Lydia Finney1, Sarah Elkin2, Samuel Todd2, Vijay Padmanaban2 and
and Lung Institute, Imperial College, London, United Kingdom,
2Respiratory Medicine, Imperial College Health Care NHS Trust, London,
is associated with increased morbidity and mortality in COPD. We aimed
to characterise the infectious aetiology and radiology of pneumonia in
COPD and identify potential risk factors.
Retrospective analysis of medical records for patients with COPD
(FEV1/FVC <0.70) admitted to 3 London hospitals between
01/01/2010 and 31/12/2012 with pneumonia or exacerbation (AECOPD).
Patients were identified by COPD bundles. Admission CXRs were assessed
by 2 independent observers. Pneumonia was defined as new consolidation
There were 235 pneumonia admissions and 706 for AECOPD. 43% of patients
with radiological diagnosis of pneumonia had a diagnosis of pneumonia
on their discharge summary. Patients admitted with pneumonia were more
likely to have had a previous admission with pneumonia in the last year
(21% v 6% p<0.01, RR 3.5) and be age ≥ 75 years (50% vs
36% p=0.01, RR 1.58). There was no difference between pneumonia and
AECOPD for mean FEV1 % predicted (40 +/- 18 v 41 +/- 19) or proportion
of patients using inhaled corticosteroids (79% v 83%).
Gram negative organisms were identified in 21% of pneumonias and 13% of
exacerbations. The most commonly isolated organism was Pseudomonas
aeruginosa (7% of pneumonia, 5% of AECOPD). 44% of pneumonias were
lobar, 42% multifocal and 13% bibasal.
Incidence of pneumonia in COPD is likely to be underestimated. The most
commonly isolated bacteria were gram negatives which may not respond to
current antibiotic guidelines for community acquired pneumonia. Risk
factors for pneumonia are previous history of pneumonia and age≥