Myers, M.

Abstract Background: Intravenous fluid resuscitation is fundamental for sepsis-induced hypoperfusion. The 2021 Surviving Sepsis Campaign recommends 30 mL/kg of intravenous crystalloid within the first three hours. Fluid management in patients with sepsis is challenging due to risk for volume overload hence edema in the lungs and other organs. Conflicting literature debates over whether liberal or restricted fluid administration is optimal in this population. We hypothesized that, despite X-ray findings suggesting volume overload, liberal fluid resuscitation in patients with severe sepsis or septic shock does not increase in-hospital mortality or ventilation requirements. Methods: The study is approved by the Mayo Clinic IRB. We retrospectively studied adult patients admitted to the medical ICU with severe sepsis or septic shock from 2014-2018. Data collected included admission weight, chest Xray (CXR) fluid status (“wet” vs. “dry”), fluid resuscitation volumes at 3 and 24 hours (liberal [≥30 mL/kg] vs. restricted), in-hospital mortality, ventilation need, ICU/hospital length of stay (LOS), admission lactate, baseline NT-proBNP, home oxygen use, and history of heart failure or chronic pulmonary disease. For accuracy, an independent pulmonologist reviewed CXRs, and patients with radiologist-pulmonologist discrepancies were excluded. Results: Of 346 patients, 68% had “wet” CXRs. Patients with “dry” CXRs received greater median fluid resuscitation at both 3 hours (3913 mL vs. 3126 mL, p=0.001) and 24 hours (6085 mL vs. 5435 mL, p=0.0005). Liberal fluid resuscitation was more frequent in patients with “dry” CXR than those with “wet” CXR (72% vs. 57%, p=0.006). In-hospital mortality was lower among patients receiving liberal resuscitation compared to restricted (13% vs. 21%; OR 0.573, 95% CI 0.32-1.02, p=0.04), although in patients with “wet” CXRs, liberal resuscitation did not significantly impact mortality (p=0.17). Need for invasive or noninvasive ventilation tended to be higher in the restricted group compared to the liberal group, though this was not statistically significant (45% vs. 29%; OR 0.48, 95% CI 0.21-1.12, p=0.07). There was no significant difference in ICU or hospital LOS between patients receiving liberal and restricted resuscitation (p=0.14 and p=0.9, respectively). In multivariate analysis, in-hospital mortality was significantly associated with chronic pulmonary disease (p=0.04) and admission lactate (p=0.007) but not with CXR findings (p=0.72) or fluid resuscitation volume (p=0.38). Additionally, neither CXR findings (p=1) nor liberal fluid resuscitation (p=0.07) significantly increased the need for ventilation. Conclusion: Liberal fluid resuscitation in patients with severe sepsis or septic shock appears safe and does not increase mortality or ventilation needs, even in those with radiographic volume overload.