- Design
- Pragmatic, adaptive, double-blind randomised trial
- Population
- 500 critically ill adults with metabolic acidosis receiving vasopressors
- Primary outcome
- Major adverse kidney event within 30 days
- Effect
- 40.2% vs 39.4%; adjusted difference 1.2 points (95% CI -7.1 to 9.4), p=0.78
Sodium bicarbonate is widely given to correct metabolic acidosis in critically ill patients on vasopressors, but whether it helps has been uncertain. SODa-BIC, a pragmatic double-blind trial in 500 adults across 55 ICUs, tested it directly.
Patients with metabolic acidosis (pH below 7.30) receiving vasopressors were randomised to bicarbonate or placebo, titrated to a pH of at least 7.30. There was no benefit. A major adverse kidney event within 30 days occurred in 40.2% with bicarbonate and 39.4% with placebo, with the confidence interval spanning no effect. Renal-replacement therapy use and 30-day mortality did not differ meaningfully, and a few adverse effects occurred only in the bicarbonate group.
The practical change is to stop giving sodium bicarbonate reflexively to correct the number in acidosis with shock. This trial does not address extreme acidaemia or specific indications such as hyperkalaemia or certain poisonings, but for the common scenario it removes the rationale.
- Pragmatic double-blind trial of 500 ICU adults with metabolic acidosis on vasopressors.
- Bicarbonate or placebo was titrated to a target pH of at least 7.30.
- Major adverse kidney events were 40.2% with bicarbonate versus 39.4% with placebo.
- Renal-replacement therapy use and 30-day mortality did not differ meaningfully.
- The trial does not cover extreme acidaemia or specific indications like hyperkalaemia.
Why it matters
It removes the rationale for a widespread reflex, correcting the pH number, that this trial shows does not help.
Don't overread it
This tested acidosis with shock at pH below 7.30; it does not speak to severe acidaemia or indications such as hyperkalaemia or specific poisonings.
The statistics, in plain English
A 40.2% versus 39.4% event rate, with an adjusted difference of 1.2 percentage points whose confidence interval (-7.1 to 9.4) crosses zero, means no effect was detected. The trial was powered to find a clinically relevant benefit, so its absence is informative.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for emergency & critical care, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free