Epidemiology of Patients with Ventricular Assist Devices Presenting to the Emergency Room from 2006-2014

Abstract

Data related to the epidemiology and resource utilization of ventricular assist device (VAD) related emergency department (ED) visits are limited. However, an improved understanding of the burden of VADs on the healthcare system is critical to designing interventions that improve outcomes. We examined national estimates of VAD-related ED visits and described medical complexity, admissions, resource utilization, and mortality. This study utilizes data from ED encounter-level data via the 2006-2014 Nationwide Emergency Department Sample (NEDS). ICD-9 codes were used to identify patient-encounters with VADs. Demographic and clinical factors are reported via descriptive statistics. The primary outcome was death; secondary outcomes included median inflation-adjusted charge and hospital admission. From 2006 to 2014, over 900 million ED visits were evaluated, of which 17,356 (.002%) VAD-related ED visits were identified, for a median of 1,028 (SD 1,489) visits per year. Most VAD ED encounters were represented by patients that were male (74%), ≥ 45 years old (73%), and living in an urban environment (85%). 73% of VAD related ED visits resulted in inpatient admission and the median inflation-adjusted charge was $23,862 (IQR 7,129; 58,265) per visit. Mortality, either in the ED or during an associated admission, was 4.7%. The most common primary diagnoses were cardiac (32%) followed by bleeding (25%) and infection (18%). More than half of the encounters were with patients with ≥1 chronic medical condition, with diabetes (34%) and hypertension (31%) the most prevalent. This is the first study to use national level data to describe the clinical characteristics and outcomes of patients with VADs who present to the ED. With nearly 1 in 20 VAD ED encounters resulting in death, strategies to reduce mortality are urgently needed.