Published ahead of print on December 3, 2004, doi:10.1164/rccm.200408-1054OC
© 2005 American Thoracic Society doi: 10.1164/rccm.200408-1054OC
Exacerbations and Time Spent Outdoors in Chronic Obstructive Pulmonary DiseaseAcademic Unit of Respiratory Medicine, St. Bartholomew's and Royal London School of Medicine and Dentistry, London, United Kingdom Correspondence and requests for reprints should be addressed to J. A. Wedzicha, M.D., Academic Unit of Respiratory Medicine, Dominion House, St. Bartholomew's Hospital, London EC1A 7BE, UK. E-mail: j.a.wedzicha{at}qmul.ac.uk
Patients with chronic obstructive pulmonary disease have a progressive reduction in activity, although its time scale and the contribution of exacerbations are unknown. A rolling cohort of 147 patients (101 male; mean age, 88.5 years; and forced expiratory volume in 1 second as percent predicted, 38.4%) were monitored for a median of 1,044 days (interquartile range, 685 to 1,779) over an 8-year period starting in March 1996. Patients recorded any increase in daily respiratory symptoms and time spent outside their home. They completed the St. George's Respiratory Questionnaire yearly. They experienced 1,465 exacerbations and time outdoors decreased by –0.16 hours/day per year (p < 0.001). This decline was faster in frequent exacerbators (p = 0.011). Before exacerbation, the patients stayed indoors all day for 2.1 days/week (on any day 34.1% were at home), but for 5 weeks postexacerbation they spent 2.5 days/week at home (p < 0.001) (44.4% remaining at home at onset; p = 0.021). St. George's total, activity, and impact scores were independently associated with time outdoors (p < 0.005), but not with symptom score. In conclusion, time spent outside the home declines over time and acutely at exacerbation. Patients with frequent exacerbations are more likely to become housebound and need targeting in rehabilitation programs.
Key Words: chronic obstructive pulmonary disease • exacerbations • FEV1 decline • housebound • quality of life Patients with chronic obstructive pulmonary disease (COPD) experience over many years an accelerated decline in forced expiratory volume in 1 second (FEV1). Thus patients become progressively more dyspneic and lose exercise capacity—although these two factors are independent of each other (1). Consequently, patients present with respiratory and skeletal muscle weakness and deconditioning (2, 3). At some point, with disease progression, a patient may become housebound (4). In some patients, this event may coincide with the prescription of long-term oxygen therapy, as 45% of these patients do not go out except for medical visits to their physician or hospital (5). In a cross-sectional study, we reported that patients with COPD classified as Grade 5 on the Medical Research Council (MRC) Dsypnea Scale, that is, "too breathless to leave the house," had low exercise capacity with a shuttle walking distance of only 90 m compared with 209 m for patients at MRC Grades 3 and 4. They were also older, with a worse health status score and greater impairment of their activities of daily living, but importantly, no difference in the degree of airflow obstruction as assessed by FEV1 (6). To date, no studies have yet examined over a prolonged period the slow reduction in amount of time that patients with COPD spent outside their home. Patients with COPD also experience exacerbations, the majority of which are triggered by viral or bacterial infections (7, 8). Exacerbations have become recognized as important events in the progression of the disease. Patients with frequent exacerbations suffer a faster decline in FEV1 (9, 10), have a poorer health-related quality of life (11), and raised airway inflammatory markers in the stable (nonexacerbation) state (12). The impact of exacerbations on the behavior of patients with COPD has not been studied. We hypothesized that exacerbations would impact on time spent outside the home, both short-term during exacerbation and over the longer term. To investigate this we have used the time spent outside the patients' homes as an outcome measure that could be collected accurately with respect to timing before the onset of an exacerbation. A walking test or a questionnaire on the nature of daily activities would have to be applied retrospectively because it is impossible to forecast an exacerbation, and impractical to frequently assess patients. The East London COPD Study started in November 1995. From March 11, 1996 we asked the patients to record on their diary cards the time they spent outside their home each day. We have now analyzed these data to (1) describe the time course of confinement at home (or its converse, going out) during an exacerbation, (2) determine whether patients who experience frequent exacerbations have a faster decline in outdoor time, and (3) identify which respiratory symptoms at exacerbation most impair outdoor activity. Some of the results of this study have been previously reported in the form of an abstract (13).
Patients One hundred forty-seven patients with COPD recruited into the East London COPD Study were selected on the basis of the fact that they had recorded daily data for 1 year or more between March 11, 1996 and March 31, 2004. The recruitment and monitoring of these patients have been previously described (7, 9, 11, 12, 14). Briefly, patients were initially recruited consecutively from an outpatient clinic. There were 59 patients in March 1996. New patients were added to this rolling cohort when patients decided to leave or died. COPD was diagnosed as FEV1 less than 70% predicted from age, height, and sex; ß2-agonist reversibility less than 15% and/or less than 200 ml; and FEV1/forced vital capacity (FVC) less than 70%. Patients with a significant respiratory disease other than COPD or an inability to complete diary cards were excluded.
Recruitment The study had ethics approval from the Ethics Committee of the East London and City Health Authority and the patients provided written informed consent.
Monitoring Patients attended a clinic every 3 months at which FEV1 was measured. Once per year, they were asked to complete unsupervised a St. George's Respiratory Questionnaire (SGRQ) and indicate their disability using the MRC Dyspnea Scale (16). The SGRQ total and component scores for symptoms, activity, and impact were calculated on the basis of empirically derived weightings. Scores varied from 0 (no disability) to 100 (maximal disability). The SGRQ data collected before November 2002 have been previously reported (11, 14), but we have not previously reported longitudinal changes, or related such changes to time spent outdoors. The MRC score is a five-item questionnaire (16), in which Grade 5 corresponds to "too breathless to leave the house."
Exacerbations The severity and recovery time of PEF and symptom count were first calculated for each exacerbation, and then an average was calculated for each patient. This avoided bias by patients with frequent exacerbations or who were in the study for longer times. Exacerbation severity was taken as the change from a baseline determined as the mean value over Days –14 to –8 preceding exacerbation onset, and the day of exacerbation onset. Recovery was defined as the time from onset for a 3-day moving average of the parameter to equal or exceed baseline within a period of 35 days. These measures have been previously described and justified (14, 18). To investigate the time course of time outside the home, data on how long the patient spent outside was converted to a binary variable of whether they remained confined indoors all day or went outside. The severity of exacerbation was defined as the difference between the number of days per week, over the period (Days 1 to 35 after onset) they remained indoors all day, as a change from the number of days per week they were indoors all day during the baseline period (Days –14 to –8 relative to onset). Recovery was taken as the number of consecutive days from the exacerbation onset that they remained confined indoors. Again, patient average values were calculated to avoid the bias described above. Records of hospitalization for exacerbation of COPD were kept throughout the study. None of the patients underwent a pulmonary rehabilitation program during the time of the study.
Statistical Analysis Analysis of longitudinal changes over time in SGRQ, FEV1, and time outdoors was made with regression models that used a generalized estimating equation approach, with adjustments made for the distributions described above in the dependent variable. These models allow for the correlation between repeated measures data typically found in panel data (19, 20). Whether or not patients with frequent exacerbations had a faster decline in outdoor excursion time was examined with these models by dividing the patients into either a high– or low–exacerbation frequency group (greater than or equal to, or less than, the cohort median of 2.47/year) and by then testing the interaction between time and group with allowance for the main effects of group and time. These models, with allowance for the binomial data distribution and with a logit link function, were use to assess whether the odds of being at MRC Grade 5 or not changed over time. For the analysis of longitudinal changes in time outdoors, there were 184,957 data points. This number exceeded our computer's memory for manipulation, and so we used only data recorded on the day before we had either FEV1 or SGRQ measurements. All statistical analysis was performed with Stata 5.0 software (Stata, College Station, TX).
Eurowinter Study
Patient Characteristics The 147 patients (101 male, 46 female) studied had moderate to very severe COPD (see Table 1). Of these, at the start of the study 129 patients were inhaling steroids daily and 12 patients were taking oral prednisolone (mean [SD], 6.5 [4.0] mg/day). Between March 11, 1996 and March 31, 2004, the patients recorded diary card data on a median of 1,044 (IQR, 685–1,779) days. During this period 62 patients asked to leave the study and another 32 died. There were a total of 1,465 exacerbations. The median number of exacerbations per patient per year was 2.47 (IQR, 1.45 to 3.84). Nine patients had no exacerbations.
Time Course of Time Spent Outdoors at Exacerbation Figure 1A illustrates the time course over 51 days of all exacerbations, in the percentage of patients confined indoor all day together with changes in symptom count. Figure 1B shows the converse of the above, that is, the percentage going out, together with changes in PEF. Over the baseline period, patients remained indoors on 3,957 of 9,663 days for which patients recorded data, that is, 34.1% of days, whereas 44.4% (552 of 1,244 days) remained indoors on the day of exacerbation onset (Day 0) and 39.7% remained indoors (17,032 of 42,864 days during a postexacerbation period from Days 1 to 35). In comparing the baseline period with onset and postexacerbation periods, there was significantly reduced outdoors activity, p values were 0.021 and 0.024, respectively.
After averaging data from repeat exacerbations by the same patient to avoid bias, the median number of days per week they stayed at home all day during the baseline period was 2.1 days/week (IQR, 0.7 to 3.8 days/week). This rose by 0.4 days/week (IQR, 0.03 to 0.82 days/week; p < 0.001) to 2.7 days/week (IQR, 1.3 to 4.5 days/week) in the postexacerbation period (Days 1 to 35). The median number of consecutive days they were confined at home was 1.5 days (IQR, 0.5 to 3.0 days). Table 2 shows that the median symptom count at exacerbation was 2.5, a rise of 2.05 from 0.21 at baseline. Similarly, PEF at exacerbation was 215 L/min, a fall of –8.9 L/min from 219 L/min at baseline. The increase in home confinement pre- to postexacerbation was correlated with the fall in PEF (rho = –0.182; p = 0.0387) but not with the rise in symptom count (rho = 0.142; p = 0.110). At baseline, when the patients were stable, confinement indoors was correlated with baseline PEF (rho = –0.327, p = 0.002) and baseline symptom count (rho = 0.177; p = 0.0449).
Influence of Exacerbation Symptoms on Outdoor Excursion Time Table 3 shows the effects of single-symptom and double-symptom combinations on the severity of confinement at home. With a Bonferroni adjustment for 29 statistical tests, only probabilities at the p = 0.00172 level were considered significant. A sore throat either alone or in combination with dyspnea or cough caused a significant increase in the number of days per week confined indoors postexacerbation.
Justification for Choosing Excursion Time on the Day before to Related to SGRQ Scores Time outdoors was recorded for the same day as SGRQ administration on 412 of 511 occasions (lag, 0) and 418 times on the day before (lag, –1). Outdoor time on the day of administration was significantly higher (Wilcoxon, p < 0.001) at 3 hours (IQR, 1.5–4 hours) compared with the day before, when the median excursion time was 2.0 hours (IQR, 0–3.5 hours). As attending the clinic may not have been representative of the patient's normal activity, the following analyses use time outdoors recorded for the day before the questionnaires were administered.
Relationships between SGRQ Scores and Time Outdoors Table 4 gives the results, based on the data described above, of a multivariate analysis of the relationship between SGRQ scores against outdoor time, age, exacerbation frequency, and FEV1. Total, impact, and activity scores were, respectively, 1.14, 1.31, and 1.15 units lower if the patients were able to spend an extra hour per day outdoors (all, p < 0.05). The symptom score was not significantly related to time outdoors. These results were essentially unchanged if FEV1 was expressed as a percentage of the predicted FEV1, except that for the regression involving the total score, the p value for the independent effect of the time outdoors was 0.051.
Longitudinal Changes Table 5 and Figure 2 show, respectively, that daily time outdoors fell significantly by 0.14 hours/year and FEV1 fell significantly by 41.1 ml/year—5.2 and 3.6%/year, respectively—relative to estimates of their values at the start of the study on March 11, 1996. Over the same period and calculated in a similar manner, the total SGRQ score rose by 1.94%/year, the activity score rose by 2.1%/year, and the impact score rose by 3.8%/year, whereas the symptom score fell by 2.0%/year (see Figure 3).
Patients with frequent exacerbations (2.47 exacerbations or more per year) had a faster decline in time outdoors of –0.07 hours/year (95% confidence interval [CI], –0.13 to –0.02 hours/year; p < 0.011) in addition to the decline of –0.10 hours/year seen in infrequent exacerbators. There was no difference in age between the frequent and infrequent exacerbators (66.6 [SD 1] and 68.5 [SD 0.8] years, respectively; p = 0.132).
Hospitalization
Longitudinal Changes in Recording MRC Grade 5
Normative Data from the Eurowinter Study The Eurowinter Study (21) collected data on the time and frequency of trips made by London residents to unheated areas outside the home during the winter of 1994–1995. Data from this survey show that 21.7% (112 of 516) of a 50- to 59-year age group, equally divided between men and women, remained indoors all day, whereas 27.8% (149 of 537) of a 65- to 74-year older group were indoors on the day of interview. This difference between the two age groups was significant (p = 0.023). In comparison with the normative data given above on people living in the same city and of the same age, patients with COPD when stable (exacerbation free) spent more of their days indoors (34%; 3,957 of 9,663 days) throughout the year compared with 27.8% of normal people during the winter months when fewer people venture out (p < 0.001).
This is the first study that has recorded, over a number of years, the time spent each day outside the home by patients with moderate to very severe COPD. With these unique and prospectively collected data, we have shown that patients with a history of frequent exacerbations reduced the time they spent outdoors at a faster rate compared with those with infrequent exacerbations, and thus are more likely to become housebound. We also showed that patients tend to stay indoors during an exacerbation, which might explain why approximately some 50% of exacerbations are not reported to a physician (14). In comparison with the normative data given above from people living in the same city and of the same age, patients with COPD when stable (exacerbation free) were indoors for a greater proportion of days during the week. We found that our patients with COPD progressively reduced the time they spent outdoors per day by 0.14 hours/day per year. The Eurowinter Study found that a significantly greater proportion of older people remained indoors (7). Our group of patients was in the older age range, and we found a significantly faster decline among patients with frequent exacerbations (2.47 exacerbations or more per year), although there was no difference in age between the frequent and infrequent exacerbators. This could be explained in part by the faster decline in lung function found in frequent exacerbators (9, 10). However, frequent exacerbators also experience high levels of both airway inflammatory markers when stable (12, 22). The systemic inflammatory response characteristic of more severe COPD has been associated with weight loss and muscle wasting (23–25) and this loss of physical strength could contribute to the faster reduction in activity. It is also possible that the increased systemic inflammatory response at exacerbation contributes to peripheral muscle weakness (26). We have reported that patients who are receiving long-term oxygen therapy are more likely to be housebound (5). Patients who are receiving long-term oxygen therapy are more anxious and depressed than other patients with COPD (27). This may be due to being severely hypoxemic as well as housebound (27–29). This depression and a dependence on a stationary oxygen source for long-term oxygen therapy provision would both contribute to a reduction in outdoor activity. The combination of these factors might explain why the decline in time outdoors of 5.2%/year was greater than the deterioration in total SGRQ score of 1.9%/year or the 3.6%/year reduction in FEV1. Patients in the MRC Grade 5 housebound group are an important target group for pulmonary rehabilitation programs. However, we have previously reported that a home-based individualized exercise program was not effective in this group (6). This could be due to (1) the patients' poor physical fitness and muscle conditions, or (2) the difficulties in achieving the exercise intensities required to produce a meaningful gain in exercise capacity. Thus new strategies for rehabilitation are required, aimed at encouraging patients to undertake normal indoor activities and to leave the home with adequate support.
We found that the total, activity, and impact SGRQ indices worsened over time. No previous study has reported trends in health-related quality of life over as long a period as 8 years. Our annual rate of increase in the total score of 1.14 units/year is lower than that previously reported for a group of patients with COPD who, like our patients, were taking daily inhaled steroids at 2.0 units/year (30). Most of this discrepancy may be due to the symptom component. In this study, the symptom score fell (improved) by –1.1/year in contrast to the ISOLDE (Inhaled Steroids in Obstructive Lung Disease in Europe) Study (30), in which it rose (became worse) by 1.02 units/year. Dowson (31) in a study of In a pre- and postexacerbation comparison, time spent outdoors was reduced by a median of 0.4 days/week and 1.5 consecutive days were spent indoors immediately postexacerbation. This quick return to normality may reflect a reduction in the severity of the symptoms, the need to go shopping, a desire to visit friends or relatives, or a determination to keep appointments. Our finding of a rapid recovery in going outdoors after exacerbation differs from the prolonged impact of an exacerbation on health-related quality of life for up to 26 weeks, reported by Spencer and colleagues (30). This suggests that exacerbations have a greater impact on the patient's perception of health rather than any physiological or behavioral change. We found that exacerbations associated with a sore throat had the greatest effect on the severity and recovery in time outdoors. This may be consistent with our earlier findings that exacerbations involving colds and thus viral infections, which could be preceded by symptoms of a sore throat, are of greater severity when assessed by changes in symptoms, lung function, and airway inflammatory markers (7, 32). The finding that patients with frequent exacerbations have a faster decline in the time they spend outdoors has some important implications. This group of patients will become housebound at a younger age. This may have considerable health economic implications for the community as home visits by nurses and physicians are expensive. It also suggests that pulmonary rehabilitation programs should specifically target this group of patients early in the progression of their disease. We have also shown in this study that patients with little outdoor activity are more likely to be hospitalized, which is consistent with the report of a strong association between usual physical activity and a reduced risk of readmission to hospital (33). As pulmonary rehabilitation programs have been shown to reduce hospitalization, they will be of particular benefit to patients with a history of frequent exacerbations and are also more cost-effective (34). In this study, we have used the measurement of time outdoors as an outcome measure. It will depend not only on the patient's respiratory health but also on walking ability, access to transport, lifestyle, proximity to shops and friends, and restrictions due to therapy. However, it has the advantages of simplicity of understanding and we have in this study shown that it can be collected on a daily basis over a considerable number of years. Time outdoors cannot be equated with physical or muscular activity. It is also likely that much of the time spent outside the person's home is spent indoors elsewhere. However, as an outcome measure it is responsive to transient events such as exacerbations, and as we have shown it is an independent factor determining the patient's health-related quality of life. Thus, we believe it to be a helpful and usable measure in these patients with COPD. In conclusion, time outdoors is a simple and useful outcome measure for monitoring the progress of COPD in behavioral terms. Patients with frequent exacerbations are likely to become housebound at an earlier stage in the disease, and thus have a poor quality of life and are at risk of hospitalization. Pulmonary rehabilitation programs should be particularly targeted at patients with a history of frequent exacerbations, so that optimal physical activity can be maintained.
The authors are grateful to all the patients with COPD in the East London cohort who have contributed to the study. The authors also thank Terry Seemungal, Angshu Bhomik, Mark Roland, Simon Leedham, Simon Lloyd-Owen, and Irem Patel for assistance with the collection of clinical data. The authors also acknowledge the Eurowinter coordinating team: W. R. Keatinge and G. C. Donaldson, and project members K. Bucher, G. Jendritsky, E. Cordioli, M. Martinelli, L. Dardanoni, K. Katsouyanni, A. E. Kunst, J. P. Mackenbach, C. McDonald, S. Nayha, and I. Vuori.
Supported by the British Lung Foundation and the Joint Research Board of St. Bartholomew's Hospital. This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org Conflict of Interest Statement: G.C.D. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; T.M.A.W. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; J.R.H. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; W.R.P. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; J.A.W. has received honoraria for lectures at meetings and/or attendance at advisory boards from Glaxo and then Glaxo-SmithKline (GSK), Boehringer Ingelheim, AstraZeneca, SmithKline Beecham, Zambon, Bayer, Merck, Roche, Teijin, Medivir UK, and VitalAire and she has received research grants totaling approximately $600,000 from GSK for studies of COPD questionnaires, studies of COPD exacerbation time course, flow cytometric studies of the interaction of the upper and lower airway in COPD, a study of the time course of inflammatory markers at COPD exacerbation, an analysis of long-term changes in airway and systemic markers, a clinical trial of anticough therapy in COPD, and a clinical trial of Seretide in COPD and she has received a grant of approximately $450,000 from Boehringer for a study of tiotropium in COPD and $25,000 from AstraZeneca for a health economic study and $300,000 from Aventis Pasteur for a study of viral epidemiology of COPD. Received in original form August 13, 2004; accepted in final form November 30, 2004
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
2 test. Recruitment data were described by a mean and standard deviation (SD). 
-1 antitrypsin–deficient patients over 24 months also showed a fall in symptom score of 3.45 units/year. This finding could be explained by early or optimized treatment that lessens the impact of respiratory symptoms. Alternatively, it may be due to the inhaled steroids that 89% of our patients were taking. A third possibility is that patients become more tolerant of their respiratory symptoms over time. A fourth is that they modify their behavior and reduce activity such that they experience fewer periods of dyspnea and thus their perception of their symptoms might lessen.