoxemia (Pao 2 55 mm Hg). A somewhat surprising finding was the relationship between marital status and survival. Patients who were married and had a spouse at home lived longer than those who were widowed, divorced, or separated. Married and not married patients had similar demographic and severity variables, except that the former were more frequently male and had longer walk distances. However, even after controlling for walk distance, marital status remained a significant prognostic factor. The reasons for this favorable effect are unclear, but being married with a spouse at home has been associated with prolonged survival in prostate cancer, 16 and with a decrease in early readmission to the hospital or death in patients with congestive heart failure. 17 The results of our study are in contrast to those of Schols and colleagues 18 who found older age and low BMI to be predictors of mortality in their retrospective analysis of 400 patients with moderate to severe COPD. Since older age tended to predict reduced survival in our patients (p 0.08), this discrepancy in results probably reflects the lower statistical power in our study. Although the patients of Schols and colleagues 18 and ours had similar mean BMI (24.0 kg/m 2 and 24.8 kg/m 2, respectively), this variable was not a predictor of survival in our analysis either when entered as a continuous variable or when dichotomized into underweight, normal, and overweight categories. Although the study by Schols and colleagues 18 had a preponderance of male patients while ours had more female patients, gender was not a significant factor in either study. The previously referenced study by Gerardi and colleagues 2 did show a small albeit weak relationship between BMI and survival. The reasons why the present study was unable to show a significant relationship between nutritional depletion and poorer survival are unknown. In summary, for patients with chronic respiratory disease completing pulmonary rehabilitation, both functional status indicators the PFSS functional activities subscore and the 6-min walk distance were better predictors of survival than more traditional measures of respiratory disease severity. This finding underscores the importance of this type of measurement in individuals with advanced lung disease. ACKNOWLEDGMENT: The authors wish to acknowledge the following individuals and institutions of The Connecticut Pulmonary Rehabilitation Consortium: Renee Clinton, RRT, and Lori Giahnbattista, RRT, Windham Hospital, Windham, CT; Gina Chase, RRT, Manchester Memorial Hospital, Manchester, CT; 702 Clinical Investigations

7 Mary Lou Golebieski, RRT, RPFT, Griffon Hospital, Derby, CT; Patricia A. Palmer, APRN, Danbury Hospital, Danbury, CT; Margarita Rios, APRN, Gaylord Hospital, Wallingford, CT; and Kathryn Sullivan, CRTT, Waterbury Hospital, Waterbury, CT References 1 Ries AL, Kaplan RM, Limberg TM, et al. Effects of pulmonary rehabilitation on physiologic and psychosocial outcomes in patients with chronic obstructive pulmonary disease. Ann Intern Med 1995; 122: Gerardi DA, Lovett L, Benoit-Connors ML, et al. Variables related to increased mortality following out-patient pulmonary rehabilitation. Eur Respir J 1996; 9: Boushey SF, Thompson HK, North LB, et al. Prognosis in chronic obstructive pulmonary disease. Am Rev Respir Dis 1973; 108: Traver GA, Cline MG, Burrows B. Predictors of mortality in chronic obstructive pulmonary disease: a 15-year follow-up study. Am Rev Respir Dis 1979; 119: The Medical Research Council Working Party (Stuart-Harris C, chairman). Long term domiciliary oxygen therapy in chronic hypoxemic cor pulmonale complicating chronic bronchitis and emphysema. Lancet 1981; 1: Weitzenblum E, Hirth C, Ducolone A, et al. Prognostic value of pulmonary artery pressure in chronic obstructive pulmonary disease. Thorax 1981; 36: ZuWallack RL, Patel K, Reardon JZ, et al. Predictors of improvement in the 12-min walking distance following a six-week outpatient pulmonary rehabilitation program. Chest 1991; 99: Haggerty MC, Stockdale-Woolley R, ZuWallack R. Functional status in pulmonary rehabilitation participants. J Cardiopulm Rehabil 1999; 19: Weaver TE, Narsavage GL. Physiological and psychological variables related to functional status in chronic obstructive pulmonary disease. Nurs Res 1992; 41: The PHREG Procedure. In: SAS Technical Report P-229 SAS/STAT Software. Changes and Enhancements manual. Cary, NC: SAS Institute, 1992; Allison PD. Survival analysis using the SAS System: a practical guide. Cary, NC: SAS Institute, 1995; SAS/STAT User s Guide Volume 2. Version 6. 4th ed. Cary, NC: SAS Institute, 1990; Connors AF, Dawson NV, Thomas C, et al. Outcomes following acute exacerbation of severe chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1996; 154: Leidy NK. Functional status and the forward progress of merry-go-rounds: toward a coherent analytical framework. Nurs Res 1994; 43: Dubois P, Jamart J, Machiels J, et al. Prognosis of severely hypoxemic patients receiving long-term oxygen therapy. Chest 1994; 105: Krongrad A, Lai H, Burke MA, et al. Marriage and mortality in prostate cancer. J Urol 1996; 156: Chin MH, Goldman L. Correlates of early hospital readmission or death in patients with congestive heart failure. Am J Cardiol 1997; 79: Schols AMJ, Slangen J, Volovics L, et al. Weight loss is a reversible factor in the prognosis of chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1998; 157: CHEST / 118 / 3/ SEPTEMBER,