Acute exacerbations of COPD: The Challenge is Early Treatment

Approximately 15 million Americans have COPD, leading to an estimated annual direct medical cost of US $15 billion ([1]). Indirect costs due to morbidity (loss of work time and productivity) and premature mortality were estimated at $32.1 million in 2002 ([1]). But because COPD may be overlooked as a cause of death these costs may be an underestimate ([2]) and the total annual cost of COPD using data from the ‘Confronting COPD’ survey has been estimated to be as high as $70 billion in the USA alone ([2]). The global burden of COPD is huge and by 2020 it is estimated that there will be 3.5 million deaths from COPD worldwide ([3]). Most of the health burden from COPD arises from hospital admissions in late disease ([4]).

Hospitalisation rates have depended on type of study but range from as low as 0.09 to 2.4 per patient per year ([5]). In one American study, length of in-patient stay and number of hospital admissions both decreased over a three year period with a commensurate increase in outpatient visits ([6]). The study also found a 22% decrease in COPD-related total medical costs mainly because of a decrease in the hospital costs for COPD, thus showing that COPD costs can be effectively reduced through decreasing hospitalization ([6]). Previous studies have shown that hospitalisations are related to poorer quality of life scores ([7], [8]) and can be reduced through patient-focused interventions such as integrated care ([9]). Further the interventional arms of the TORCH and UPLIFT studies were associated with a 12 to 18% reduction in the rate of hospitalisations for COPD exacerbations compared to placebo ([10], [11]). Taken together, these data suggest that hospital admission rates are a suitable measure of health burden in COPD and respond to interventions. Thus factors that are associated with an increased risk of hospitalisation are of interest to both health care planners and physicians alike.

One study of physiologic changes occurring during COPD exacerbations suggested that a COPD exacerbation may be considered to have three sequential non-overlapping intervals: prodrome, onset and recovery ([12]). Detection of patients during the prodrome has proved difficult but two subsequent large studies of COPD exacerbations, one British and the other Canadian, have looked at factors related to early reporting of exacerbations in the community. Wilkinson et al from the London COPD Study looked at 1099 and Langsetmo et al, 501 COPD exacerbations. Both studies found that the total number of symptoms at onset of exacerbations and presence of sputum at exacerbation were correlated with earlier reporting, but patients with colds at onset were less likely to report their exacerbations ([13], [14]). Interestingly, the London group also showed that early treatment after onset of exacerbation was associated with a faster recovery from exacerbation and importantly, that patients who had a tendency not to report exacerbations had more hospital admissions for exacerbations ([13]). These data have come from well-studied cohorts in the community but an important question that remains is whether or not the delay in reporting a COPD exacerbation is of importance for hospital admission for that particular exacerbation?

In this issue of The Journal, Chandra and colleagues have gone some way to answer this important question ([15]). These authors have found that, of those exacerbations reported to an emergency department (ED), delay in presentation to the ED after the onset of symptoms increases the risk of hospital admission. The authors found that females, patients with a high body mass index and probable lack of patient familiarity with in-hospital care were more likely to be associated with late presentation at the ED. As with the two outpatient studies above, Chandra et al found that late presenters were more likely to report that a respiratory infection had triggered their exacerbation. This latter observation may be associated with an underestimation of the potential severity of viral infections in COPD. Colds have been associated with prolonged exacerbations ([12]) thus these results of the Chandra study suggest a need for more patient education about the nature and consequences of exacerbation triggers.

We need to know more about the characteristics of these late presenting patients: for example prior to presentation did these patients fail exacerbation therapy in the community? Has a behavior pattern been exposed here, in which late presenters are patients who are more likely to under-report their exacerbations? There is no doubt that certain behaviors in COPD are associated with poorer outcomes ([13]). In the patients in the Chandra study there is some data on treatment prior to the ED visit. Amongst the early presenters about 29% or 12% had already been started on systemic corticosteroids or antibiotics respectively which contrasts with 13% and 7% respectively for late presenters. Further 70% of late presenters came to the ED more than 3 days after onset of their exacerbation. Thus late presenters were less likely to have had prior exacerbation-directed therapy in spite of having symptoms. These data allow us to speculate on a behavior profile for the late presenter as someone who may not have presented to the family physician and who simply tolerated the symptoms. Chandra et al did not assess anxiety and depression scores but these may also affect outcomes in COPD ([16]). Taking these data together with those of the London group we can speculate further that perhaps late presenters may have a pattern of under-reporting their exacerbations to their family physicians and are more likely to end up as in-patients though there is evidence that American patients may differ from Canadian (and so perhaps British) patients in their use of the ED ([17]).

There are some limitations, which the authors have attempted to address. How do we diagnose COPD in the ED? The current GOLD guidelines require a spirometric diagnosis but these authors took a self-reported past history of COPD ([18]). This is a very real problem that we as clinicians face because of the wide differential diagnosis of a COPD exacerbation. The authors point out that it is difficult to validate a diagnosis of COPD in the ED with spirometry because of the acute nature of the illness. However they have previously studied self-reported COPD in the ED in 60 consecutive such patients and have validated this approach ([19]).

Another important factor is how to distinguish between pneumonia and COPD exacerbations. The TORCH and INSPIRE studies have had similar problems in differentiating reported pneumonia events many of which did not have radiological confirmation ([10], [20]). The infectious disease society of America requires “a demonstrable infiltrate by chest radiograph or other imaging technique” for diagnosis of community acquired pneumonia but this is not a necessary requirement according to the British Thoracic Society guidelines ([21], [22]). In the Chandra study it appears that chest radiographs were done on all patients but infiltrates were found in about 8% patients who were diagnosed by ED physicians with a COPD exacerbation with an apparent concomitant diagnosis of pneumonia. However it is to be noted that these patients all met the pre-assigned case definition of a COPD exacerbation and this illustrates one of the problems we have as practitioners with a definition of a COPD exacerbation that is based only on symptoms ([23], [24]). Thus the results of the Chandra study cannot be refuted on these grounds.

Another issue addressed by the authors was that of repeat visits after the same exacerbation and these were excluded from the analysis. The readmission rate in the 2003 UK National COPD audit was 31.4% and readmission has been related to mortality in COPD ([25], [26]). It is now recognized that exacerbations may cluster in time and a significant number of exacerbations are associated with a recurrent or second exacerbation within 8 weeks of the first event ([27]) We hope that the study by Chandra and colleagues will be followed by one looking precisely at those patients who have either relapsed or developed a recurrent exacerbation as we may discern more about the phenotype of the late presenter and importantly implications for mortality.

Finally, what does this study tell us about where the COPD patients with an acute exacerbation should go? While the Chandra study was conducted in the ED we should not conclude that the ED should be the only site of treatment for COPD exacerbations. About ten years ago, the Respiratory Network Group in Québec conducted a randomized clinical trial of COPD self management in the community ([28]). In the Quebec study, module 3 of the methodology called for understanding and using a plan of action for acute exacerbations which included self treatment along fixed guidelines with antibiotics and steroids in the community. The study showed that self management was associated with a 40% reduction in hospital admissions and 41% reduction in ED visits ([28]). Thus taking this study together with the Chandra and London Studies, we conclude that early intervention is desirable in COPD exacerbations. One way to encourage earlier intervention is with self-management and home monitoring strategies, though these need more evaluation in the COPD patient group, particularly those with complex comorbidity. In the future we need to ensure that the appropriate patients can access hospitals when they really need them in a timely manner. This is particularly important for patients with COPD exacerbations who also have respiratory failure or complicated comorbidity.