Buffering a Number: Sodium Bicarbonate in Severe AKI-Associated Acidosis
A single patient with severe combined metabolic acidosis and acute kidney injury from suspected mesenteric ischemia. The disagreement is what treating the pH number can still be expected to buy, now that the trial built to test that subgroup has reported.
D.L., a 39-year-old construction foreman, spent the morning on a scaffold before the abdominal pain that started as a dull ache became, within three hours, the kind of pain that had him doubled over and unable to finish a sentence. He has no prior abdominal surgery, no known vascular disease, and by his own account has never spent a night in a hospital — a genuinely healthy 39-year-old until a CT angiogram this afternoon showed a superior mesenteric artery occlusion with a segment of bowel already showing signs of poor enhancement. He is in the operating room now for exploration and likely resection; this consultation is happening in the interval while the surgical team works, because his metabolic picture has become its own emergency running in parallel.
His arterial pH is 7.06, bicarbonate 8 mEq/L, with a lactate of 14 — ischemic bowel producing lactate faster than any buffering system can plausibly keep pace with, and his creatinine has already risen to 2.8 mg/dL from a baseline the trauma team confirmed as normal on a physical exam three months ago, AKIN stage 2 and moving in the wrong direction. The question the team is weighing is not whether he is critically ill — that is not in dispute — but whether sodium bicarbonate, given now, actually changes what happens to him, or simply corrects a number on a blood gas while the ischemic bowel driving it keeps producing acid regardless. BICAR-ICU found no overall mortality benefit for bicarbonate in severe acidemia, but in its pre-specified AKIN stage 2–3 stratum — where his own AKIN 2 puts him, a staging score rather than the SOFA threshold that governed entry — 28-day mortality was lower, 46% against 63%, and renal replacement therapy was used less. That stratum looked promising enough that a second trial, BICARICU-2, was built to enroll it prospectively: pH at or below 7.20, bicarbonate at or below 20, KDIGO stage 2–3 AKI. His 7.06, his 8, and his stage 2 meet all three, which is unusual — he is squarely inside that population rather than adjacent to it. It did not reproduce the mortality signal: 62.1% against 61.7% at ninety days. What it did reproduce was the renal one, roughly doubling the median time to starting dialysis. So the live question is no longer whether bicarbonate improves his survival, since the trial built to answer that has answered no, but whether hours bought before renal replacement are worth anything in a patient whose acid is still being manufactured by bowel that has not yet come out.
ICU-surgical liaison call, mid-operation
A pH of 7.06 is its own physiologic threat independent of the cause — it impairs catecholamine responsiveness, which is part of why his pressor requirement keeps climbing, and it worsens myocardial contractility. I want sodium bicarbonate now, not as a cure for the ischemia but to buy hemodynamic stability while the surgical source control finishes. I am not claiming it changes whether he survives; I am claiming it changes what I have to do to his norepinephrine over the next hour.
I have to be careful here, because the evidence moved and it moved against the argument I would have made a few years ago. BICAR-ICU's AKIN 2–3 stratum did show lower 28-day mortality and less RRT, and he fits that stratum on the numbers. But BICARICU-2 then enrolled precisely that population prospectively and found no mortality difference at 28, 90, or 180 days. What survived was the renal endpoint — time to dialysis initiation roughly doubled. I am not going to argue survival to you, because I would be arguing from a subgroup that has since been tested head-on and did not replicate.
The limitation that does still stand is cause: both trials enrolled acidosis driven mostly by sepsis and shock, not by bowel that is still ischemic. Hours bought before dialysis are worth more if source control lands inside those hours than if it does not.
From where I am standing, the honest answer is that nothing changes his acidosis until this bowel is out, and I do not think either of you is arguing otherwise. My only input is timing: whatever bicarbonate strategy you choose, correct it gradually rather than in one large bolus — rapid correction can transiently worsen intracellular acidosis and shift ionized calcium in ways that complicate an already difficult field.
I am not taking a position on whether to give it, only on how, if the two of you decide to.
Agreed: sodium bicarbonate started as a gradual infusion rather than a rapid bolus, continued until source control is achieved and the metabolic picture can be reassessed off the operating table.
Explicitly NOT claimed, and named as such rather than left to be inferred: that this improves his odds of surviving. BICARICU-2 tested exactly that in exactly his enrollment population and found no difference, so the team treated on the narrower ground that delaying renal replacement has real value in a patient this unstable.
Not agreed: whether even that narrower benefit, established in acidosis driven mostly by sepsis and shock, holds where the acid is being actively manufactured by ischemic bowel. The decision to treat was made while that uncertainty stood, not after it was settled.