Routine administration of sodium bicarbonate during in-hospital cardiac arrest did not statistically significantly increase sustained return of spontaneous circulation in adults who received epinephrine, according to findings from the Bicarbonate for In-Hospital Cardiac Arrest trial.
In the randomized, parallel-group, double-blind, placebo-controlled superiority trial, researchers randomized 913 adults with in-hospital cardiac arrest at 21 hospitals in Denmark. Of these, 779 patients were eligible for the primary modified intention-to-treat analyses, including 372 assigned to sodium bicarbonate and 407 assigned to placebo.
The trial is the first randomized clinical trial to evaluate sodium bicarbonate specifically during in-hospital cardiac arrest. Prior randomized trials of buffer therapy were conducted in out-of-hospital cardiac arrest, including the 2 largest, conducted more than 30 years ago. In 1 prior trial that reported timing, the median time to trial-drug delivery was 31 minutes. In the current trial, the median time from cardiac arrest to trial-drug administration was 8 minutes.
Eligible patients were aged 18 years or older, had in-hospital cardiac arrest, and received at least 1 dose of epinephrine during resuscitation. Patients were excluded if they had a clearly documented do-not-resuscitate order, prior trial enrollment, invasive mechanical circulatory support at the time of arrest, known or suspected pregnancy, known objection to trial participation, or a clinical indication for sodium bicarbonate.
Most patients had an initial nonshockable rhythm, and the median time from cardiac arrest to epinephrine administration was 5 minutes.
Patients assigned to sodium bicarbonate received 50 mL of 8.4% sodium bicarbonate, equivalent to 50 mmol, as soon as possible following the first epinephrine dose. If they remained in cardiac arrest, 1 additional 50-mmol dose was given following the next epinephrine dose, for a maximum of 100 mmol. Patients assigned to placebo received 0.9% sodium chloride from identical vials. The researchers wrote that fixed dosing was selected because accurate weight-based dosing during cardiac arrest was not considered feasible.
The primary outcome was sustained return of spontaneous circulation, defined as a palpable pulse or other signs of circulation with no further need for chest compressions for at least 20 minutes. Sustained return of spontaneous circulation occurred in 39% of patients in the sodium bicarbonate group and 37% of patients in the placebo group.
Key secondary outcomes numerically favored sodium bicarbonate but did not show clear between-group differences. At 30 days, 12% of patients in the sodium bicarbonate group and 9% of patients in the placebo group were alive. Survival at 30 days with a favorable neurologic outcome, defined as a modified Rankin Scale score of 0 to 3, occurred in 8% and 5% of patients, respectively. Because the primary outcome was not statistically significant, additional P values were not calculated for these key secondary outcomes under the prespecified hierarchical testing plan.
The researchers noted that the relatively narrow confidence interval for sustained return of spontaneous circulation made a large difference in that proximate outcome unlikely, and that a large difference in longer-term outcomes therefore also appeared unlikely.
Findings were generally consistent across prespecified subgroup analyses, including analyses by age, initial rhythm, witnessed arrest, time from arrest to trial-drug administration, and known metabolic acidosis prior to cardiac arrest. However, the subgroup analyses were not designed to provide definitive evidence of treatment-effect heterogeneity, and subgroup estimates were unadjusted.
Additional outcomes at 90 days followed the same pattern, again favoring sodium bicarbonate without reaching a clear difference. Survival at 90 days was 11% in the sodium bicarbonate group and 8% in the placebo group. Survival at 90 days with a favorable neurologic outcome occurred in 9% and 7% of patients, respectively. Quality-of-life analyses among survivors showed no clear differences between groups.
Sodium bicarbonate produced expected physiologic changes but did not translate into improved clinical outcomes. Patients in the sodium bicarbonate group had higher pH and standard bicarbonate values after cardiac arrest, although patients who achieved sustained return of spontaneous circulation in the sodium bicarbonate group remained severely acidotic following cardiac arrest, with a median pH of 7.11. Whether higher doses of sodium bicarbonate during cardiac arrest could improve outcomes remains unknown, the researchers wrote.
Alkalosis and hypernatremia were more common with sodium bicarbonate, occurring in 35% vs 20% and 42% vs 29% of patients, respectively.
The researchers noted several limitations. The trial was powered for sustained return of spontaneous circulation rather than longer-term patient-centered outcomes, including survival and neurologic recovery. Laboratory values such as pH and standard bicarbonate could not be measured during cardiac arrest prior to enrollment, resulting in a heterogeneous population with varying degrees of baseline acidosis. The trial was also conducted exclusively in Denmark, and generalizability to other health care systems is uncertain.
The findings apply to routine sodium bicarbonate use during in-hospital cardiac arrest and do not address use in special circumstances in which sodium bicarbonate may be clinically indicated, such as hyperkalemia or tricyclic antidepressant overdose.
Current European and US resuscitation guidelines suggest against routine sodium bicarbonate administration during cardiac arrest. The researchers noted that these recommendations have been based on limited clinical trial evidence, despite common use of sodium bicarbonate during in-hospital cardiac arrest. Sodium bicarbonate has been administered in more than half of in-hospital cardiac arrests in the US, according to prior research cited in the study.
In an accompanying editorial, Clifton W. Callaway, MD, PhD, of the University of Pittsburgh School of Medicine, wrote that the trial's design and execution strengthened confidence in the findings, citing clear eligibility criteria, concealed randomization, achievement of the target sample size, minimal missingness for the primary outcome, and similar results in adjusted and unadjusted analyses. He noted that the modified intention-to-treat analysis could raise concern but argued that excluding randomized patients who never received the blinded study drug was unlikely to bias the results. He also wrote that sustained return of spontaneous circulation was a reasonable primary outcome because it was closely linked in time to the intervention and less vulnerable than longer-term outcomes to postrandomization treatment decisions.
However, Dr. Callaway cautioned that the findings may not change practice. He wrote that clinicians often focus treatment decisions on individual patients rather than average treatment effects in trial populations and may continue to look for subgroups or individual patients who could benefit. He pointed to analyses of heterogeneity of treatment effect, including machine learning approaches, as a possible next step, while emphasizing that any such post hoc findings would need confirmation in separate data sets or trials. He also acknowledged that some clinicians may continue to administer sodium bicarbonate in individual resuscitation attempts despite the neutral findings. He described the neutral result as at least reassuring that sodium bicarbonate is not harmful. The trial also showed physiologic effects consistent with bicarbonate delivery, including higher pH and blood bicarbonate values among survivors in the sodium bicarbonate group.
Dr. Callaway also cautioned that the results may not directly settle questions about out-of-hospital cardiac arrest, where patient characteristics, arrest etiologies, outcomes, time to drug delivery, and degree of metabolic acidosis may differ substantially. Still, he wrote that the trial "reinforces and provides stronger evidence" for current guideline recommendations against routine sodium bicarbonate use during in-hospital cardiac arrest.
"These findings do not support routine administration of sodium bicarbonate for patients with in-hospital cardiac arrest," wrote lead study author Asger Granfeldt, DMSc, of Aarhus University Hospital in Denmark, and colleagues.
Disclosures can be found in the published investigation and editorial.
Source: JAMA