Validação no Children's Hospital Early Warning Score (CHEWS) em português para avaliação do risco de deterioração clínica em pacientes de unidades pediátricas

Abstract

Clinical deterioration is conceptualized as a condition that changes from one clinical condition to another worse. The Children’s Hospital Early Warning Score (CHEWS) is an algorithm that can detect early warning signs of this deterioration, improving the patient’s response time to avoid complications and optimize prognosis. The objectives were: to validate the content of CHEWS, Brazilian Portuguese version (CHEWS-Br); verify the inter-observer reliability of CHEWS-Br; and verify the association of the CHEWS-Br score with the patients' sociodemographic and clinical characteristics. Methodological study, divided into two phases: Phase 1- CHEWS-Br content validation, using the Content Validity Coefficient (CVC), with seven judges in the area of cardiac children's health; and Phase 2- interobserver reliability for assessing the risk of clinical deterioration in pediatric patients, from the application of CHEWS-Br, at the initial moment and after 4 and 6 hours, in a public tertiary hospital in Fortaleza-Ceará, with 163 patients, by means of application of a sociodemographic and clinical characterization instrument and CHEWS-Br. The application of CHEWSBr was carried out by two observers (A and B), simultaneously, to analyze the agreement between them. The analyzes of the proportions, between the observers, were made by the Chi-Square and Fisher-Freeman-Halton tests. The comparison of averages between observers was performed using the Mann-Whitney and Friendman tests. The analysis of the linear correlation of the score between the two observers, at each moment, was performed using Spearman's correlation coefficient. As a result, CHEWS-Br has valid content for early detection of signs of clinical deterioration in children and adolescents admitted to pediatric units, with the following total CVC: clarity of language (0.88), practical relevance (0.90) and theoretical relevance (0.91). Most of the children were male, with an average age of 5.6 ± 5.0 years, of brown color (64.4%), coming from the interior of Ceará or even from another state (64.4%). In the three evaluation times, there was no significant difference (p> 0.05) in all components of CHEWS-Br (neurological, cardiological, respiratory, professional concern and family concern); in the CHEWS evaluation algorithms (green, yellow and red); and in the mean of the CHEWS-Br scores, confirming agreement between observers A and B. There was a directly proportional linear correlation between the CHEWS-Br score between and among the observers, in the three times (p <0.0001). In the association of the mean scores of the CHEWSBr, in the three periods, with the sociodemographic characteristics of children and adolescents, there was a significant difference (p <0.05) only in the variable female gender. The odds ratio of the CHEWS-Br scores did not present a greater risk of clinical deterioration, according to sociodemographic variables. Children and adolescents who had previous ICU admissions or neurological disorders were 4.2 times more likely to risk clinical deterioration with CHEWS-Br scores> 3. It was concluded that CHEWS-Br is a valid and reliable algorithm to detect early clinical deterioration of hospitalized children and adolescents, allowing early intervention to prevent the worsening of the clinical condition.