OUP user menu

Fatigue in bronchiectasis

K.L.M. Hester, J.G. Macfarlane, H. Tedd, H. Jary, P. McAlinden, L. Rostron, T. Small, J.L. Newton, A. De Soyza
DOI: http://dx.doi.org/10.1093/qjmed/hcr184 235-240 First published online: 20 October 2011

Abstract

Background: Fatigue is a complex, disabling symptom in non-CF bronchiectasis (nCF-Br). Fatigue can be formally measured using the validated fatigue impact scale (FIS). The relationship between fatigue and clinically important factors such as airflow obstruction, breathlessness or Pseudomonas aeruginosa infection in nCF-Br is unclear.

Aim: To measure the correlation between FIS scores and markers of disease severity in nCF-Br.

Design: A prospective cohort study.

Methods: Patients attending a specialist service were studied. Lung function (FEV1% predicted), Medical Research Council dyspnoea score (MRCD), sputum culture results and FIS were recorded. Patients were categorized according to sputum culture into three subgroups: Pseudomonas ‘colonization’, ‘isolation’ and neither.

Results: One hundred and seventeen consecutive patients were included. Average FEV1% predicted was 64% (SD ±28%). Twelve (10%) patients had Pseudomonas aeruginosa isolation; 47 (40%) patients had P. aeruginosa colonization. Fatigue levels were similar in patients with and without colonization (median 38 versus 32, P = 0.155). Significant fatigue (FIS > 40) was similar in all three Pseudomonas subgroups (P = 0.31, chi-square). Fatigue correlated with MRCD score (r = 0.57, P < 0.001) and FEV1% predicted (r = −0.30, P = 0.001). FEV1% predicted was lower in patients who had ever isolated or been colonized with P. aeruginosa (P ≤ 0.001).

Conclusions: There are significant correlations between FIS score and MRCD score and FEV1% predicted in bronchiectasis. Pseudomonas aeruginosa infection appears to be associated with poorer lung function, and higher MRCD scores, yet there is no significant association between P. aeruginosa status and fatigue.

Introduction

Non-cystic fibrosis bronchiectasis (nCF-Br) has been associated with high levels of fatigue, with one study reporting fatigue as a symptom in up to 74% of patients.1 There is, however, limited objective questionnaire data for bronchiectasis patients. The symptom of fatigue is common2 and debilitating and is recognized as a symptom in patients with a variety of chronic diseases.3 In order to formally quantify fatigue, a number of measures including the fatigue impact scale (FIS) can be used. The FIS examines patients’ perception of their fatigue, and its impact on cognitive, social and psychosocial functioning.4 FIS scores have been well validated in numerous chronic health conditions, including respiratory diseases such as chronic obstructive pulmonary disease (COPD).5 Levels have been demonstrated to be higher in COPD than healthy age-matched subjects,6 impacting on quality of life and exercise tolerance.7 Higher FIS scores have been associated with premature death over 9 years in other chronic conditions, such as the autoimmune cholestatic liver disease primary biliary cirrhosis (PBC).8 In patients with PBC, a previous study has found the median FIS score to be 40, and this has been used as a cut-off point to signify high fatigue levels.9 Recent work carried out at our centre demonstrated that nearly half of patients with COPD or nCF-Br had FIS scores >40 with fatigue severity being comparable between both groups.10 The specific reasons for fatigue in nCF-Br are unclear, but potentially include chronic infection and exacerbation rates, physical deconditioning, dyspnoea, anxiety and depression or other co-morbid conditions. Overall poorer outcomes such as increased exacerbation rates and higher risk of death have been reported in Pseudomonas aeruginosa infected nCF-Br patients.11 In the same study, fatigue in patients with nCF-Br had a positive correlation with mortality.

Moreover, P. aeruginosa colonization is known to be associated with lower FEV1.12 We, therefore, aimed to investigate fatigue in the context of other clinically important factors. We focused on exploring the relationship between FIS scores and P. aeruginosa infection as well as other parameters of disease severity represented by dyspnoea and airflow limitation, i.e. FEV1.

Methods

Consecutive patients from the specialist bronchiectasis clinic at the Freeman Hospital were studied and all had a diagnosis of nCF-Br defined by clinical history and high-resolution computed tomography scan (HRCT).

Patients attending routine outpatient appointments were asked to complete an MRCD questionnaire and the FIS assessment tool. The FIS assessment tool consists of 40 questions, having a maximum possible score of 160. There are three subscales covering psychosocial, cognitive and physical domains. Any total score >40 was considered to be clinically significant based on previous work.9

This study was carried out as an audit and as such formal ethical approval was not required.

Lung function tests

Spirometry was performed in clinic using Vitalograph wedge bellows according to European Respiratory Society (ERS) guidelines13 and using European Community for Coal and Steel (ECCS) predicted values.14 FEV1, FVC and FEV1% predicted were recorded.

Sputum microbiology

Sputum cultures of participating patients from the preceding 12 years were retrospectively reviewed. All patients attending clinic are expected to produce a spontaneous sputum sample. Current sputum microbiology was reviewed in addition to all available previous results.

Patients were subdivided for further analysis dependent on their P. aeruginosa status as per criteria defined by Pasteur et al.15 Three cohorts were formed: those who had never isolated P. aeruginosa, those with prior ‘isolation’ (organism cultured on at least one occasion) and those with ‘colonization’ (organism cultured on two or more occasions, 3 months apart within a 1-year period).

Statistical analysis

Statistical analysis was performed using SPSS statistics 17.0 to examine relationships between FIS score and P. aeruginosa status, FEV1% predicted and MRCD score. Spearman's rank correlation, Mann–Whitney U-test and chi-squared tests were used. Kruskal–Wallis tests were used to identify differences in the distributions between groups. A probability (P) value of <0.05 was deemed to indicate statistical significance.

Results

One hundred and seventeen consecutive patients were included in the study, 71 females and 46 males. The overall median age of the patients was 60, with a range of 16–83 years.

Lung function

Spirometry was performed on all patients demonstrating a mean FEV1 of 64% predicted (SD ±28%) and median 60% (range 18–133%).

MRCD scores

All patients completed an MRCD questionnaire. Overall the median score was 3 (range 1–5).

FIS scores

FIS scores were recorded for all patients. The median FIS score was 35 (range 0–160). Notably 53 patients (45%) had a clinically significant FIS score (>40).

Microbiology

The mean follow-up in terms of documented microbiological samples was 7.9 years (SD ±4.17 years). Sputum microbiology data demonstrated that 47 patients (40%) were colonized with P. aeruginosa as per previously described criteria. Twelve (10%) patients were classed as having isolation of P. aeruginosa only and the remaining 58 (50%) had never isolated P. aeruginosa from a sputum sample.

The relationships between fatigue and lung function, dyspnoea and microbiological status were examined further.

FIS and microbiology

There was no significant difference in the distributions of the FIS scores between the three P. aeruginosa subgroups (P = 0.364, Kruskal–Wallis). Fatigue levels were similar in the 47 patients with persistent P. aeruginosa colonization (median FIS 38) and the 70 patients without colonization (median FIS 32) (P = 0.155, Mann–Whitney). Although FIS scores were higher in those who had isolated P. aeruginosa (median 44) than those who had never isolated the organism (Median 31), this difference was not statistically significant (P = 0.981, Mann–Whitney).

The incidence of significant fatigue (FIS >40) was not associated with the P. aeruginosa status (P = 0.31, chi-squared) (Table 1).

View this table:
Table 1

The relationship between P. aeruginosa isolation (organism cultured on ≥1 occasion) and P. aeruginosa colonization (organism cultured ≥2 occasions, 3 months apart in a 1-year period) and fatigue in adults with non-CF bronchiectasis

Pseudomonas aeruginosa isolatedPseudomonas aeruginosa never isolatedPseudomonas aeruginosa colonizedP-value (Kruskal– Wallis)
Number of patients, n (%)12 (10)58 (50)47 (40)
Median FIS score4431380.364
Significant fatigue (FIS > 40) (%)5043570.31
Median FEV1% predicted (%)5776470.001

Lung function and microbiology

FEV1% predicted value differed significantly between the three subgroups (P = 0.001, Kruskal–Wallis). When further investigating these differences, patients who had never isolated P. aeruginosa in their sputum had a median FEV1% predicted value of 76% (range 25–133). This was significantly higher than that seen with P. aeruginosa colonization (median 47%, P = 0.001, Mann–Whitney), but not when compared to the group with P. aeruginosa isolation only (median 57%, range 18–106) (P = 0.091, Mann–Whitney; Table 1.) Additionally, there was no difference in FEV1% predicted between those who were colonized and those who had only previously isolated P. aeruginosa (P = 0.37, Mann–Whitney).

Lung function, dyspnoea and FIS

There was a positive correlation between FIS scores and MRCD scores (P < 0.001, r = 0.57, Spearman's rank correlation). FIS scores also correlated with FEV1% predicted value (P = 0.001, r = −0.30).

There was a significant difference in MRCD between the P. aeruginosa groups (P = 0.014, Kruskal–Wallis). There was a significant difference in MRCD between those who had never isolated P. aeruginosa (median 2, range 1–5) and those who had ever isolated or been colonized (median 3, range 1–5) (P = 0.004, Mann–Whitney). There was no difference, however, when comparing those who had isolated P. aeruginosa or were P. aeruginosa colonized (P = 0.83, Mann–Whitney).

Discussion

Fatigue is an important and prominent symptom for patients with bronchiectasis and impacts on both quality of life and exercise tolerance.4 Based on previous pilot data from our centre,10 patients with nCF-Br had similar fatigue levels as patients with COPD at a younger age, with comparable MRCD scores and with more preserved lung function. This led to the hypothesis that the infective burden in nCF-Br plays an important role in driving fatigue in this cohort.

We found that overall FIS scores were clinically significant in 45% of the nCF-Br cohort, demonstrating a significant incidence of fatigue within the patient group. In a previous study in 2006, 74% of a cohort of 103 patients with nCF-Br reported fatigue as a symptom.1 This was based on a simple survey without objective questionnaires (P. King, personal communication). Our incidence of fatigue was much lower than that previously reported1 despite our cohort having potentially more severe disease as suggested by a lower FEV1% predicted and a higher incidence of P. aeruginosa infection. These differences in fatigue rates observed previously and herein may reflect the difference between our use of a validated objective score rather than more subjective patient-reporting of a symptom. It is clear, however, that collectively the two studies demonstrate a high non-pulmonary symptom burden in patients with bronchiectasis. Importantly, as a number of new therapies in bronchiectasis are being developed, a key patient-reported outcome may be improvement in fatigue. We, therefore, propose that validated fatigue questionnaires are included in study design as a key patient-reported outcome measure.

The prior literature is limited in ascertaining the potential link between fatigue and P. aeruginosa infection. To address this specific gap in the literature, we looked at the relationship between fatigue and other parameters of disease severity. We have found within our cohort that P. aeruginosa isolation or colonization is associated with a significantly lower FEV1% predicted in keeping with previous studies. The absolute FIS scores in the P. aeruginosa colonized group and the P. aeruginosa isolator groups were not significantly different. This remains difficult to explain. One factor may be that those classed as ‘isolators’ are actually colonized with P. aeruginosa but some samples are culture negative. With the benefit of up to 12 years of longitudinal data for this cohort, it is apparent that many ‘isolators’ continue to have culture positive samples intermittently over the years, yet do not meet previously defined criteria for ‘colonization’. There was however no difference in FIS scores between those who had ever isolated or been colonized with P. aeruginosa, and those who had never isolated the organism. Within this cohort, we have therefore found no convincing evidence of a relationship between P. aeruginosa infection and fatigue. This is interesting considering its association with declining lung function and the correlation found between lung function and fatigue. A previous study by Wilson et al. in 1997 investigated the association between sputum bacteriology and quality of life in 87 patients with nCF-Br.16 In this cohort, 22 (25%) were colonized with P. aeruginosa. Quality of life was examined in the present study using three different scores, the St George's Respiratory Questionnaire, the SF-36 and the fatigue scale.17 The fatigue scale used in their study is a 14-item scale covering physical and mental domains, and they found significantly worse scores in only the physical domain in P. aeruginosa colonized patients. This prior study was more focused on quality of life in patients colonized with P. aeruginosa and did not examine how other factors interrelated with fatigue. The authors acknowledged that a number of factors potentially influenced the quality of life in patients with P. aeruginosa colonization and called for longitudinal studies in this area. Importantly, the FIS used herein consists of 40 questions, differing from the fatigue scale previously used.

Pseudomonas aeruginosa has been shown to be associated with poorer outcomes in nCF-Br,11 though the reasons for this are not fully understood. We have found relatively high rates of P. aeruginosa colonization (40%) compared to previously reported data. King et al.18 conducted a longitudinal microbiological follow-up study involving 89 patients, with just 12% isolating P. aeruginosa. Loebinger et al.11 found that 22% of the population of 91 nCF-Br patients were infected with P. aeruginosa, which was also found to be an independent predictor of mortality. A study of the effect of P. aeruginosa on lung function in bronchiectasis (from the same centre) found that in a cohort of 163 patients, 82 (50%) isolated P. aeruginosa but only 14 (9%) were ‘chronically infected’.19 In their cohort, infection with P. aeruginosa was associated with a lower FEV1% predicted but not with accelerated decline. Fatigue, however, was not measured.

In this study, we found those with higher MRCD scores had higher FIS scores, demonstrating a significant relationship between these factors. There was also a significant relationship between FIS and FEV1% predicted, and, between MRCD and FEV1% predicted.

In COPD, systemic effects are proposed as contributing to the observed impairment of quality of life and function, and ‘systemic spill over’ of pulmonary inflammation has been proposed;20 there may be similar effects in bronchiectasis.

Limitations of the present study include the specialist clinic setting from which this consecutive cohort of nCF-Br patients were available for recruitment. This potentially pre-selects a cohort with more severe disease or other complex management issues. We note, however, that the clinic provides secondary care to local patients as well as tertiary care. Although our rates of P. aeruginosa infection are slightly higher than some prior case series but the observed levels of lung function are generally similar.

We did not prospectively undertake health-related quality of life scores or anxiety and depression scores, so it is not possible to comment on the influence these factors may have on the incidence of reported fatigue. This indeed has the potential to be a confounder. Another study from our centre has demonstrated a significant relationship between anxiety and depression (as scored by the Hospital Anxiety and Depression Scale: HADS) and MRCD score. There was, however, no significant relationship between HADS scores and FEV121. Work to evaluate relationships between anxiety and depression scores and fatigue in a single population is important to assess this further. This work is underway but incomplete and has not been included in these data.

Importantly, the effects of exacerbations on fatigue were not studied here; it is possible that an exacerbation phenotype is more closely linked with symptoms of fatigue. Prior data suggest that although hospital admissions were more common in P. aeruginosa-infected patients as compared to those with other pathogens, exacerbation frequency did not differ.

Selected patients with significant fatigue may benefit from systematic management of fatigue. This has been implemented with success in other patient groups where chronic fatigue is a prominent symptom22. Monitoring fatigue as a measure of quality of life in patients with nCF-Br may prove useful in clinical practice and in clinical trials to objectively assess response to interventions.

This work has provided a preliminary evaluation of the FIS score as a potential tool to assess the common and poorly understood symptom of fatigue in patients with nCF-Br. This has potential to be a new objective tool that assesses patients in a different way to existing accepted parameters. We have shown here some of the characteristics and associations of the FIS score when assessing patients with nCF-Br for the first time. Although there are some data on the validity of the FIS score as an endpoint in other conditions,23 this has not been used in a bronchiectasis population before. This tool merits further longitudinal studies assessing relationships with systemic inflammation, co-morbid conditions, measures of depression, exacerbation frequency and indeed its sensitivity to change, in order to facilitate its use to further investigate this clinically important symptom in bronchiectasis.

Acknowledgements

A.D.S is a HEFCE senior lecturer, and K.L.M.H. is a Northumbria Tyne and Wear Comprehensive Local Research Network (CLRN) clinical fellow. We gratefully acknowledge the help and assistance of the patients and Kate Best, Research Assistant in Epidemiology and Statistics, Newcastle University for statistical advice.

Conflict of interest: None declared.

References

View Abstract