High-risk PICU refers to patients with sepsis, traumatic brain injury, or burn or those requiring mechanical ventilation and/or vasopressors. Data were derived from multiple sources.^11-32

Many of these retrospective studies also investigated whether there was an association between hyperglycemia and poor outcome. Reports over time have indicated that there is indeed an association between hyperglycemia and morbidity and mortality. Similar to the adult literature, the pediatric literature in general has reported a positive association between ventilator days, infections, use of high-frequency ventilation, use of inotropes, and renal insufficiency and failure. In the general PICU population, mortality rates are up to 6 times higher in patients with hyperglycemia.^10,13,20 In mechanically ventilated patients and those with septic shock, mortality is 2 to 3 times higher in groups with hyperglycemia.^11,12 In a study by Cochran et al,²⁷ a BG of greater than 300 mg/dL in patients with TBI was predictive of nonsurvival.

Figure 12-2. Incidence of hyperglycemia in a PICU according to risk factor

Abbreviations: CRRT, continuous renal replacement therapy; ECMO, extracorporeal membrane oxygenation; MV, mechanical ventilation; VP, vasopressors. Data were derived from multiple sources where hyperglycemia was defined as 2 consecutive blood glucose readings of greater than 140 mg/dL.^27,32

Recommendations for Glycemic Control in Critical Care and Practice in Pediatric ICUs

Relatively soon after the publication of van den Berghe’s 2001 article (arguably with little confirmatory reports), a number of prominent national and international oversight groups adopted recommendations for routine tight glycemic control. The term tight glycemic control (or TGC) was used in the van den Berghe studies to describe the patients who had blood glucose controlled in the range of 80 to 110 mg/dL (vs the “conventionally” controlled group whose BG was kept in the range of 180-200 mg/dL). By 2004, organizations such as the Institute for Healthcare Improvement, the American Diabetes Association, and the American College of Endocrinologists strongly recommended TCG for all ICU patients.³³ In the 2004 guidelines of the Surviving Sepsis Campaign, the Society of Critical Care Medicine suggested using an insulin infusion to keep BG less than 150 mg/dL.³⁴ It appears that most adult ICUs adopted some form of glycemic control (if not the true TGC approach), and hospital systems began to use glycemic control as a critical quality and compliance metric. In 2009, the NICE-SUGAR study was published, which questioned some of the prevailing notions and raised more questions regarding routine TGC. This was a multicenter study that enrolled more than 6,000 critically ill adults and randomized patients to blood glucose control of either 81 to 108 mg/dL or 180 mg/dL or less.⁷ In reality the average BG for each group was 107 and 142 mg/dL, with substantially overlapping ranges. This study found a small but statistically significant increase in mortality in the TGC group (27.5%) versus the conventional group (24.9%); in addition, there was a significantly higher incidence of hypoglycemia in the TGC group, 6.8% versus 0.5%, respectively. Following this study, the optimal goal for glucose control was significantly questioned, and some extended the discussion to the utility of glycemic control at all (although this certainly was not supported by the approach or findings of this study). Following the NICE-SUGAR findings and evaluation of other studies,^6,7,35 the oversight committees relaxed their recommendation. The Institute for Healthcare Improvement now recommends maintaining BG less than 180 mg/dL, the American Diabetes Association and American College of Endocrinologists recommend keeping BG in the range of 140 to 180 mg/dL, and the Surviving Sepsis Campaign recommends a BG of less than 150 mg/dL if the BG is greater than 180 mg/dL in patients with shock.^36,37

Throughout this time it has been unclear to what extent pediatric intensivists have adopted glycemic control measures philosophically or in practice. A 2008 survey of adult and pediatric intensivists suggested that the majority of pediatric intensivists practice glycemic control. Specifically, about 90% of pediatric intensivists said that the desired BG for their patients is less than 180 mg/dL with approximately 75% indicting the goal range between 80 and 140 mg/dL.³⁸ Approximately 85% of pediatric intensivists consider hypoglycemia to be more dangerous than hyperglycemia. Although this suggests that glycemic control may be routine in PICUs, a more focused survey published in 2009 suggested that this is not the case.³⁹ Specifically, in only 3 of 30 centers did physicians believe that all children with critical illness hyperglycemia should undergo glycemic control, and in two-thirds of centers physicians believed that fewer than 25% of their patients with hyperglycemia were actually managed. Fewer than 10% of centers had adopted a standard approach to manage hyperglycemia, and approximately 70% listed fear of hypoglycemia as the main barrier to initiating a routine, protocol-based approach to standard management.

In recent years, a number of groups have reported that pediatric-specific glycemic control protocols can be adopted that appear to be safe and to effectively manage glucose levels (Table 12-1). In 2008, experience with a physician-initiated, nurse-managed protocol was published. In this approach, patients with select “risk factors” (ie, MV, vasopressors, or CRRT) have BG checked at least every 12 hours; if there is a reading of greater than 140 mg/dL, a second BG is checked. If the second reading is confirmed greater than 140 mg/dL, the nurse automatically starts an insulin infusion and titrates it via a paper protocol. In about 70% of days in which patients received insulin therapy, all BG checks were less than 160 mg/dL, and hypoglycemia (<40 mg/dL) occurred in 4% of patients.³⁰ This protocol is undergoing external validation and is being tested in additional PICUs via a study sponsored by the National Institute of Diabetes and Digestive and Kidney Disease (NCT01116674). Preliminary results indicate good glycemic control with approximately 5% incidence of hypoglycemia (M. Rigby, unpublished data, August 2012). In 2009, Verhoeven et al⁴⁰ reported a PICU-specific protocol in which insulin was initiated when BG was greater than 145 mg/dL, and in 50 patients there were no episodes of severe hypoglycemia. Most recently, Faraon-Pogaceanu et al⁴¹ published experience with paper and electronic protocols. The initiating BG for both protocols was a reading of greater than 140 mg/dL, and BG was controlled at 90 to 119 mg/dL under the paper-based protocol and 80 to 110 mg/dL under the computerized approach. Although BG seemed reasonable with both approaches, hypoglycemia was 10% and 20%, respectively. Taken together, studies provide strong data that routine approaches to identify and control hyperglycemia can be successfully implemented in pediatric ICUs.

Related posts:

Neuroscience in Pediatric Critical Care Considerations in the Determination of Death Critical Care in Resource-Poor Environments: The Reality and the Challenge for Tomorrow, Leadership for Today Role of Simulation in Improving Patient Outcomes Costs and Reducing Waste in Pediatric Critical Care