Background
I’ve been managing cardiac arrests for a long time, and sodium bicarbonate is one of those drugs that has never quite died died from resuscitation practice. My understanding has ben that it is unlikely to work expect in specific circumstances (e.g. tricyclic OD), but I still see it used. Interestingly often at the end of a prolonged resuscitation when everything else has failed to work and a last throw of the dice is called for. These days I see it most often used in paediatric arrests……. I don’t know why……., but that’s what I see paediatricians and PICU folk do.
In my early years bicarbonate featured much more prominently during cardiac arrest. The physiological rationale was appealing as it can make the blood gas analyser happy by reducing the acidaemia. During prolonged arrest, severe metabolic acidosis develops, myocardial function deteriorates, the response to catecholamines may be impaired and intracellular acidosis contributes to ongoing organ injury. Correcting that acidosis seemed intuitively attractive, but although bicarb may make the analyser happy, this may come at the expense of a worsening of intracellular pH which is probably at least equally important.
Unfortunately, the cognitive link between low pH on a blood gas = give bicarb seems a bit of a sticky one, and yet we know that physiology/lab measures and clinical outcomes do not always point in the same direction.
That said, I do see it used less and less, and it’s no longer in any of the mainstream guidelines. Current international guidelines recommend against its routine administration, reserving it for specific situations such as hyperkalaemia, tricyclic antidepressant overdose and selected cases of severe metabolic acidosis. Despite this, in some parts of the world, and in some patient groups, bicarbonate continues to be used surprisingly frequently during cardiac arrest, particularly in hospital.
It was therefore great to see an RCT on the use of bicarb in cardiac arrest published recently. The BIHCA investigators have performed a large randomised trial examining whether routine bicarbonate administration improves outcomes during in-hospital cardiac arrest. The abstract is below, but as always please read the full paper yourself and come to your own conclusions.
Abstract
Objective: To determine whether administration of sodium bicarbonate during in-hospital cardiac arrest increases the proportion of patients with return of spontaneous circulation.
Design, setting, and participants: Randomized, parallel-group, double-blind, placebo-controlled clinical trial conducted at 21 hospitals in Denmark. Participants were adults with in-hospital cardiac arrest, who received at least 1 dose of epinephrine. Patients were enrolled from February 6, 2023, to February 11, 2026, with the last 90-day follow-up conducted on May 4, 2026. Final statistical analysis was conducted on May 5, 2026.
Intervention: Sodium bicarbonate (up to 100 mmol) or placebo intravenously.
Main outcomes and measures: The primary outcome was sustained return of spontaneous circulation. Key secondary outcomes were survival at 30 days and survival at 30 days with a favorable neurologic outcome, defined by a score of 0 to 3 on the modified Rankin Scale (scores range from 0 to 6, with higher scores indicating greater disability).
Results: A total of 2913 patients with in-hospital cardiac arrest were screened; 913 patients were randomized, of which 779 were eligible for the primary analyses, with 372 randomized to receive sodium bicarbonate and 407 randomized to receive placebo. The median (IQR) age of patients was 73 (64-79) years and 502 were male (64%). Sustained return of spontaneous circulation occurred in 146 patients (39%) in the sodium bicarbonate group and 150 (37%) in the placebo group (risk ratio, 1.05 [95% CI, 0.88-1.24]; P = .62). At 30 days, 45 patients (12%) in the sodium bicarbonate group and 37 (9.1%) in the placebo group were alive (risk ratio, 1.25 [95% CI, 0.84-1.88]); a favorable neurologic outcome at 30 days occurred in 30 patients (8.1%) and 22 patients (5.4%), respectively (risk ratio, 1.39 [95% CI, 0.82-2.34]). Alkalosis and hypernatremia after cardiac arrest were more common in the sodium bicarbonate group.
Conclusions and relevance: There was no significant difference in sustained return of spontaneous circulation between sodium bicarbonate and placebo in adults with in-hospital cardiac arrest. These findings do not support routine administration of sodium bicarbonate for patients with in-hospital cardiac arrest.
Trial registration: ClinicalTrials.gov Identifier: NCT05564130; ClinicalTrialsRegister.eu Identifier: 2022-501304-10-00.
What kind of study is this?
This was a multicentre, randomised, double-blind, placebo-controlled superiority trial (exactly what we want to see for this kind of question) conducted in 21 Danish hospitals between 2023 and 2026.
Adults with an in-hospital cardiac arrest who required at least one dose of adrenaline were randomised to receive either intravenous sodium bicarbonate (50mL of 8.4%) or placebo immediately after the first adrenaline dose, with a second study dose (50mL of 8.4%) administered if cardiac arrest persisted. Patients with recognised indications for bicarbonate, including hyperkalaemia and sodium channel blocker toxicity, were excluded. The primary outcome was sustained return of spontaneous circulation (ROSC) lasting at least 20 minutes, with secondary outcomes including 30-day survival and neurological outcome.
So a good methodoloigcal design, with a clear protocol and reasonable outcomes.
Tell me about the patients
The trial enrolled 779 patients in the primary analysis, making it the largest randomised study of bicarbonate during in-hospital cardiac arrest to date. The median age was 73 years and almost two-thirds were male.
These patients are important to understand because they differ from many of the cardiac arrests encountered in emergency medicine and almost all of those seen in pre-hospital critical care. The age is probably older than those I see in the UK in EM/PHEM work as in my part of the world the ambulance service has a robust system of stopping resuscation in the prehospital setting for those patients who will not survive (e.g. asystole for 40 mins with no response). This takes out the majority of futile resuscitations. In addition the EM/PHEM group often do not have a defined pathology, unlike the patients in hospital in whom you would expect the clinicians have a significant background knowledge of. However, in the supplemental data it does states that 16% of patients in the trial were in ED. How that is defined is unclear, but I suspect that these are patients who arrested whilst in the ED, rather than patients who were out of hospital cardiac arrests brought to the ED. I base this assumption on numbers, but it is a little unclear.
Most patients were already admitted with significant medical pathology. Infection, cardiac disease and other acute medical illnesses accounted for the majority of admissions, while trauma represented only around 7% of the cohort. Many patients were already physiologically compromised before arrest, with a substantial proportion requiring organ support or demonstrating frailty before the cardiac arrest occurred. So not really an EM/PHEM cohort, but there may still be something to learn so let’s carry on.
What were the measured outcomes?
The primary outcome was sustained return of spontaneous circulation. That’s fair enough as during a resuscitation this is a vital way-point, but it’s not that patient focused as there have been many cardiac resuscitation interventions over the years that improve ROSC , but which don’t improve subsequent patient outcomes. To be fair to the authors though, they did follow up the patients to 30 days to look at survival and neurological outcome, defined using the modified Rankin Scale. The investigators also examined adverse effects, particularly disturbances in acid-base physiology and electrolyte balance following resuscitation.
Using ROSC as a way-point probably means a smaller trial was possible from a sample size perspective, and it is reasonble, but always be cautious around ROSC alone as an outcome measure. At St Emlyn’s we really like trials that follow patients up to at least 6 months for neurological recovery outcomes. ROSC is clearly important, but ultimately we are interested in survival and neurological recovery rather than simply restarting the heart.
What are the main results?
The trial did not find any benefit to bicarbonate adminstration. Sustained ROSC occurred in 39% of patients receiving bicarbonate and 37% receiving placebo, a difference that was neither statistically nor clinically significant. Thirty-day survival and favourable neurological outcome were numerically higher in the bicarbonate group, but again small numbers and the confidence intervals crossed unity and the study found no convincing evidence of benefit.
Administration of bicarbonate did achieve what we would expect physiologically. Patients were more likely to develop alkalosis and hypernatraemia following resuscitation, confirming that the intervention altered acid-base balance, but these biochemical changes did not translate into improved clinical outcomes. As I said at the beginning, bicarb will make your blood gas analyser happy, but does that really matter? It would appear not.
So the bottom line is that this study provides little evidence that routine bicarbonate administration improves outcomes in this patient population.
What does this paper add to the literature?
This paper confirms my initial thoughts that bicarb treats blood gas results and not patients (as an aside, that’s disappointing as I love it much more when papers change my practice). International guidelines have recommended against routine bicarbonate administration for many years, but those recommendations have rested largely on physiological concerns, observational data and extrapolation from out-of-hospital cardiac arrest. Randomised evidence has been remarkably limited, and this trial is therefore very welcome.
In particular, this study looked at the administration of bicarb much earlier in the arrest than I remember us doing in the past, and where I see it still used as a last throw of the dice in hospital. That’s a really useful difference in approach, and yet despite this, there is no effect. Whether it has a role in prolonged cardiac arrest is not tested here, but it seems unlikely. If bicarbonate has a meaningful role during routine in-hospital cardiac arrest, this trial, giving it early probably gave it the best opportunity to demonstrate benefit.
It did not.
Does this data apply to different populations such as kids? We just don’t know, but it seems very unlikely.
What do I think about this paper?
I like this study, as it addresses a common intervention that has persisted despite relatively little high-quality evidence. The investigators have conducted a pragmatic, well-designed randomised trial and produced a clear answer for the population they studied. But we do need to be careful in thinking about how far we can take this evidence. Emergency clinicians often move between environments. Many of us work in emergency departments, intensive care units and pre-hospital critical care. It is tempting to assume that evidence generated in one setting applies equally well in another, particularly in resuscitation where the algorithms appear superficially similar, which of course is incorrect.
Patients who arrest in hospital are fundamentally different from many of those encountered in emergency medicine and prehospital practice. They already have an established diagnosis, often have significant physiological derangement before the arrest and usually receive advanced life support within minutes. In contrast, patients presenting to the emergency department or managed in the prehospital environment frequently have undifferentiated pathology. Trauma, toxicology, profound hyperkalaemia, severe diabetic ketoacidosis and prolonged out-of-hospital cardiac arrest all present different physiological problems, and some of those remain recognised indications for bicarbonate therapy. In this study most of those patients were deliberately excluded. Does this apply to kids? We just don’t know as they were not included, but it seems very unlikely.
So for me, I think the paper tells me that routine use of bicarb won’t work, and if I find myself reaching for it (very unlikely) or see someone else suggesting it (likely) the question should be – what specific reason are you giving that for (and the answer ‘acidaemia’ is not a good one).
Should we change practice based on this study?
For most UK emergency physicians, probably not. Routine bicarbonate administration has already disappeared from most advanced life support practice, and this trial provides support for continuing that approach.
Patients with hyperkalaemia, sodium channel blockade, severe metabolic acidosis or some forms of traumatic arrest remain separate clinical questions. Those decisions still require an understanding of the underlying physiology rather than a blanket application of this trial.
Bottom line
The BIHCA trial provides high-quality evidence that routine sodium bicarbonate administration does not improve return of spontaneous circulation or survival during in-hospital cardiac arrest. It strengthens existing guideline recommendations against routine use and should probably close the chapter on bicarbonate as a standard component of advanced life support in this setting.
For emergency and prehospital clinicians, however, the more important message is slightly different. This trial tells us that bicarbonate should not be given routinely. It does not tell us that bicarbonate should never be given. Those are very different conclusions, and recognising the difference is probably the most useful lesson from this study.
vb
S
References and further reading
- Granfeldt A, Kirkegaard BL, Vallentin MF, Holmberg MJ, Stankovic N, Lind PC, Kjærgaard AG, Houe N, Vested M, Isbye DL, Kristensen CM, Kjærgaard J, Zwisler ST, Lundsfryd SN, Petersen LB, Folke F, Hansen CM, Iversen KK, Schou M, Wiberg S, Riddersholm SJ, Rasmussen BS, Karlsson K, Plovsing RR, Juhl CS, Krag M, Sørensen CD, Kuhne-Qvist P, Oxlund J, Sølling CG, Hansen JR, Andersen LPK, Jensen JMS, Itenov TS, Lauritzen M, Pedersen MJ, Nielsen RP, Fjeldsøe-Nielsen H, Mirzada N, Secher N, Espelund US, Eskol JR, Andersen K, Brøchner AC, Nedergaard HK, Hansen PM, Henschel W, Walli AR, Sandgaard M, Asferg CL, Ciubotariu A, Rasmussen MHR, Le Breton V, Nielsen LK, Hansen FG, Brændstrup N, Thomsen IK, Sindberg B, Andersen LW; Bicarbonate for In-Hospital Cardiac Arrest (BIHCA) Investigators. Sodium Bicarbonate for In-Hospital Cardiac Arrest: A Randomized Clinical Trial. JAMA. 2026 Jun 11:e2610628. doi: 10.1001/jama.2026.10628. Epub ahead of print. PMID: 42273960; PMCID: PMC13261510.
