• chronic obstructive pulmonary disease
• hospital admission
• physical activity

A hospital admission or readmission for an exacerbation of chronic obstructive pulmonary disease (COPD) is bad news for everyone. For the patient it may signal the beginning of the terminal phase of the illness. For the health services it is a significant component of the cost of care for a condition that is increasingly burdensome. In recent years knowledge has been growing about the important role of the exacerbation in the deteriorating progress of COPD. The last few years of life with COPD may be characterised by repeated episodes of illness culminating in hospital admission. Exacerbations only lead to hospital admission in about 16% of cases, but their increasing frequency is associated with declining state of health.¹

In UK hospitals emergency inpatient admission is the largest component of the total cost of respiratory disease to the NHS that amounts to over £2.5 billion.² In recent years attention has focused on reducing the cost and impact of hospital admissions for COPD. The emphasis of the strategy to date has been on admission avoidance and early supported discharge schemes.^3–5 These have had some success in curtailing admissions and reducing lengths of hospital stay, and are beginning to be introduced more widely into clinical practice. Contrary to popular perception, COPD is not a stable condition and there are inevitably day to day variations in both symptoms and ability to function. The periodic episodes of worsening are known as exacerbations, although a precise agreed definition is lacking. However, the meaning of the term is generally understood and a working appreciation of the impact of these events is becoming clear. Not all episodes of exacerbation lead to hospital admission, but if these can be contained or prevented by the patient or primary care services, then the burden of COPD will be reduced. A study of the factors that can prevent hospital admission in these circumstances would therefore make an important and valuable contribution.

This issue of Thorax contains the third major publication from a group in Barcelona (EFRAM) that has been examining the factors associated with hospital admission.⁶ This latest paper is the report of a prospective study that examines the risk factors for readmission to hospital for exacerbations of COPD. In previous publications the group have described the prevalence and relative risk of modifiable risk factors for hospital admission.^7,8 The initial study population was a sample of 404 patients who were admitted to hospital in Barcelona over a period of 1 year.⁷ During the admission the patients completed a questionnaire and 353 later underwent spirometric tests and arterial blood gas measurements. Identification of risk factors that were potentially amenable to modification showed that lack of rehabilitation and poor inhaler technique were the most frequent associations. Other important factors were continued smoking, inadequate oxygen prescription, and lack of influenza immunisation. In the second paper the authors reported a case-control study of the factors associated with hospital admission in a subsample of the original cohort.⁸ The control subjects were patients with COPD who had had a previous admission but were stable at the time of the comparison. The only apparent risk factors found in 86 pairs of cases and controls were a previous history of three or more admissions, lower FEV₁, and underprescription of oxygen. Perversely, continued smoking appeared to convey an advantage and inhaled corticosteroids offered no benefit. Some other potentially modifiable factors including rehabilitation could not be tested because of the very small numbers receiving it. In their conclusion the authors acknowledged that the case-control methodology had limitations in terms of selection bias and small numbers and that a prospective cohort study was required.

The latest paper describes a prospective study of the risk of readmission in a cohort of 340 patients from the original study population who were followed for over 1 year after their index admission.⁶ At the end of the study 63% of the patients had been readmitted and 29% had died, suggesting that previous admission is an important risk factor. Other expected risk factors were lower FEV₁ and hypoxaemia. One unexpected finding was that a high level of usual physical activity was associated with a 46% reduction in the risk of admission. The activity profile was obtained through self-reporting and no objective test of exercise capacity was made. Nevertheless, the association was very strong. Once again the authors could not test the influence of rehabilitation because the numbers were too small. Other factors weakly associated with increased admission were supervision by a respiratory specialist, oral corticosteroids, and anticholinergic drugs. These latter associations are unlikely to be causal and probably reflect confounding by severity of disease.

Similar findings were reported by the British Thoracic Society and Royal College of Physicians’ audit of admission for acute exacerbation of COPD.⁹ In this audit of 1400 admissions two thirds of the patients had had a previous admission for COPD and one third had had a similar episode in the previous 4 months. In a follow up audit of readmission within 3 months, poor performance status was a predictor of mortality at the first admission but not readmission.¹⁰ Once again the best predictors of admission were low FEV₁ and previous admission.

The pattern of hospital readmission may be affected by both patient and healthcare delivery factors. It could be that general practitioners find it easier to admit the patient to hospital than to deal with the exacerbation at home. However, high rates of readmission to hospital are common to both studies from different healthcare systems. The implication of the latest EFRAM study is that the risk of readmission to hospital for COPD can be reduced by improving spontaneous domestic activity and thereby breaking the cycle of hospital dependency. This may be true but, so far, trials of pulmonary rehabilitation that improve exercise capacity have been unable to show a reduction in hospital admission. However, they have shown a reduction in the length of stay once admitted.¹¹ It is possible that rehabilitation may reduce hospital admissions but studies with an appropriately sensitive design have not been performed. In addition, the capacity to provide rehabilitation in most countries is so poor that it has never been testable. In the British Thoracic Society audit only 3% of the patients were recommended for rehabilitation, while in Spain only 14% of the study group had received it. An alternative explanation for the results of the study is that patients with less severe COPD simply feel better, do more, cope better, and are not admitted so frequently. It is known that physical performance, as reflected by a walking test and functional performance questionnaire, is a strong predictor of outcome in COPD.¹² The answer to the important question of whether the prognosis can be altered by increasing functional performance is not known. The arguments are somewhat analogous to those about the presence of nutritional depletion in severe COPD, where a low body mass index is associated with increased mortality but attempts to change the situation by supplementation have been ineffective.^13,14 In the case of physical function, however, it is clear that the simple intervention of exercise training will improve performance and have a prolonged effect on lifestyle.¹⁵ It would therefore be reasonable to test the hypothesis that rehabilitation can reduce hospital readmission in an appropriately susceptible group.

In the UK a hospital admission for COPD generally involves a length of stay of 8 days.⁹ From the patient’s perspective, there is a flurry of attention at the beginning of the admission followed by 7 days of observed inactivity. As result, patients may leave hospital less well equipped for independent life than when they were admitted. Characteristically, our hospital and community services are attuned to dealing with one crisis but make little attempt to prevent the next. The huge financial costs of hospital admissions for exacerbations of COPD deserve some exploration of the value of actions that may prevent the initial admission or reduce the frequency of readmission. The cost of drug treatment for these patients is second only to the cost of hospitalisation, but to adhere to the 20th century view that drug treatment alone will provide all the answers is a delusion. The diverse factors associated with advanced COPD require a multi-modality approach by a multi-professional team. There is now increasing evidence that non-pharmacological interventions may play a major role. Let us hope that this evidence is now persuasive enough to promote appropriate investment in research and services for this neglected condition.