No difference in overcorrection risk with rapid bolus vs slow infusion of hypertonic saline for severe hyponatremia (SALSA)

Référence

Baek SH, Jo YH, Ahn S, et al. Risk of overcorrection in rapid intermittent bolus vs slow continuous infusion therapies of hypertonic saline for patients with symptomatic hyponatremia. JAMA Intern Med 2020 Oct 26:e205519. doi: 10.1001/jamainternmed.2020.5519. Online ahead of print.

Sommaire

The safest mode of hypertonic saline administration to avoid overcorrection of severe hyponatremia is unknown. These authors randomized 178 adults in emergency departments and hospital wards who had evidence of symptomatic severe hyponatremia to receive either rapid intermittent boluses or a slow continuous infusion of 3% hypertonic saline over 24 to 48 hours. Enrolled patients had moderately severe to severe symptoms of hyponatremia and a sodium level of 125 mmol/L or less. In the rapid group, 3% hypertonic saline was administered over 20 to 40 minutes initially and repeated every 6 hours as needed. Dosing was based on the patient's weight. In the slow group, hypertonic saline was infused at a rate of 0.5 to 1 mL/kg/h and then adjusted based on sodium levels every 6 hours. The target sodium correction rate was an increase by 5 to 9 mmmol/L in the first 24 hours, and either by 10 to 17 mmol/L in the first 48 hours or to an overall sodium level of 130 mmol/L or higher. The 2 groups were similar at baseline: mean age was 73 years, mean sodium level was 118 mmol/L, and the top 2 causes of hyponatremia were thiazide use and syndrome of inappropriate antidiuresis. There were multiple patients excluded in both groups, mainly due to protocol violation. For the primary outcome of the incidence of overcorrection (an increase in sodium level > 12 mmol/L in the first 24 hours or > 18 mmol/L in the first 48 hours), there was no significant difference in the 2 groups (17.2% in rapid group vs 24.2% in slow group; P = .26). Results were similar in the per-protocol analysis. There were no incidences of osmotic demyelination syndrome in either group, but the study was not powered to detect a difference in this rare adverse event. Sodium re-lowering treatment was less frequently required in the rapid group than in the slow group (41.4% vs 57.1%; P = .04; number needed to treat [NNT] = 6.3). In a post-hoc analysis, patients in the rapid group were more likely than those in the slow group to achieve the target sodium correction rate within one hour (32.2% vs 17.6%; P = .02; NNT = 6.8).