Risk factors for hospital admission with a respiratory illness in a comparative effectiveness cohort of patients with chronic obstructive pulmonary disease (COPD).
Veljanovski J, and Ouellette DR. Risk factors for hospital admission with a respiratory illness in a comparative effectiveness cohort of patients with chronic obstructive pulmonary disease (COPD). Am J Respir Crit Care Med 2017; 195.
Am J Respir Crit Care Med
COPD is the 3rd leading cause of death in the United States per year. It affects over 6.3% of Americans and is associated with high resource allocation including multiple hospitalizations, readmissions and close outpatient follow up. Awareness of clinical risk factors that are predictive of hospital admissions is important for clinicians as well as policy makers. Although many therapies exist for the management of COPD, most have been studied only in placebo-controlled efficacy trials in highly selected populations. Comparative Effectiveness Research (CER) allows for comparisons in a real world environment. This study uses a CER cohort of COPD patients from an academic internal medicine clinic to identify clinical risk factors for respiratory hospital admissions (RHA), with a secondary aim of establishing the framework for a future COPD CER database. Methods This single center retrospective observational study analyzed 242 patients who had both a clinical diagnosis of COPD during an ambulatory internal medicine encounter in 2014 as well as pulmonary function tests performed in 2013-14. Data was obtained from an institutional pulmonary function test database and the electronic medical record. GOLD criteria was applied to define obstructive versus non-obstructive disease. The need for informed consent was waived by an institutional IRB for this study. Univariate analysis of associations between potential risk factors and the outcome variable of RHA was performed by Chi-square analysis for dichotomous variables, and the student's t-test for continuous variables, with a p value <0.05 being considered significant. Risk factors significant in the univariable analysis were included in a multivariable analysis model. Statistics were performed using SPSS 20 (IBM Corp., Armonk, NY). Results The cohort was comprised of 242 patients; 76 % with obstruction and 24% without obstruction. Refer to Table 1 where a univariate analysis identified associations of the following medications with RHA; LAMA, SAMA, LABA, beta adrenergic agents, inhaled corticosteroids, combination inhaler SABA+SAMA, combination inhaler LABA+ICS, macrolides and PDE-4 inhibitors. Interestingly, no pulmonary function variables were found to be predictive of RHA including the presence of obstructive disease. Three medication classes remained significantly associated with RHA in a multivariable analysis; SAMA, combination inhaler SABA+SAMA and combination inhaler LABA+ICS. Conclusion Three medication classes were found to be associated with RHA; SAMA, combined SABA+SAMA and combined LABA+ICS. These findings were consistent with both obstructive and non-obstructive patients. Our secondary aim of establishing a foundation for future COPD CER was also achieved.