Hospitalization Predictors in Acute Exacerbation of Chronic Obstructive Pulmonary Disease: A Post Hoc Study of a Multicentric Retrospective Analysis

Abstract

Background/Objectives: Hospitalizations for acute exacerbations in patients with chronic obstructive pulmonary disease (AECOPD) are connected with considerable mortality and morbidity and have a great impact on healthcare costs. We aimed to identify potentially important characteristics that distinguish patients with AECOPD that required hospitalization between those that did not. Methods: We performed a post hoc study of a previously conducted observational study assessing the discharge records of AECOPD patients who were either hospitalized or discharged directly from the emergency department (ED). Results: A total of N = 593 AECOPD patients (37.4% female) were included: N = 225 (37.9%) were hospitalized, while 368 (62.1%) were discharged from the ED. Patients had a mean age of 71 (±9.14) years. Further, 64.7% had arterial hypertension, and 60.4% of hospitalized and 42.1% of discharged patients had cardiovascular disease (excluding hypertension) (p < 0.001). In addition, 24% of hospitalized patients suffered from diabetes mellitus (vs. 16.8% of those discharged, p = 0.042). Patients that were discharged had a higher median eosinophil count than hospitalized patients (p < 0.001). Among the 368 patients discharged from the ED, 23.6% visited the ED due to AECOPD, and 50.6% were hospitalized in the subsequent three months. Patients that had at least one AECOPD in the subsequent three months had significantly lower initial eosinophil counts than those that did not (p = 0.015). Conclusions: Hospitalized AECOPD patients had a higher prevalence of pre-existing cardiovascular disease, diabetes mellitus and lower blood eosinophil counts. Patients that had subsequently visited ED in the following three months initially had lower blood eosinophil counts than those that did not make subsequent visits.

Keywords: COPD; acute exacerbations of COPD; discharge record; post hoc analysis; predictors.

Figure 1

Flowchart depicting main study groups. Primary analysis consisted of patients that underwent emergency department (ED) evaluation for AECOPD, which were then either discharged home from the ED or admitted to the hospital. Follow-up analysis was performed among patients that were discharged (in the primary analysis) and were again evaluated in the ED for AECOPD in subsequent 30 days and in subsequent 3 months from initial ED discharge.