Twenty-four-hour fluid administration in emergency department patients with suspected infection: A multicenter, prospective, observational study.

Abstract

To describe 24-hour fluid administration in emergency department (ED) patients with suspected infection. A prospective, multicenter, observational study conducted in three Danish hospitals, January 20 to March 2, 2020. We included consecutive adult ED patients with suspected infection (drawing of blood culture and/or intravenous antibiotic administration within 6hours of admission). Oral and intravenous fluids were registered for 24hours. 24-hour total fluid volume. We used linear regression to investigate patient and disease characteristics' effect on 24-hour fluids and to estimate the proportion of the variance in fluid administration explained by potential predictors. 734 patients had 24-hour fluids available: 387 patients had simple infection, 339 sepsis, eight septic shock. Mean total 24-hour fluid volumes were 3656mL (standard deviation [SD]:1675), 3762mL (SD: 1839), and 6080mL (SD: 3978) for the groups, respectively. Fluid volumes varied markedly. Increasing age (mean difference [MD]: 60-79years: -470mL [95% CI: -789, -150], +80years; -974mL [95% CI: -1307, -640]), do-not-resuscitate orders (MD: -466mL [95% CI: -797, -135]), and preexisting atrial fibrillation (MD: -367mL [95% CI: -661, -72) were associated with less fluid. Systolic blood pressure<100mmHg (MD: 1182mL [95% CI: 820, 1543]), mean arterial pressure<65mmHg (MD: 1317mL [95% CI: 770, 1864]), lactate≥2mmol/L (MD: 655mL [95% CI: 306, 1005]), heart rate>120min (MD: 566 [95% CI: 169, 962]), low (MD: 1963mL [95% CI: 813, 3112]) and high temperature (MD: 489mL [95% CI: 234, 742]), SOFA score>5 (MD: 1005mL [95% CI: 501, 510]), and new-onset atrial fibrillation (MD: 498mL [95% CI: 30, 965]) were associated with more fluid. Clinical variables explained 37% of fluid variation among patients. Patients with simple infection and sepsis received equal fluid volumes. Fluid volumes varied markedly, a variation that was partly explained by clinical characteristics.