Hyponatraemia and intracellular water in sepsis: an experimental comparison of the effect of fluid replacement with either 0.9% saline or 5% dextrose.

Hyponatraemia associated with sepsis is known to have an increased morbidity and mortality. The cause of this phenomenon is unknown, but may be related to dilution of the extracellular space with retained exogenous fluid. Fluid and ion redistribution across the cell membrane of striated muscle was investigated in an animal sepsis model and compared with sham controls. The objective was to study the effect of different volumes of fluid replacement with either 0.9% saline or 5% dextrose. Significant shifts of sodium, chloride, and water occurred into the cell in all septic animals but not in controls. This trend was exacerbated by the use of dextrose for intravenous replacement even when the estimated normal fluid requirements had not been exceeded. Hyponatraemia and plasma hypoosmolality were induced only in septic animals, which received 100% of their fluid requirements as dextrose. These animals at the same time had significantly reduced extracellular and increased intracellular volumes compared with controls and the septic animals that received saline replacement. It is concluded that the hyponatraemia and plasma hypoosmolality that occurs in these animals is caused by a combination of intracellular shift of sodium and water, and dilution of the extracellular space, probably on the basis of physiological antidiuretic hormone (ADH) secretion. Dextrose (and by implication 4% dextrose/0.18% saline) is inappropriate, potentially dangerous, and should be avoided in these circumstances.