SGEM#394: Say Bye Bye Bicarb for Pediatric In-Hospital Cardiac Arrest

Reference: Cashen K, Reeder RW, Ahmed T, et al. Sodium bicarbonate use during pediatric cardiopulmonary resuscitation: a secondary analysis of the icu-resuscitation project trial. Pediatric Crit Care Med. 2022 Date: February 15, 2023 Guest Skeptic: Dr. Carlie Myers is Pediatric Critical Care Attending at Cincinnati Children’s Hospital Medical Center. Dr. Carlie Myers Case: A 6-month-old boy presents to the emergency department (ED) with three days of worsening cough, cold symptoms, and fever. Parents note that he has been progressively more tired and difficult to arouse. He is found to be in hypoxic respiratory failure and septic shock. Intravenous (IV) access is obtained. He is quickly intubated. Despite multiple fluid boluses, he remains hypotensive and is started on vasoactive support. His blood gas reveals a mixed respiratory and metabolic acidosis with a lactate of 5.0. Despite your best efforts, he has an episode of agitation leading to hypoxia and subsequent cardiac arrest. Your team begins high quality cardiopulmonary resuscitation (CPR). An arterial blood gas is obtained demonstrates a pH of 7.0, PaCO2 of 70, PaO2 of 28, HCO3- of 7, Base Deficit of -10, and Lactate 10.0. A team member asks if you want to administer some sodium bicarbonate (1mEq/kg). Background: We often manage patients in cardiac arrest in the ED or the intensive care unit (ICU). Apart from high-quality CPR and early defibrillation, many other interventions we try lack a strong evidence base. But that does not stop us from trying to save the patient’s life and may represent some intervention bias.[1] The SGEM has covered the use of epinephrine, vasopressin, methylprednisolone, and calcium for cardiac arrest in SGEM#238, SGEM#350, and SGEM#353. Today we are focusing on sodium bicarbonate. Sodium bicarbonate has historically been used during CPR with the goal of alkalizing blood pH and treating metabolic acidosis. There are a few key assumptions about the use of sodium bicarbonate.  Low pH decreases cardiac function and responsiveness to catecholamines. Sodium bicarbonate administration will increase the pH. The increase in pH will lead to improved responsiveness to catecholamines and cardiac function. But it’s not that straightforward. Many of the studies supporting these claims were conducted on animal models or in vitro. [2] It is unclear if we see the same effects of acidosis and sodium bicarbonate in vivo. HCO3- + H+ ↔ H2O + CO2 Rapid bicarbonate infusion can cause an imbalance in CO2 across the cell membrane. HCO3- + H+ converts to H2CO3 and then to CO2 +H20. Extracellular CO2 rises rapidly, it diffuses across cell membranes and the reverse reaction occurs H2O + CO2→ HCO3- + H+; therefore, creating intracellular acidosis. There was a lack of evidence about the benefits and potential harm from using sodium bicarbonate in cardiac arrest [3], so it was removed from the American Heart Association’s (AHA) guidelines. The latest guidelines from the AHA in 2020 state, “clinical trials and observational studies since the 2010 guidelines have yielded no new evidence that routine administration of sodium bicarbonate improves outcomes from undifferentiated cardiac arrest and evidence suggests that it may worsen survival and neurological recovery.” [4] This association seems to hold true in the pediatric literature as well. [5-6] Clinical Question: What is the association between sodium bicarbonate use and pediatric in-hospital cardiac arrest mortality and morbidity? Reference: Cashen K, Reeder RW, Ahmed T, et al. Sodium bicarbonate use during pediatric cardiopulmonary resuscitation: a secondary analysis of the icu-resuscitation project trial. Pediatric Crit Care Med. 2022 Population: Pediatric patients 37 weeks to 18 years of age who received chest compressions across 18 pediatric intensive care units (PICU) or pediatric cardiac intensive care units (PCICU) from Oct 2016 to March 2021. Excluded: Children were excluded if prior to the arrest, they had terminal disease and were not expected to survive, documented lack of commitment to aggressive ICU therapies, brain death, and out of hospital cardiac arrests. Additionally, in the secondary analysis patients on extracorporeal membrane oxygenation (ECMO) at the time of CPR were excluded. Intervention: Sodium Bicarbonate administration during CPR Comparison: No sodium Bicarbonate administered during CPR Outcome: Primary Outcome: Survival to hospital discharge Secondary Outcomes: Return of spontaneous circulation (ROSC), survival to hospital discharge with favorable neurologic outcome (Pediatric Cerebral Performance Category [PCPC] of 1), functional status at the time of discharge using the Functional Status Scale (FSS), and presence of a new morbidity defined as worsening from baseline FSS by 3 or more points. Authors’ Conclusions: “Sodium bicarbonate use was common and associated with lower rates of survival to hospital discharge.” Quality Checklist for Observational Study: Did the study address a clearly focused issue? Yes Did the authors use an appropriate method to answer their question? Yes Was the cohort recruited in an acceptable way? Yes Was the exposure accurately measured to minimize bias? Unsure. Was the outcome accurately measured to minimize bias? Yes Have the authors identified all-important confounding factors? Unsure. Was the follow up of subjects complete enough? Yes How precise are the results? Unsure Do you believe the results? Yes Can the results be applied to the local population? Unsure Do the results of this study fit with other available evidence? Yes Financial Conflicts of Interest: No Results: 1,100 CPR events were included in the study. Approximately half (48%) received sodium bicarbonate. The median age was 0.63 years (IQR 0.19-3.81 years) and 53.5% were male. The most common pre-existing medical condition was respiratory insufficiency. Key Result: Sodium bicarbonate use had no association with ROSC and was associated with lower survival to hospital discharge and lower survival to hospital discharge with favorable neurologic outcomes. Primary Outcome: Survival to hospital discharge was 42.2% in patients who received sodium bicarbonate vs 73.3% (aOR 0.7, 95% CI; 0.54-0.92) Secondary Outcomes: ROSC: Sodium bicarbonate use had no association Survival to hospital discharge favorable neurologic outcomes and new morbidity 1)Patient Population: There are a few things to say about the cohort included in this study. One issue was excluding patients who were terminal and not expected to survive the hospitalization. It can be difficult at times to predict when patients will die. This subjective exclusion criteria could have introduced some selection bias. Another issue is a large portion of the patients included in this analysis had underlying are medical or surgical cardiac disease (58%) that included congestive heart failure, pulmonary hypertension, congenital heart disease, and single ventricle heart disease. Sodium bicarbonate was used more commonly in PCICU compared to those in the PICU.  These were in-hospital cardiac arrests. If outcomes with bicarb were this poor in the hospital (a better place to experience cardiac arrest than out of hospital), this doesn’t bode well for sodium bicarbonate use for out of hospital cardiac arrests. 2) Time Dependent Propensity Matching: The time dependent propensity matching of bicarbonate administration was not available in original database. The patients in this study may have received bicarbonate at different times. In the presence of time-varying treatment or exposure (HCO3), the conventional method (propensity scoring is traditionally time fixed) may cause bias because subjects with early and late exposure are treated as the same. Any patient receiving HCO3 was treated the same, regardless of the time frame the HCO3 was administered. After time-dependent propensity matching, the matched cohort can be analyzed with conventional Cox regression model or conditional logistic regression. For more information on propensity score matching, you can read Peter Austin’s introductory article on the topic. [7] 3) Confounding Factors: This was an observational study with only measured confounding factors that could be controlled for with adjustments. It means we can only conclude associations from this data not causation. Sodium bicarbonate use was associated with prolonged resuscitation time and additional pharmacologic interventions during CPR (epi, atropine, calcium, vasopressin, amiodarone, lidocaine, and fluid boluses). Sodium bicarbonate was also used more often in children with higher (Pediatric Risk Mortality Score) PRISM and VIS (Vasoactive Inotropic Score)– both markers of severity of illness. Were outcomes worse in the group receiving sodium bicarbonate because of the sodium bicarbonate or because the length of resuscitation was longer, the patients were sicker, and what about all the other medications or interventions? 4) Futility: When it comes to using sodium bicarbonate in cardiac arrest, we’ve heard some clinicians say, “well what’s the harm? The patient is dead already.” We want to caution against this mindset. This is an extremely vulnerable patient population, and our interventions should be deliberate and targeted to maximize potential benefit and reduce potential harm. Additionally, spending time performing interventions that don’t help can take away time and energy spent on performing more meaningful interventions. 5) Racial Demographics: While this study did not particularly look at variable outcomes based on race or gender of the patient, race and ethnicity are worthy demographics to report for 1) Generalizability of study findings and 2) Further examination of results in the context of equitable care.