PURPOSE: Antibiotic use during an acute exacerbation of chronic obstructive pulmonary disease (AECOPD) remains controversial. Limited data is available about clinical practice of antibiotic treatment of hospitalized patients with AECOPD and its prognostic implications.
METHODS: This was a retrospective study of all admissions to a single tertiary respiratory centre in Slovenia from 2002 to 2007. Medical charts were reviewed for presence of AECOPD, sputum sample and in-hospital management. Vital status was obtained from Central Population Registry.
RESULTS: We included 679 patients (age 70±9 years, 75% men, 65% GOLD III/IV), of which 560 (80%) received antibiotics. Antibiotic group had a more advanced COPD (GOLD III/IV 69.1 vs. 57.2%, p<0.01), higher frequency of long-term oxygen treatment (LTOT) (22.9% vs. 13.7%, p<0.05), higher value of admission C-reactive protein (32(11.3-80.4) vs. 13.4(5.4-32.8), p<0.001). No difference was detected in the rate of positive sputum culture (54 % vs. 45%), intensive care admission or admission pH. Single agent was prescribed in 85% of patients (amoxicillin-clavulanate 54%, moxifloxacin 20%, ciprofloxacin 7%). 15% received several antibiotics (ciprofloxacin and ceftazidime 9%). Sputum microbiology showed that prescribed antibiotics were inappropriate in 45 patients (13%). In a multivariate model adjusted for age, GOLD, LTOT, intensive care admission and cardiovascular disease (CVD), antibiotic use was associated with length of stay (LOS) >14days (OR 1.91; 95% CI 1.09-3.32). During 6 months of follow-up a composite event of death or rehospitalisation for AECOPD occurred in 227 (30%) patients. Antibiotic use didn't influence the occurrence of composite event in univariate Cox regression (HR 1.37; CI 0.95-1.97) nor in multivariate model (HR 1.20; CI 0.83-1.74), adjusted for age, GOLD, LTOT, intensive care admission, LOS and CVD.
CONCLUSIONS: Antibiotic use was frequent and associated with prolonged length of stay. Their use didn't improve clinical outcomes during 6 months of follow-up.
CLINICAL IMPLICATIONS: To avoid widespread use of broad spectrum antibiotics with no clear benefit on clinical outcomes in AECOPD, further studies are needed to better define their appropriate use.
DISCLOSURE: The following authors have nothing to disclose: Ursa Bones, Irena Sarc, Kristina Ziherl, Miha Zabret, Tina Jeric, Mitja Kosnik, Stanislav Suskovic, Mitja Lainscak
No Product/Research Disclosure Information