Sodium Bicarbonate Use in Shock and Cardiac Arrest

OBJECTIVES: To evaluate the preferences and self-reported practices of pediatric acute care physicians with respect to sodium bicarbonate administration to infants and children in shock or cardiac arrest. DESIGN: National survey study utilizing a self-administered questionnaire. SETTING: Thirteen Canadian pediatric tertiary care centers. SUBJECTS: Canadian pediatric critical care physicians, pediatric emergency physicians, and trainees in these subspecialties. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Survey items were evaluated based on Yes/No responses, frequency responses, and Likert scales. Overall response rate was 53% (151/284) with 49.0% (74/151) citing pediatric critical care as their primary practice. 82.0% of respondents (123/150) indicated they would administer sodium bicarbonate as part of ongoing resuscitation for septic shock, whereas 58.3% (88/151) would administer sodium bicarbonate in a cardiac arrest scenario (p=0.004). 47.3% (71/150) selected a pH threshold at or below which they would administer sodium bicarbonate (mean, 6.94±0.013; median, 7.00; range, 6.50-7.20; interquartile range, 6.90-7.00), whereas 20.5% (31/151) selected a base excess threshold (mean, -15.62±0.78; median, -16; range, -20 to -4; interquartile range, -20 to -14). Both pH and duration of resuscitation were strongly associated with the decision to administer sodium bicarbonate (p<0.0001). Respondents' perceptions regarding a colleague's likelihood of administering sodium bicarbonate to the same patient under the same circumstances reflect an acknowledgment of disparate practices with respect to sodium bicarbonate use. 53.0% (79/149) felt current American Heart Association guidelines help them in deciding whether to administer sodium bicarbonate to critically ill patients, and 84% would support a randomized trial. CONCLUSION: Differences of opinion exist among pediatric acute care physicians with respect to the timing and appropriateness of sodium bicarbonate administration during resuscitation. Most indicated they would support moving forward with a clinical trial.