BACKGROUND: Fluid supplement may be affected by ventilatory management due to physiological interaction between heart and lung. The aim of the present study was to explore the effects of ventilator strategies on the relationship of fluid balance and hospital mortality for sepsis patients. METHODS: This was a retrospective cohort study included sepsis patients with invasive mechanical ventilation (MV) over 24 h from Medical Information Mart for Intensive Care (MIMIC) IV database. The accumulative fluid balance increased by 6 h intervals were calculated as fluid intake minus fluid output. The modes (assisted or controlled) and levels (high or low) of positive end-expiratory pressure (PEEP) of MV every 6 h were recorded. The modification effect for modes and levels of PEEP on the relationship of fluid balance and hospital mortality were tested by multivariable regression models, respectively. RESULTS: A total of 4466 sepsis patients with invasive MV were included, of which hospital mortality was 26.5%. Fluid balance seemed to have U-shape relationship with hospital mortality. The majority of patients used controlled ventilation at the beginning, and switched to assisted ventilation gradually
however, the PEEP level did not change a lot during the first 24 h. The relationship between fluid balance and hospital mortality was not modified by the ventilator mode
while the PEEP level may modify the relationship. CONCLUSIONS: For sepsis patients admitted to ICU with invasive MV, the PEEP level, but not the mode of MV, appeared to modify the relationship of fluid balance and hospital mortality. The setting of mechanical ventilation may be an important consideration for fluid therapy.