Sodium bicarbonate corrected the acidosis and changed nothing that mattered

Design

pragmatic, adaptive, double-blind, placebo-controlled randomised trial in 55 ICUs across seven countries (SODa-BIC)

Population

500 adults with metabolic acidosis (pH below 7.30, base excess no more than -4 mmol/L) receiving vasopressors in the ICU

Primary outcome

major adverse kidney event (death, renal-replacement therapy or persistent renal dysfunction) within 30 days

Effect

40.2% vs 39.4% (adjusted difference 1.2 percentage points, 95% CI -7.1 to 9.4; P = 0.78); renal-replacement therapy 16.8% vs 20.9%; 30-day in-hospital mortality 25.4% vs 24.0%

SODa-BIC randomised 500 adults across 55 ICUs in seven countries with metabolic acidosis — pH below 7.30, base excess no more than -4 mmol/L, with a controlled carbon dioxide — who were receiving vasopressors. They were given either sodium bicarbonate or 5% dextrose placebo, infused for up to five hours and titrated to a pH of at least 7.30 and a base excess of at least zero.

Major adverse kidney events within 30 days — death, renal-replacement therapy or persistent renal dysfunction — occurred in 40.2% of the bicarbonate group and 39.4% of the placebo group. In-hospital mortality by day 30 was 25.4% against 24.0%. Renal-replacement therapy was used in 16.8% against 20.9%, a difference whose interval crosses zero.

This is a double-blind, placebo-controlled test of a practice that is deeply established and almost entirely reasoned rather than evidenced. The pH is a number that looks wrong and can be made to look right, and correcting it feels like treatment. This trial gave bicarbonate in the population where the case for it is strongest — acidotic and on vasopressors — titrated properly, and found nothing.

The practical change is to stop treating the number. A pH below 7.30 in a patient on vasopressors is a signal to find and fix the cause: the sepsis, the hypoperfusion, the toxin, the renal failure. Giving bicarbonate treats the display.

• Do not give sodium bicarbonate to correct pH in a vasopressor-dependent patient with metabolic acidosis
• Treat the cause the acidosis points to — source control, perfusion, toxin, renal replacement where indicated
• Keep the specific indications where bicarbonate has separate grounds, such as severe hyperkalaemia or sodium-channel-blocker toxicity
• Remove 'correct the acidosis' from admission order sets that still carry it as a default
• Note the entry criteria: pH below 7.30 with a controlled PaCO2 — this was not tested in respiratory acidosis

The statistics, in plain English

The adjusted difference of 1.2 percentage points with a 95% confidence interval of -7.1 to 9.4 and P = 0.78 is a clear null on the primary outcome, though with 500 patients the interval is wide enough that a benefit or harm of up to about 7 to 9 percentage points cannot be excluded. The renal-replacement difference of -3.9 points (95% CI -10.6 to 2.7) favours bicarbonate numerically but crosses zero, and was one of several secondary outcomes — reading a trend from it would be exactly the sort of over-interpretation the primary result argues against. Four adverse effects in the bicarbonate group against none on placebo (P = 0.06) is a small signal in a small sample, but it is not in the reassuring direction.

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.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.