Design: Pragmatic, adaptive, double-blind, placebo-controlled RCT
Population: 500 ICU adults with metabolic acidosis on vasopressors
Primary outcome: Major adverse kidney events within 30 days
Effect: 40.2% vs 39.4%; adjusted difference 1.2 points (95% CI −7.1 to 9.4)

SODa-BIC was a pragmatic, adaptive, double-blind trial in 55 ICUs across seven countries, published in June. It randomised 500 adults on vasopressors with metabolic acidosis (pH below 7.30, base excess −4 or lower, no respiratory acidosis) to sodium bicarbonate infused for up to five hours towards pH 7.30, or 5% dextrose placebo. The primary outcome was the major adverse kidney event composite at 30 days.

MAKE30 occurred in 40.2% with bicarbonate and 39.4% with placebo (adjusted difference 1.2 points, 95% CI −7.1 to 9.4). Renal replacement therapy was used in 16.8% vs 20.9% (difference −3.9 points, −10.6 to 2.7) and 30-day mortality was 25.4% vs 24.0%. Four patients had adverse effects with bicarbonate, none with placebo.

Bicarbonate has often been given in acidotic shock partly to delay or avoid dialysis. This trial did not find a kidney benefit. For nephrologists consulted on acidotic, vasopressor-dependent patients, the decision about renal replacement should rest on the usual indications rather than a trial of bicarbonate.

Do not use bicarbonate infusion to try to avoid renal replacement therapy in acidotic, vasopressor-dependent patients.
Base the decision to start dialysis on standard indications: refractory acidosis, hyperkalaemia, fluid overload, uraemic complications.
Continue bicarbonate for specific indications, including hyperkalaemia with acidosis and bicarbonate-losing states.
Watch sodium load and ionised calcium if bicarbonate is given.

Why it matters

It removes a commonly cited kidney rationale for bicarbonate in shock.

The statistics, in plain English

The adjusted difference of 1.2 percentage points in MAKE30 has a confidence interval from −7.1 to 9.4, which crosses zero. The renal replacement difference (−3.9 points) is not significant and should not be read as a hidden benefit.