© 2003 American Thoracic Society
Role of Exercise and Lung Function in Predicting Work Status in Cystic FibrosisDepartment of Medicine, University of British Columbia; McDonald Research Laboratories/iCAPTURE Centre, and Adult Cystic Fibrosis Clinic, St. Paul's Hospital, Vancouver, British Columbia, Canada Correspondence and requests for reprints should be addressed to Dr. P. G. Wilcox, M.D., Department of Medicine, Pulmonary Research Laboratory, McDonald Research Wing, Room 292, 1081 Burrard Street, St. Paul's Hospital, Vancouver, BC, V6Z 1Y6 Canada
With larger numbers of adult patients with cystic fibrosis (CF) in the workplace, the issue of disability has arisen increasingly. We examined relationships between measures of pulmonary impairment and work/school capability and then determined whether quantification of aerobic fitness improved predictability of disease-related disability. We studied 73 patients with CF who performed lung function and exercise capacity tests, completed a work/education questionnaire, and were scored for clinical and chest radiographic status. Patients who were characterized as unemployed and in poor health had more severe pulmonary disease according to American Thoracic Society impairment/disability criteria. Subjects were further classified into three groups based on employment or education status over the preceding 12 months. FEV1, maximal oxygen consumption, Schwachman–Kulczycki clinical and Brasfield radiographic scores, and frequency of pulmonary exacerbations over 2 years were associated with disability, but change in FEV1 over 2 years and oxygen saturation at rest or exertion were not. FEV1 and Schwachman–Kulczycki scores were the best independent predictors of impairment/disability; specific thresholds used in other pulmonary diseases were of limited utility. We conclude that after accounting for either current level of FEV1 or Schwachman–Kulczycki scores, no other physiological or clinical measures contribute to predicting limitation in a work or school environment.
Key Words: patients with cystic fibrosis • disability • forced expiratory volume • impairment
Advances in both the diagnosis and treatment of cystic fibrosis (CF) have led to substantial increases in survival of this population. As a result, adult patients commonly outnumber pediatric patients with CF. With more adult patients with CF in the workplace and in light of the progressive nature of the disease, the issue of disability inevitably arises. The medical assessment of job performance or determination of potential work restrictions is particularly challenging. The American Thoracic Society (ATS) publication concerning evaluation of impairment/disability represents one of the guidelines used in determining the presence and degree of impairment caused by respiratory disorders (1). This approach uses resting pulmonary function tests and in some instances measures of exercise capacity. Other national and state guidelines for making impairment ratings also exist and also use pulmonary function and exercise test measures. Impairment has been defined as temporary or permanent change in pulmonary function and/or maximal oxygen consumption (  O2max) (1). Disability is defined as the effect of impairment in an individual's life. Severe impairment is defined as FEV1% pred < 40%, FVC % pred < 50%, FEV1/FVC < 40%, or diffusing capacity of the lung for CO (DLCO) < 40% (1). Exercise testing is recommended if any of the resting pulmonary function measurements are abnormal and above the energy demand threshold. A patient is characterized as severely impaired if O2max < 15 ml/kg per minute, or if the occupational energy demand exceeds 40% of the patient's O2max (1). Ortega and associates have shown that a combination of resting pulmonary function measurements and exercise performance assessment can more accurately determine impairment in patients with chronic obstructive pulmonary disease (2). These thresholds have not been specifically evaluated in patients with CF. Applicability to this patient group is uncertain given the younger age and effects of other manifestations of this disease.
The aim of our study was threefold: first, to determine the proportion of a cohort of adult patients with CF who are limited in their employment/education on the basis of current ATS impairment criteria; second, to investigate the utility of different measurements of resting pulmonary function and exercise capacity used to apportion impairment caused by other respiratory diseases in patients with CF; and finally, to evaluate the energy cost of the work of patients as a fraction of their measured
Sample We studied 73 stable patients with CF attending the Adult Cystic Fibrosis Clinic at St. Paul's Hospital (Vancouver, BC, Canada). We explained experimental procedures and risks before obtaining written consent, which was approved by the Ethics Committee of the University of British Columbia and St. Paul's Hospital.
Pulmonary Function
Resting and Exercise Tests
Clinical Scores
Questionnaire and Patient Grouping ATS impairment criteria for FEV1% pred (four groups):
Normal: FEV1% pred
Mildly impaired: FEV1% pred
Moderately impaired: FEV1% pred Severely impaired: FEV1% pred < 41%
ATS impairment criteria for
Capable of continuous heavy exertion for an 8-hour shift:
Capable of work that is
Limited or incapable of most jobs: Study subjects were also categorized by questionnaire responses into the following three work/school status groups: Employed: Employed or attending school and did not miss work in past year because of CF Employed-limited: Employed or attending school but missed work in past year because of CF Unemployed-limited: Unemployed because of CF
We used the intensity codes described by Wilson and associates to calculate the energy cost of subjects' jobs (8). We calculated the cumulative energy cost in metabolic equivalents based on activities performed over an 8-hour shift. We then calculated equivalent values in units of oxygen consumption (ml/kg per minute) and then the percentage of energy cost of the job from their
Statistical Analysis
Of a total of 130 patients with CF attending our clinic, 73 patients were working or going to school full-time (or were classified as homemakers), 30 patients were working or attending school part-time or volunteering, and 16 patients were unemployed due to poor health. There were four patients actively searching for work. Table 1 presents data on the study sample and clinic patients who did not participate in the study and for whom we had complete demographic and work status data. Values for nonparticipants were obtained during periods of clinical stability. Stable clinical status was defined as the absence of pulmonary exacerbation over the previous 6 weeks (i.e., requiring intravenous antibiotics), the absence of a current mild exacerbation requiring treatment with oral antibiotics, and the absence of more than one of the following clinical symptoms: increased cough, sputum volume and purulence, increased dyspnea, reduced weight, and a fall in FEV1 > 10%. We looked at the distribution of study and nonstudy (remaining clinic) patients on the basis of work status according to ATS impairment/disability criteria for FEV1% pred (data on employment/education status were not available for seven patients). No significant differences were detected between the study patients and the remaining clinic patients in employment status or ATS disability grouping (p = 0.23). However, study patients were diagnosed at an earlier age and had lower levels of lung function.
Figure 1 shows the ATS disability criteria group for FEV1% pred by questionnaire-based employment group. Patients who were employed part-time and attending school part-time, or who were homemakers, were placed in the Employed group. We placed study patients who only volunteered in the Employed-limited group on the basis of responses that they were volunteering because they were unable to maintain employment as a result of their CF. Study patients who noted in the questionnaire that they were unemployed because of having CF were grouped in the Unemployed-limited group. According to ATS disability criteria grouping for FEV1% pred there were 11, 14, 23, and 25 patients classified as normal, mildly impaired, moderately impaired, and severely impaired, respectively.
Patients with CF designated as Unemployed-limited because of their CF had more severe pulmonary disease and, in fact, had mean FEV1% pred values similar to those of the Employed-limited group (Figure 1 and Table 2) . Neither change in FEV1 (i.e.,  FEV1) nor resting peripheral oxygen saturation (SpO2) was discriminating of disability group. Although there were no statistically significant differences in either resting SpO2 or change in SpO2 with maximal exercise [i.e., change in SpO2 from start of O2max protocol to maximal effort, abbreviated as SpO2(start–max)] between the three disability groups, the numbers suggest a more pronounced change in exercise SpO2(start–max) with maximal exercise in the Unemployed-limited group. Data on frequency and duration of treatment for pulmonary infections over 2 years and absenteeism are presented in Table 2. The results of applying the ATS O2max criteria for determining impairment in our study group are shown in Figure 2
. Results from the O2max test are presented in Table 3
. According to ATS impairment criteria grouping for O2max there were 57 patients classified as capable of heavy labor, 15 patients capable of work at < 40% O2max, and one patient classified as limited. Only one patient was classified as severely limited according to ATS criteria. This patient had recently stopped working because of disease-related symptoms and was being evaluated for double lung transplantation. Associations between potential exercise and clinical predictors are presented in Table 4
.
On the basis of the results of ordered logistic regression analysis (data not shown), FEV1% pred and S–K score were the only variables independently predictive of disability group. Given the high correlation (r = 0.75) between FEV1% pred and S–K score (Table 4), neither was preferred over the other; when both FEV1% pred and S–K score were in the same model, neither variable was statistically significant (p = 0.15). We also collected data on marital status, living arrangements, place of residence, and income sources. Approximately half of the patients in each group were single and the remainder of the patients were either married or living common-law. Of those patients who were single, the majority were living with their parents, or alternatively, with roommates. Sixty-two and 88% of patients in the Employed and Employed-limited groups, respectively, responded that their income was based predominantly on employment income, whereas patients in the Unemployed-limited group were supported by income assistance or disability programs. We next calculated the energy cost of the study subjects' jobs according to Wilson and coworkers (8). For the Unemployed-limited group, we used each subject's last recorded job for the calculations (Table 3). We showed a higher energy cost of work for the Employed-limited group compared with the Employed group. There was a moderate association between increasing pulmonary disease severity (i.e., FEV1% pred) and higher energy cost of job performance (r = -0.44, p = 0.0001) and similarly for energy cost of job performance and days hospitalized for pulmonary infections (r = 0.32, p = 0.009).
In the questionnaire, patients were asked whether CF was a consideration in their job choice. Twenty-nine percent of our patients stated that they took into consideration having CF in choosing their postsecondary schooling. All study patients had completed elementary school and all but three patients had completed high school. A small proportion of patients stated that they required additional time to complete their secondary education and obtain their high school diploma (two, two, and four patients, respectively, in the Employed, Employed-limited, and Unemployed-limited groups; i.e., 11% of the total study population) and the delay was attributed to CF-related complications. The majority of our study subjects had attended postsecondary institutions (only two, two, and five from the Employed, Employed-limited, and Unemployed-limited groups, respectively, did not attend postsecondary institutions) and only a few required additional time beyond that allotted to complete their programs (five, four, and two patients, respectively, in the Employed, Employed-limited, and Unemployed-limited groups). In the Employed group there were 12, 11, and 7 patients who had either completed or were attending university, college, or a technical school, respectively, and 8 patients who had not completed their postsecondary education program. In the Employed-limited group there were nine, two, and five patients who had either completed or were currently attending university, college, or technical school, respectively, and eight patients who had not completed their education program. There were only three patients in the Unemployed-limited group who had attended postsecondary institutions and three other patients in this group stated that they had dropped out of their postsecondary education program. Table 5
summarizes study subject responses about educational and career goals and whether they thought having CF affected the goals they set. Of those who had been unemployed at some time, only those who were currently not working because of their CF (i.e., the Unemployed-limited group; 10 of 11) stated that CF complications were responsible for their unemployment and that they previously had been employed. Patients were also asked to rate the physical intensity of their job as sedentary, semiactive, active, or labor intensive. Figure 3 shows the Employed and Employed-limited groups based on work intensity rating by %pred
In this study we determined the proportion of a cohort of adult patients with CF who are disabled from employment or schooling. We examined the utility of different measurements of resting pulmonary function and exercise capacity that are used to apportion disability/impairment in other respiratory diseases, for patients with CF. This study shows that pulmonary function and exercise capacity measures have limited utility in predicting work/school status. Moreover, other common clinical parameters do not reliably predict those unable to work/attend school because of CF-related health issues. Cystic fibrosis is generally characterized by a longitudinal decline in lung function, punctuated by acute exacerbations. Disease-related disability will of necessity become an issue for most patients with CF who have entered the workforce. Criteria to apportion disability would therefore be of relevance to individuals with this diagnosis. Ortega and associates have shown that a combination of resting pulmonary function measurements and exercise performance assessment can more accurately determine impairment in patients with chronic obstructive pulmonary disease (2). FEV1 is useful for the detection of severe impairment based on resting pulmonary function. Determination of other resting pulmonary function measures and arterial blood gases is invasive and may not provide much more additional information. In this study we found measures of airflow obstruction to be reduced in patients subjectively disabled from work compared with those fully employed. Furthermore, FEV1 was identified as a factor in predicting overall work-related disability. Several observations, however, indicate that the clinical utility of this parameter is limited in discriminating impairment. We found comparable group means of FEV1% pred between those unable to work and those working part-time. A threshold of FEV1% pred < 40% predicted has been used to denote severe impairment (1). We found this threshold to be a poor discriminator of those unable to work. Although a large number of those subjectively unable to work had a FEV1% pred < 40% predicted (5 of 11), appreciable numbers of those working full-time (8 of 37) or part-time (12 of 25) had similar levels of lung function. Similarly, other clinical parameters were of limited utility in apportioning disability. Resting SpO2, body mass index, and chest radiography scores were comparable between groups.
We reasoned that cardiopulmonary exercise testing would offer a better characterization of work ability. This is an integrative test that examines not only the ventilatory response to exercise but also a composite of cardiac, peripheral muscle, and volitional aspects all relevant to patients with CF. In other respiratory diseases, limited correlation has been shown between lung function parameters (FEV1, FVC, and DLCO) and
We next looked at measures of the oxygen cost of the work performed to determine whether those working with more advanced disease responded by selecting more sedentary occupations. We hypothesized that as clinical status deteriorated, individuals would perform at a higher percentage of their maximal oxygen consumption and were thus more likely to perceive their jobs as more labor intensive. We did find higher energy cost for work performed by those employed part-time and who believed themselves limited (i.e., Employed-limited group) compared with our full-time, not limited group (i.e., Employed group). We also showed patients with CF who left the workforce because they felt unable to perform their jobs adequately owing to CF (i.e., Unemployed-limited group) to have similar energy costs of work compared with those still employed part-time but limited (i.e., Employed-limited group). The majority of our patients at all stages of disease severity were employed in occupations with low metabolic costs, such as secretarial or clerical positions, and did show increasing pulmonary disease severity to be associated with higher energy cost of work. It has been proposed that the average metabolic work requirements of a job not exceed 40% of predicted
The limited utility of specific thresholds of lung function and exercise capacity used in the ATS documents for predicting disability of patients with CF is not surprising. Such factors characterize impairment and are only one aspect leading to disability. Patients with CF are younger than the populations used to define these parameters and are likely to be more motivated to pursue and maintain employment as a demonstration of independence. A fundamental question arising from our data is why a proportion of patients with CF believed themselves unable to work with lung disease severity that was not marked. As described by others for patients with chronic progressive diseases, the process of discontinuing work is multifactorial and involves interactions between specific job characteristics (physical demands of the job, job pace, flexibility of work schedule), the particular physical/clinical impairments of patients (2, 9), and the ability of patients to cope with their changing physical and financial condition (psychological factors) (10). Psychological factors of dealing with a chronic debilitating disease are not addressed by the ATS criteria. In CF, other aspects of the disease, which can also contribute to disability, include gastrointestinal disease, diabetes mellitus, and impaired skeletal muscle endurance and contractility leading to muscle fatigue. Ratings of perceived exertion (RPE) values for legs were higher than dyspnea scores in the Only a few studies to date have examined employment and work disability in a sizeable cohort of patients with CF. Gillen and colleagues studied a subgroup of adult patients with CF attending a clinic in Oregon, and reported on 49 patients, all employed at some time (11). At the time of evaluation, 13 were working full-time, 14 were working part-time, and the remainder were unemployed. Details of pursuit of higher education were not provided. The number of individuals with workforce participation is comparable to that in our study. This is to be noted, given the differences in social structures between these two jurisdictions. Disease severity was of utility, particularly rate of decline in FEV1. However, specific thresholds were not examined. They noted that age, adult CF diagnosis, female sex, and single marital status (after adjusting for disease severity parameters) were important predictors of disability risk (11). Although patients in the Unemployed-limited group were diagnosed with CF at a much older age, we did not show significant differences across our groups and we also did not show such variables as age at CF diagnosis, age, and sex as significant predictors of disability. Goldberg and associates completed a 5-year longitudinal study of 52 adolescents with CF, looking at clinical measurements as well as administering questionnaires to determine education and employment status, intelligence, and self-esteem (12). They showed that mild disease severity was the only variable predictive of current education or employment status. A study by Schechter and colleagues, using the data from the U.S. National Cystic Fibrosis Foundation Patient Registry data (1986–1994), showed that socioeconomic status is an important predictor of clinical outcome (13). They used Medicaid status as a measure of socioeconomic status and showed that patients with CF and receiving Medicaid benefits showed more respiratory infections and worse pulmonary function, deficiencies in height and weight, and a greater probability of having positive cultures for Pseudomonas aeruginosa and Burkholderia cepacia than patients not receiving Medicaid benefits. It is clear that access to medical services will directly influence the educational and career goals of this group and their employment status. It would appear that patients with CF in our cohort were not encumbered by their disease, and of those not attending school, the majority were employed. The type of work was diverse, generally representative of occupations pursued by individuals of comparable education and social background, with the exception of underrepresentation of jobs requiring sustained heavy labor. Educational and work accomplishments are indicative of overall improvements in the health status of adults with CF, perhaps reflective of their high expectations of longer term survival. Treatment advances are geared to help minimize perturbations of school/work activities. Evidence of this is the increased number of pulmonary infections treated via home intravenous antibiotic programs. There is also the observation that many patients and their families are determined to minimize the effects of their disease and optimize quality of life. Virtually our entire CF cohort completed high school (i.e., 95.9%), the majority going on to attend postsecondary institutions. This is higher than our overall current provincial graduation rate (84.5%). Given the natural history of progression of this disease, it was surprising to find that delays in completion of schooling were modest. Completion of secondary school on time for our CF group (i.e., 84.5%) was similar to the rate for the province of British Columbia (81.5%). Even with these accomplishments, a number of patients volunteered that having CF modified their educational goals and objectives. This was manifested in a number of ways, for example, by enrolling in colleges rather than universities, and by reduced likelihood of pursuing advanced degrees.
The findings of our study require evaluation in the context of certain limitations. We conclude that patients with CF attempt to lead as normal a life as possible and pursue postsecondary education and participate in the workforce. Educational and work accomplishments are indicative of overall improvements in the health status of adults with CF, perhaps reflective of their expectations of longer term survival. Many patients with CF do take their disease into account when considering their educational and career options. Pulmonary function alone does not give an accurate picture of a patient's capabilities. Exercise testing does provide additional information about a patient with CF that cannot be obtained from resting pulmonary measures. We found patients who were predicted to be impaired on the basis of resting pulmonary parameters to be incorrectly categorized when exercise parameters were considered. We propose that both measures be considered when assessing impairment and predicting disability in the CF population. Perhaps for this population it would be practical to calculate and consider S–K scores for an overall representation of clinical status, and thus also consider the frequency of pulmonary infections. It is of concern that approximately one-quarter of the study sample believed that at some point in their life they were denied employment because they had CF. Measures to ensure that such discrimination is the exception must be enacted and enforced. For example, in the United States, the Americans with Disabilities Act protects individuals who have CF and other conditions.
Supported by a Canadian Cystic Fibrosis Foundation Studentship and a Michael Smith Foundation for Health Research Trainee Award to D.D.F. Received in original form February 20, 2002; accepted in final form October 7, 2002
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
INTRODUCTION
80% 
40% of observed 
2 analysis. Ordered logistic regression analysis was used to identify those predictor variables that were independently predictive of work/school status group. For this analysis, disability group was regressed on each pair of potential predictor variables, using Stata statistical software (Stata, College Station, TX). 
