LAUSR

In this issue of Pediatric Critical Care Medicine, Stenson et al (1) report a retrospective analysis demonstrating that plasma chloride concentrations greater than or equal to 110 mmol/L are associated with a complicated course and mortality in children less than or equal to 10 years old with septic shock. Additionally, they show the association of a mean chloride greater than or equal to 110 mmol/L and an increased risk of mortality in this patient group. The study is novel in that it is the first study to specifically investigate the risk of poor outcomes in pediatric septic patients with hyperchloremia. Importantly, the authors adjusted for baseline risk of mortality as defined by the Pediatric Sepsis Biomarker Risk Model (2). The study sample size is robust with 890 subjects included from a 29-center pediatric septic shock database, providing good generalizability of the study findings. Strikingly, similar results have been reported in two studies in adults, both with a similar threshold for increased risk for mortality (3, 4). Underlying acute disease and severity of illness may increase the likelihood of uncovering the hazard of hyperchloremia. Patients in the analysis by Stenson et al (1) had septic shock, a condition with high mortality risk and, importantly, an increased susceptibility for hyperchloremia. Primarily, as a result of the routine use of 0.9% NaCl for fluid resuscitation in the ICU, patients with sepsis are at risk for hyperchloremia (5). Neyra et al (3) reported in critically ill adult septic patients with hyperchloremia at admission (≥ 110 mmol/L) an independent association of a higher chloride levels in the first 72 hours of admission with increased hospital mortality. Animals with experimental sepsis randomized to saline resuscitation (vs fluid with a more physiologic chloride concentration) had decreased survival and increased kidney injury by creatinine as well as histology (6). Unfortunately, Stenson et al (1) do not report the amount and type of fluid administered to patients in their study. However, we can assume that it is due primarily to exogenous chloride; 0.9% NaCl resuscitation being the most likely source. Numerous physiologic explanations have been proposed for the association of hyperchloremia and poor outcomes. The role of increased circulating levels of interleukin (IL)–6, IL-10, and tumor necrosis factor in a proinflammatory response as a result of hyperchloremic metabolic acidosis has been demonstrated in animal models of sepsis (7). However, in both animal and human studies, reductions in glomerular filtration rate and renal blood flow have been demonstrated with exposure to chloride loads (8, 9). Acute kidney injury (AKI) was not included as an outcome in the study by Stenson et al (1), which would have been useful given that prior investigations have shown increased rates of AKI and need for dialysis in patients receiving saline when compared with fluids with more physiologic chloride concentrations (10–12). The association of chloride levels and AKI in the pediatric critically ill population has not yet been reported. The study findings by Stenson et al (1) have important implications for the pediatric critical care practitioner. There is a significant body of evidence in the pediatric literature indicating that the routine use of hypotonic fluids in hospitalized patients contributes to the development of iatrogenic hyponatremia, leading to likely an increased use of isotonic fluids in patients even beyond the acute resuscitation period (13, 14). In choosing fluids not only for resuscitation, but also for maintenance of IV infusions, practitioners should carefully consider the potential risks of 0.9% NaCl. The early initiation of enteral feeding in pediatric intensive care patients has been shown to lead to improved nutrition and improvement in clinical outcomes (15). For patients requiring IV fluid administration (including parenteral nutrition), pediatric intensivists and other clinicians should be monitoring for hyperchloremia and strongly consider the use of more physiologic solutions in an effort to avoid iatrogenic hyperchloremic acidosis. Ringer’s lactate and Plasma-Lyte 148 (Baxter Worldwide, Deerfield, IL) are examples of balanced crystalloid solutions that more closely represent the physiologic composition of plasma with less chloride when compared with 0.9% NaCl. Importantly, regardless of the decision to provide 0.9% NaCl or a more physiologic electrolyte solution, a careful consideration of the indication for IV fluid administration and the amount of such fluid should occur similar to the administration of any other drug. Despite its near universal use in PICUs, 0.9% saline carries a risk for hyperchloremia; a condition we can no longer ignore.