© 2008 American Thoracic Society
Inhaled Corticosteroids and Pneumonia in COPD: An Association Looking for EvidenceTo the Editor:A recent database study by Dr. Ernst and colleagues (1) identified an association between inhaled corticosteroid (ICS) prescriptions and increased pneumonia hospitalization and 30-day mortality in patients with chronic obstructive pulmonary disease (COPD) between 1988 and 2003 from data obtained from the Quebec health insurance registry. Pneumonia is a common condition and may affect patients with advanced COPD. Two recent prospective studies with primary outcomes of mortality (2) and exacerbations of COPD (3) have reported increased lower respiratory tract infections (LRTIs) in subjects receiving inhaled steroid preparations. Interestingly, the increase in LRTIs (although case definitions and treatments were undefined in these two studies) did not increase mortality or COPD exacerbations. Although there may be an association between ICS and LRTIs, especially with higher ICS doses, there is always a concern about biases that may not be accounted for in nonexperimental studies. Until there is a prospective study with detailed clinical and imaging data, there may continue to be parallels with the many studies that investigated associations between short-acting β-agonists and severe life-threatening and fatal asthma. Confounding by severity and by indication were recurrent issues that needed clarification in those studies (4, 5). It would be interesting to know if the association between pneumonia and ICS in the Quebec study (1) is independent of other medications that may be used by patients with severe symptomatic COPD, such as ipratropium bromide and oral theophyllines. Defining the pattern of ICS use more precisely during the preceding year may also help define risk. The same authors performed a very elegant analysis of patterns of β-agonist use and impact on fatal and near-fatal asthma by using a profile score, which if applied to the current study would not only allow categorization of patients as constant, increasing, or decreasing users of ICS but also quantify the temporal patterns of utilization (6). Another question raised by the current study and examined carefully in previous studies of β-agonists and fatal and near-fatal asthma is whether the ICS association with pneumonia is a class effect or whether certain ICS increase risk. With the large numbers of patients in the current study, these data may be available and helpful for investigators planning studies of pneumonia in COPD. Until better data are available, clinicians and patients alike await future studies that may confirm or refute the postulate that ICS predispose the diseased airways and lungs of patients with COPD to LRTIs.
University of British Columbia FOOTNOTES Conflict of Interest Statement: M.O.T. has received a lecture fee from Bayer and served on an advisory board for GlaxoSmithKline. REFERENCES
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
