P0459 - Diagnosis and Management of Central Diabetes Insipidus in Critically Ill Children Widely Varies by Critical Care Medicine, Endocrinology and General Pediatrics Providers

Background/Purpose: Central Diabetes Insipidus (CDI) results from vasopressin (AVP) deficiency following suprasellar surgery, cerebral ischemia, or trauma. Definitive diagnosis of CDI is through fluid deprivation and AVP stimulation testing, but clinical features like polyuria, hypernatremia, high plasma and low urine osmolality are used to diagnose new-onset CDI in critically ill children. Management of new-onset CDI is complex and not standardized. We hypothesized that diagnosis and management of new-onset CDI in critically ill children widely varies by critical care medicine (CCM), endocrinology (EN) and general pediatric (GP) providers.

Methods: With IRB approval, a 10-question survey based on a post-operative neurosurgical case with new-onset CDI was sent to CCM and EN attendings, fellows, and advanced practice providers, and GP residents at a quaternary academic pediatric center. The survey was active from 8/2018 to 12/2018. Survey questions were single or multiple choice with option for free text. Difference in responses among groups was assessed with Chi2 analysis.

Results: Of 275 CCM, EN and GP providers surveyed, 158 (58%) responded. Among respondents, 29% were CCM, 20% were EN, and 51% were GP residents. For new diagnosis of CDI, CCM and EN compared to GP providers relied on urine output (p<0.001), while CCM favored serum sodium over EN and GP providers (p<0.001). Fewer CCM providers compared to EN and GP used serum and urine osmolality for diagnosis (both p<0.001). EN providers preferred IV fluids at 2/3 maintenance rate compared to GP and CCM (p=0.001). There were no differences among providers for the initial dose of vasopressin infusion (0.5 milli-U/kg/hr), frequency of infusion titration (every 30 min), or frequency of lab studies such as serum sodium (every 1-2 hrs) and urine output (every 30 min-1 hr). Compared to GP, more CCM and EN providers favored stopping titration based on urine output of 3 ml/kg/hr (p<0.001). More GP than CCM and EN providers perceived variability in CDI diagnosis (p=0.002). Compared to CCM, more EN and GP providers agreed that clinical pathway development would be useful for diagnosis and management of new-onset CDI (p<0.001).

Conclusion: Diagnosis and management of new-onset CDI in critically ill children widely varies by CCM, EN and GP providers. Efforts to standardize clinical care and improve education for diagnosis and management of CDI in critically ill children should be explored.