Comparing pneumonia incidence in COPD patients treated with or without inhaled corticosteroid

Abstract

<b>Introduction:</b> Prescriber consider the risk and benefit when choosing different maintenance therapies for treatment of chronic obstructive pulmonary disease (COPD). Pneumonia is a common infectious disease in COPD and is also risk associated with inhaled corticosteroid use. However, risk factors for pneumonia in COPD have not been fully elucidated. <b>Aims and objectives:</b> This study compares the incidence of pneumonia in COPD patients, and, associated risk factors, comparing cohorts of patients prescribed with either long-acting muscarinic antagonist (LAMA) or inhaled corticosteroid plus long-acting beta agonist (ICS+LABA). <b>Methods:</b> This nationwide cohort study used Korean National Health Insurance claim data to select patients who received either LAMA or ICS+LABA from Jan 2005 to April 2015, with COPD diagnostic code. We have enrolled patients with good adherence, defined as medication possession ratio ≥ 80%. The observation period for pneumonia was a minimum of 12 months and a maximum of 136 months. <b>Results:</b> The incidence rate of pneumonia per 1,000 person per year was 90.20 in LAMA (n=3,479) and 147.09 in ICS+LABA (n=1,220) with P value &lt; 0.0001. Adjusted hazard ratio (HR) [95% confidence interval (CI)] for pneumonia in ICS+LABA was 1.431 [1.213-1.689] compared with LAMA (P &lt; 0.001). In multivariable analysis, other risk factors for pneumonia included age≥75 (HR [95% CI]: 1.431 [1.235-1.657]), low social economic status (1.464 [1.143-1.875]), previous history of pneumonia (2.400 [1.978-2.911]). <b>Conclusions:</b> The incidence of pneumonia was higher in COPD patients prescribed with ICS+LABA compared with LAMA.