Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

COPD and treatment

Exacerbation and hospital admission
Health economics and perspective
The capacity question

Bandolier thought the diagnostic algorithm for chronic obstructive pulmonary disease (COPD, or chronic obstructive airways disease COAD) determined through the CARE process was wonderful ( Bandolier 78 ), though the limitations of treatment sobering. Smoking more than 40 pack-years was the biggest predictor for COPD. A fifth of long-term smokers have a reduced forced expiratory volume over one second (FEV1), defined as more than two standard deviations below that for a normal population of that age. There follows increasing dyspnoea and other respiratory symptoms, with progressive deterioration of health status.


In nonsmokers the FEV1 falls with age, so that by 75 years it may be 75% or so of the value obtained at 25 years, but in some smokers it can fall dramatically faster. Stopping smoking reduces the speed of decline. Otherwise treatments are palliative, with inhaled anticholinergic drugs as the mainstay. Two recent publications of the same review are authoritative regarding the evidence available on the treatment of acute exacerbations [1,2], including diagnostic tests and prognostic factors.

About 5-15% of adults in industrialized countries have COPD defined by spirometry. In 1990, COPD was the twelfth most common cause of combined mortality and disability but is expected to become the fifth cause by the year 2020. After diagnosis the 10-year survival rate is approximately 50% with more than one-third of patients dying due to respiratory insufficiency.

There are about 16 million patients with COPD in the USA, and the prevalence in adults between 40 and 70 years is about 10%. A UK primary care organisation of 100,000 would have, on average, about 40,000 people over 40 years, with about 4,000 patients with COPD on their books. In a proportion it would be moderate or severe.

The problems when COPD is moderate or severe are several fold. First is a reduced health status, inability to work, and increasing dependence. Then there are the exacerbations of breathlessness and sputum production. Finally, of course there is hospital admission.

Exacerbation and hospital admission

A large cross-sectional observational study on ambulatory COPD patients in primary care sought to find predictors of exacerbations and hospital admission in 201 general practices in Spain. It used sensible definitions for diagnosis, and exacerbation. Information was collected on 1,001 patients, and was detailed, and included spirometry. Randomly selected patients were used to test and validate models for predicting exacerbation and hospital admission.

Two or more exacerbations a year were suffered by 55%, and 22% had a hospital admission. Two or more exacerbations occurred more often in older patients, those with chronic mucus hypersecretion and those with more severe FEV1 impairment. At least one hospital admission occurred more frequently in older patients, those with more comorbid illnesses and with more severe FEV1 impairment.

There are limits to the utility of this information except that it hints at the sorts of patients more likely to be in trouble, and probably does no more than confirm clinical feel in primary care. It also shows the importance of the surrogate measure of FEV1. In other studies, like one looking at five-year survival after discharge with first admission with COPD [4], the crude five-year mortality rate was 45%. It increased with older age and lower FEV1.


A new potential treatment heaves into view with publication of two randomised trials [5,6] from each side of the Atlantic (the way the FDA demands) plus a thoughtful editorial [7]. The interesting thing is that these studies were large and had a one-year duration, enough to investigate the effect on exacerbations and hospital admission.


These had similar designs, being randomised and double blind, patients having a smoking history of 10 pack years or more, an FEV1 of 65% or less of predicted for age, and were at least 40 years of age. Asthma, allergic rhinitis, atopy or elevated eosinophil counts were exclusions, as were patients needing regular supplementary oxygen, recent respiratory tract infection, or with significant comorbid conditions. They received tiotropium 18 μg each morning, or placebo (USA) or ipratropium 40 μg four times a day. The duration of the studies was one year.


Patients were well matched at baseline in each trial, were mostly male, and had an average age of about 65 years. There were a number of outcomes. The main results (all with statistical significance between tiotropium and control for efficacy measures) are shown in Table 1.

Table 1: Main results in two randomised trials of tiotropium in the USA and in Europe

US Study

European study





Quality score (out of 5) 4 4
Number of patients 550 371 356 179


Baseline FEV1 (L/one second) 1.04 1.00 1.25 1.18
Change in FEV1 (before dose, L) +0.12 -0.03 +0.12 -0.03


Clinically important symptoms improvement (% in five assessments) 42-47 29-34 31 18


At least one exacerbation (%) 36 42 35 46
Exacerbations per patient per year 0.76 0.95 0.73 0.96

Hospital admissions

Hospital admission (%) 5.5 9.4 7.3 11.7
Hospital admission per patient per year 0.09 0.16 0.10 0.16
Hospital days per patient per year 0.6 1.2 1.42 2.13

Adverse events

All discontinuations (%) 18.7 27.8 15.2 21.2
Adverse event discontinuations (%) 9.6 13.7 10.1 12.8
Lack of efficacy discontinuations (%) 2.4 7.0 0.8 1.7

Tiotropium improved FEV1 (measured before dosing at the end of the year) and the proportion of patients with clinically significant improvement in symptom scores. In the European study a clinically meaningful improvement of four units in the St George's Respiratory Questionnaire was found in 52% of patients on tiotropium and 35% of those on ipratropium (an NNT of about 6).

Exacerbations were reduced with tiotropium, as were hospital admissions. Combining the two studies because the results with placebo and ipratropium for these outcomes were about the same, we can calculate NNTs. For exacerbations, 323/906 (36%) of patients treated with tiotropium had at least one exacerbation, compared with 238/550 (43%) with control. The NNT to prevent one patient having at least one exacerbation in a year was 13 (95% CI 8-41). At least one hospital admission occurred in 56/906 (6%) of patients treated with tiotropium compared with 56/550 (10%) with control. The NNT to prevent one patient having at least one admission in a year was 25 (14-97).

Only about 10% of patients had a hospital admission with the control treatments, but the average number of hospital days for each patient was 1-2. That means that when a patient with COPD has a hospital admission it is for at least 10 days. Tiotropium saved at least 0.6 of a hospital day a year for each patient given tiotropium. The most common adverse event with tiotropium was dry mouth. All discontinuations, and adverse event discontinuations, were lower with tiotropium than with control.


These results hang together. There is good evidence that FEV1 is an important clinical measure, and predicts exacerbation and hospital admission for patients with COPD in primary care, as well as mortality when admitted. Improvement in FEV1 should logically then result in lower rates of exacerbation and hospital admission. This they do with tiotropium in large, high quality, long-duration trials.

So what we have is good evidence for clinical efficacy of tiotropium at various levels, from respiratory physiology to important healthcare outcomes. The question then is how important are these results for healthcare services and delivery.

Health economics and perspective

The first place to go for sensible thoughts about COPD is the British Thoracic Society guidelines [8]. As well as having useful thoughts on management of the disorder, it also tells us much about the impact of COPD on emergency services. For instance, about 12% of all medical emergency admissions in the UK are due to COPD (or were in the early '90s). Using some of the figures for annual GP consultations with COPD, we can conclude that an average primary care organisation (PCO) of 100,000 people would have about 2,600 consultations for COPD (Table 2). That's half to one GP, though other estimates for GP consultations are much higher, at about three per patient per year [9].

Table 2: COPD consultations in a PCO



Rates per 10,000

COPD consultations per PCO

45-64 19000 417 792
65-74 12000 886 1063
>75 7000 1032 722




An estimate of the burden of COPD in the UK based on data from the early '90s indicated the cost to the NHS of managing COPD was over £800 million [9]. A detailed breakdown of those costs is in Table 3. The estimate is likely to be out of date, since emergency admissions for COPD have risen by 50% since the mid-90s, to about 100,000 a year in 2000. That's about 200 emergency admissions for COPD for each PCO, at a cost of about £3000 [9], or about £600,000 per PCO.

Table 3: Estimated costs of COPD care in the UK in the early 1990s


Cost (£ million)

Inpatient stay 243
GP consultations 237
Outpatient clinic 35
Prescribing 126
Diagnostic tests 144
Other 20



Another critical issue is the consumption of bed-days in hospital. An early UK estimate is that COPD patients consumed 1.2 million bed days a year in the early 1990s. Table 4 shows that estimate to more or less agree with the results of the tiotropium clinical trials, with some back of envelope assumptions about prevalence of COPD and the proportion of COPD patients who have moderate or severe disease. It suggests that translating the results to a UK perspective could save the equivalent of 528,000 bed days, the equivalent of four 400 bed hospitals. Whether the intervention is cost effective would depend on a thorough cost-effectiveness analysis, and that includes the crucially important issue of acquisition cost, something we don't yet have as this intervention is not yet licensed.

Table 4: Some preliminary health economic calculations for COPD admissions in the UK



Per 100,000

Per million

Per 55 million

Number of patients with COPD 10% prevalence in over 40s 4,000 40,000 2,200,000
Moderate or severe COPD, FEV1 <70% predicted Assume 40% of all COPD 1,600 16,000 880,000
Number of hospital days 1.7 days per patient per year, mean of US and European studies 2,720 27,200 1,496,000
Number of bed days saved with tiotropium Average saving 0.6 days per patient per year 960 9,600 528,000
Cost saving (£) possible on hospital admission Assume £200 per day 192,000 1,920,000 105,600,000

The capacity question

These simple calculations take no account of how new technology can impact on the perennial problem of health service capacity. When beds, doctors and nurses are at a premium because we don't have enough, we should value interventions like this more highly.

One thing we do know, and that is deaths from respiratory diseases are rising (Figure 1). COPD will be a bigger problem with population ageing. More and better health economic studies will enable us to plan services better.

Figure 1: Death rates in England and Wales


  1. DC McCrory et al. Management of acute exacerbations of COPD. Chest 2001 119: 1190-1209.
  2. PB Bach et al. Management of acute exacerbations of chronic obstructive pulmonary disease: a summary and appraisal of published literature. Annals of Internal Medicine 2001 134: 600-620.
  3. M Miravitelles et al. Factors associated with increased risk of exacerbation and hospital admission in a cohort of ambulatory COPD patients: a multiple logistic regression analysis. Respiration 2000 67: 495-501.
  4. J Vestbo et al. Vital prognosis after hospitalisation for COPD: a study of a random population sample. Respiratory Medicine 1998 92: 772-776.
  5. R Casaburi et al. A long-term evaluation of once daily inhaled tiotropium in chronic obstructive pulmonary disease. Eur Respir J 2002 19: 217-224.
  6. V Vincken et al. Improved health outcomes in patients with COPD during 1 yr's treatment with tiotropium. Eur Respir J 2002 19: 209-217.
  7. PJ Rees. Tiotropium in the management of chronic obstructive pulmonary disease. Eur Respir J 2002 19: 205-206.
  8. Anon. BTS guidelines for the management of chronic obstructive pulmonary disease. Thorax 1997 52: Suppl 5.
  9. JF Guest. The annual cost of chronic obstructive pulmonary disease to the UK's National Health Service. Dis Manage Health Outcomes 1999 5: 93-100.
previous or next story in this issue