Predictors of mortality among sepsis patients transferred from a rural, low-volume ED to an urban, high-volume hospital.

Abstract

Introduction: We investigated the extent to which demographic characteristics, clinical care aspects, and relevant biomarkers predicted sepsis-related mortality among patients transferred from a rural, low-volume emergency department (ED) to an urban, high-volume, level-2 trauma center.

Methods: We conducted an observational study among adult severe sepsis patients (N = 242) who, within a community-based regional healthcare system, presented to one of the four rural, low-volume EDs and were subsequently transferred to the urban, high-volume, level-2 trauma center, and were identified as septic at either location. We evaluated in-hospital and 30 days after discharge mortality.

Results: In-hospital mortality rate was predicted by previous admission to an ICU (OR 5.02, 95 % CI: 1.89-15.94, p = 0.002), identification of sepsis prior to transfer (OR 0.29, 95 % CI: 0.11-0.74, p = 0.01), and a moderately abnormal lactate level (OR 0.22, 95 % CI: 0.05-0.79, p = 0.03). Mortality 30 days after discharge was predicted by previous admission to an ICU (OR: 3.28, 95 % CI: 1.62-6.97, p = 0.001), abnormal red cell distribution width (OR: 2.23, 95 % CI: 1.11-4.60, p = 0.03), identification of sepsis prior to transfer (OR: 0.26, 95 % CI: 0.12-0.54, p < 0.001), and a moderately abnormal lactate (OR: 0.32, 95 % CI: 0.12-0.79, p = 0.02).

Discussion: Early identification of sepsis, as well as attention to prior ICU admission or comorbidities and abnormal red cell distribution width, could facilitate better care and prevent mortality among patients with sepsis who are transferred from a rural, low-volume emergency department to an urban-high volume facility.