Correlation between Cosyntropin Stimulation Study and Disease Severity in Children with Fluid- and Catecholamine-Refractory Shock in the Pediatric and Cardiovascular Intensive Care Unit

Abstract

Background: The cosyntropin stimulation study (CSS) measures the patient’s ability to adequately mount a cortisol response. Clinically, CSS results may not be used to guide hydrocortisone use. The objective of this study was to examine how the CSS results are associated with clinical parameters, mortality/disease severity, and use of glucocorticoids in pediatric patients with catecholamine- and fluid-resistant shock. Methods: This was a retrospective cohort study of patients who had a CSS during 2009–2014 in the intensive care unit at a children’s hospital. Data collected included clinical variables, mortality, biochemical studies, and glucocorticoid use. PRISM III scores were used to determine the association between CSS results and disease severity. Adequate response to cosyntropin was defined as peak cortisol of 18 µg/dL or higher. Results: Of the 76 patients that underwent CSS, 68 (89%) had an adequate response to cosyntropin. There was a positive correlation between peak cortisol and PRISM III score (r = 0.45, r2 = 0.2). Glucocorticoid was administered in 52/76 (68%) despite several patients with normal CSS results. Conclusions: Sicker patients were more likely to have an adequate response to CSS. Clinically, glucocorticoid supplementation was not based on CSS results. Further prospective studies are needed to elucidate if CSS is a valuable clinical tool.