Model-predicted capillary leakage in graded hypotension: extended analysis of experimental spinal anesthesia.

Background Crystalloid fluid infused during the induction of spinal anesthesia is involved in a complex set of physiological responses, including vasodilatation, reactive vasoconstriction, and changes in mean arterial pressure (MAP). The present evaluation compares the modeled capillary leakage in anesthetized versus non-anesthetized body regions. Methods Ten female volunteers (mean age, 29 years) received 25 mL/kg of Ringer´s acetate over 60 min during experimental spinal anesthesia. Blood hemoglobin was measured repeatedly in the radial artery (reference), arm (cubital) vein, and leg (femoral) vein for 240 min. Each pattern of data served as a dependent variable in volume kinetic analyses that used mixed models software and MAP as covariate. Results The capillary leakage of fluid from the plasma to the extravascular space peaked at 17 mL/min when MAP was 100 mmHg, and the two venous curves were virtually identical. At MAP 60 mmHg, the rate was reduced to 10-12 mL/min when assessed in arterial blood and leg vein blood, but only 5 mmHg in blood collected from the arm vein. The distribution half-life of infused fluid was then 40 min in the leg and 80 min in arm. These results suggest that vasoconstriction in non-anesthetized body regions halves the capillary leakage that is observed in vasodilated, anesthetized body regions. Conclusion Graded hypotension during spinal anesthesia reduced the capillary filtration of fluid as determined by volume kinetic analysis. The effect was twice as great when venous blood was sampled from a non-anesthetized body region than from an anesthetized body region.