Is Serum Bicarbonate Level Associated With Negative Outcomes in Pediatric Patients? A Retrospective Cohort Study Naveen Poonai Prakash Chandra, MD,*†‡ David Mainprize, BSc,* Carolyn Travers, BSc,* Lilian Lee Yan Vivas, MD,* Peter Tryphonopoulos, MD,§ Gurinder Sangha, MD,*† Ryan Arbeau, MD,* Jamie Seabrook, PhD,‡|| Michael Miller, PhD,† Amrita Sarpal, MD,† and Rodrick Lim, MD*† Objectives: Early identification of children at risk for adverse outcomes is important. Serum bicarbonate is easily collected and widely available. We described the relationship between bicarbonate and adverse outcomes in children presenting to the emergency department (ED). Methods: We conducted a retrospective cohort study of children aged 0 to 17 years from January 1, 2007, to December 31, 2011, who had a serum bicarbonate measured in the ED. Primary outcome was the predictive abil- ity of bicarbonate for the individual components of the composite outcome that included at least one of the following: intensive care unit admission, assisted ventilation, inotropic support, cardiopulmonary resuscitation, or death. Secondary outcome was the relationship between bicarbonate level of greater and less than 13 mEq/L and the composite outcome. Results: We reviewed 16,989 charts, of which 432 had an adverse out- come. Receiver operating characteristic curve analysis showed that a bicar- bonate level of less than 18.5 mEq/L predicted inotropic support with an area under the curve of 0.69 (95% confidence interval [CI], 0.60–0.77; P < 0.001) and death with an area under the curve of 0.75 (CI, 0.66–0.85; P < 0.001). Significantly more patients with bicarbonate level of less than 13 mEq/L had at least 1 adverse outcome compared with those with bicar- bonate level of greater than 13 mEq/L (4.4% vs 2.5%, P = 0.001), odds ratio 1.96 (95% CI, 1.3–2.97). Conclusions: Among children presenting to the ED, bicarbonate level of 18.5 mEq/L had fair specificity in predicting inotropic support and death. Negative outcomes are significantly associated with bicarbonate level of less than 13 mEq/L. Bicarbonate should routinely be measured in children at risk of clinical deterioration. Key Words: bicarbonate, inotropic support, respiratory failure, death (Pediatr Emer Care 2016;00: 00–00) I n the emergency department (ED), it is important to accurately identify children at risk for negative outcomes, so timely man- agement can be implemented. 1,2 Noninvasive methods of risk as- sessment in the ED have usually relied on routine measurement of physiologic parameters such as heart rate, respiratory rate, blood pressure, oxygen saturation, and urine output. 3,4 However, in iso- lation, these indices fail to reliably predict mortality or hospital length of stay in critically ill patients. 3,5 Furthermore, because of the incredible physiologic reserve children possess, they may not exhibit obvious alterations in physiological parameters until shortly before cardiovascular collapse ensues. 6 Markers of meta- bolic acidosis such as pH, arterial base deficit/excess, and lactate are frequently measured in intensive care settings because of their clinical relevance and widespread availability. 4,7–9 Unfortunately, determination of an arterial base deficit (ABD)/excess is painful and invasive and requires a skilled operator. 10 In addition, it carries the risk of bleeding, blood-borne infections, pseudoaneurysm, distal ischemia, or thromboembolism, 10–14 factors that may limit its routine determination in children. The serum bicarbonate level can be easily obtained as part of a serum chemistry panel and offers the advantages of ease of collection and lower costs and does not require an arterial punc- ture or catheterization, thereby overcoming the drawbacks of an ABG. 10,14,15 Although published data are limited, studies have suggested that it is equivalent to ABD in predicting significant metabolic acidosis. 14,15 Although it may be intuitive that a lower bicarbonate level implies an increased risk of a poor outcome, this has not been quantified statistically. More importantly, not all cen- ters include bicarbonate in their laboratory's electrolyte panel. An estimate of a bicarbonate threshold, less than of which indicates an increased risk of an adverse outcome, would greatly assist risk stratification of the pediatric patient and prompt emergency phy- sicians to consider bicarbonate in their assessment of children. The objective of this study is to examine whether low se- rum bicarbonate levels in children who present to the ED were as- sociated with any of the following adverse outcomes: pediatric intensive care unit (PICU) admission, inotropic support, assisted ventilation, cardiopulmonary resuscitation (CPR), or death. The results may enable more timely evaluation, intervention, and con- sultation in children with low bicarbonate levels, possibly leading to improved clinical outcomes. Our hypotheses were 2-fold: (1) a threshold bicarbonate level would predict at least 1 of the 5 ad- verse outcomes, and (2) a previously described threshold bicar- bonate level of 13 mEq/L 16,17 would be associated with a significantly greater frequency of adverse outcomes as a composite. MATERIALS AND METHODS Design and Setting We conducted a retrospective cohort study involving a chart review of all patients' encounters for children aged 0 to 17 years presenting to the ED of the Children's Hospital, London Health Sciences Centre, from January 1, 2007, to December 31, 2011, who had a serum bicarbonate level measured in the ED. Our center has approximately 40,000 annual visits and serves as the pediatric regional referral base for a population of more than 1.2 million. The criteria for admission to our PICU are the following: assisted ventilation, inotropic support, continuous monitoring of oxygen saturation, renal replacement therapy, or unstable vital signs, sug- gestive of uncompensated shock. The attending pediatric emer- gency physician made all decisions regarding investigations. Charts with incomplete data pertaining to the primary outcome From the *Division of Emergency Medicine, †Department of Paediatrics, Schulich School of Medicine and Dentistry, Western University; ‡Children's Health Research Institute, London Health Sciences Centre; §Department of An- esthesia, Schulich School of Medicine and Dentistry, Western University; and || Division of Food and Nutritional Sciences, Brescia University College at West- ern University, London, Ontario, Canada. Disclosure: The authors declare no conflict of interest. Reprints: Naveen Poonai Prakash Chandra, MSc, MD, FRCPC, Paediatric Emergency Department, Children's Hospital, London Health Sciences Centre, Schulich School of Medicine and Dentistry, 800 Commissioners Rd E, London, Ontario, N6A 5W9 (e‐mail: naveen.poonai@lhsc.on.ca). Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0749-5161 ORIGINAL ARTICLE Pediatric Emergency Care • Volume 00, Number 00, Month 2016 www.pec-online.com 1 Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.