Could patient-controlled thirst-driven fluid administration lead to more rapid rehydration than clinician-directed fluid management? An early feasibility study.

Abstract

Fluid management is a major factor determining perioperative outcome, yet in reality, fluid administration practice is variable. Thirst however, is a highly sensitive and reliable indicator of fluid deficits. We explored the use of thirst sensation to trigger i.v. fluid boluses to guide individualized fluid management. We performed a randomised double crossover trial on 16 healthy male volunteers, of mean age 31 yr and BMI 24.4 kg m^-2. Twice, after administrations of oral furosemide (40 mg) and 12 h of oral fluid restriction, participants received a 4-h i.v. fluid infusion. In the experimental arm, participants pressed a trigger to relieve their thirst, administering a 200 ml bolus. In the control arm, i.v. fluid was infused following National Institute for Health and Clinical Excellence (NICE) guidelines at 1.25 ml kg^-1 h^-1 with a clinician delivered 500 ml i.v. bolus in response to clinical signs of dehydration. Plasma osmolality and urine specific gravity were measured before and after each infusion. More fluid was infused in response to thirst than by adherence to NICE guidelines, with a mean difference of 743 ml (P=0.0005). Thirst-driven fluid administration was fitted to an exponential function of time, plateauing after a mean half-life of 98.8 min. In the experimental arm there was a greater reduction in urine specific gravity and thirst score with mean differences 0.0053 g cm^-3 (P=0.002) and 3.3 (P=0.003), respectively. Plasma osmolality demonstrated no fluid overload. A system delivering i.v. fluid in response to subjective thirst corrects fluid deficits in healthy participants. A clinical feasibility study will assess the potential use of this system in the perioperative setting. NCT 03176043.