LAUSR

Abstract The aim of this study was to assess the bedside brain function monitoring of color density spectral array (CDSA) for early prognostic evaluation of coma patients in pediatric intensive care unit (PICU). Forty-two consecutive pediatric coma patients were enrolled. The individual conscious state was evaluated according to the Glasgow coma scale (GCS). CDSA parameters including CDSA pattern (CDSAP), sleep–wake cycle (SWC), sleep stage (SS), and drug-induced fast wave activity (DIFWA) were recorded. Three months later, prognosis was evaluated according to pediatric cerebral performance category (PCPC) score, based on which the patients were divided into FP-group (favorable prognosis) and PP-group (poor prognosis). The changeable type of CDSAP, appearance of SWC, SS, and DIFWA were significantly correlated with favorable prognosis. Both GCS and SWC were significantly correlated with the prognosis. However, there was substantial overlap in GCS between FP-group and PP-group. Although the absence of SWC was statistically an independent risk factor for poor prognosis but with a high false positive rate (0.143), a linear logistic regression showed the odds ratio of GCS for predicting prognosis was 0.93 (95% confidence interval: 0.48–1.80; P = .83) and that of SWC was 0.12 (95% confidence interval: 0.03–0.47; P = .03). Furthermore, the absence of SWC was correlated with poor prognosis in nonintracranial infection patients. Our study found that several CDSA factors are associated with prognosis of coma patients in PICU. SWC may be a potential indicator for evaluating the prognosis of coma patients in PICU.