A carregar...
Publicação
Desempenho da PRISM e PRISM IV numa unidade de cuidados intensivos pediátricos
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 1.58 MB | Adobe PDF |
Orientador(es)
Resumo(s)
Introdução: As escalas preditivas de mortalidade, ou prognósticas, têm como finalidade avaliar o risco de mortalidade intra-hospitalar, com base em valores de variáveis fisiológicas das primeiras horas de admissão em cuidados intensivos. Permitem quantificar o estado de saúde à data da hospitalização, independentemente do diagnóstico. São ainda úteis para avaliação da qualidade dos cuidados prestados.
Objetivos: O objetivo deste estudo foi avaliar o desempenho das escalas prognósticas Pediatric Risk of Mortality (PRISM) e Pediatric Risk of Mortality IV (PRISM IV) na Unidade de Cuidados Intensivos Pediátricos do Hospital de Santa Maria (UCIPed).
Métodos: Estudo observacional retrospetivo com revisão dos internamentos ocorridos entre 1 de janeiro de 2022 e 31 de dezembro de 2022 na UCIPed. Foram excluídos os doentes internados para vigilância ou monitorização após procedimentos eletivos. O
desempenho das escalas foi analisado com recurso à discriminação, calibração e taxa de mortalidade padronizada (SMR). A discriminação foi avaliada pela área abaixo da curva
(AUC) ROC (Receiver Operating Characteristic) e o teste de Hosmer-Lemeshow foi usado para medir a calibração.
Resultados: Duzentos e dezanove pacientes foram incluídos no estudo e a mortalidade foi de 9,6% (21 óbitos). As AUC com intervalos de confiança (IC) de 95% mostraram boa
discriminação por parte de ambas as escalas, com resultados de 0,875 (IC 95%: 0,782 –
0,967) e 0,896 (IC 95%: 0,804 – 0,987) para PRISM e PRISM IV, respetivamente. A PRISM
sobrestimou a mortalidade [SMR 0,488 (IC 95%: 0,301 – 0,719)] e mostrou calibração inadequada (p < 0,001), enquanto a PRISM IV mostrou melhor desempenho em termos
de estimativa da mortalidade [SMR 0,871 (IC 95%: 0,538 – 1,283)] e de calibração (p =
0,636).
Conclusão: A escala PRISM IV apresentou calibração e discriminação adequadas pelo que constitui a ferramenta mais útil para a avaliação prognóstica nesta unidade.
Introduction: Mortality scores, also known as prognostic, were designed to assess the risk of in-hospital mortality based on physiological variables measured during the first hours of admission to intensive care. These scores allow quantification of a patient's health status at the time of hospitalization, regardless of the clinical diagnosis, and are also useful for evaluating the quality of care provided. Objectives: The aim of this study was to assess the performance of the prognostic scores Pediatric Risk of Mortality (PRISM) and Pediatric Risk of Mortality IV (PRISM IV) in the Pediatric Intensive Care Unit of Hospital de Santa Maria (UCIPed). Methods: Retrospective observational study of all non-elective UCIPed admissions, between January 1st 2022 and December 31st 2022. Patients admitted for surveillance or monitoring after elective procedures were excluded. The performance of the scores was analyzed using discrimination, calibration and the standardized mortality ratio (SMR). Discrimination was assessed by the area under the Receiver Operating Characteristic (ROC) curve (AUC) and the Hosmer-Lemeshow test was used to measure calibration. Results: Two-hundred and nineteen patients were included, 21 (9,6%) of which died. The AUC with 95% confidence intervals (CI) indicated good discrimination by both scores, with results of 0,875 (95% CI: 0,782 – 0,967) and 0,896 (95% CI: 0,804 – 0,987) for PRISM and PRISM IV, respectively. PRISM overestimated mortality [SMR 0,488 (95% CI: 0,301 – 0,719)] and showed poor calibration (p < 0,001), while PRISM IV performed better in terms of mortality prediction [SMR 0,871 (95% CI: 0,538 – 1,283)] and calibration (p = 0,636). Conclusion: PRISM IV showed adequate calibration and discrimination, making it a more useful prognostic tool in this unit.
Introduction: Mortality scores, also known as prognostic, were designed to assess the risk of in-hospital mortality based on physiological variables measured during the first hours of admission to intensive care. These scores allow quantification of a patient's health status at the time of hospitalization, regardless of the clinical diagnosis, and are also useful for evaluating the quality of care provided. Objectives: The aim of this study was to assess the performance of the prognostic scores Pediatric Risk of Mortality (PRISM) and Pediatric Risk of Mortality IV (PRISM IV) in the Pediatric Intensive Care Unit of Hospital de Santa Maria (UCIPed). Methods: Retrospective observational study of all non-elective UCIPed admissions, between January 1st 2022 and December 31st 2022. Patients admitted for surveillance or monitoring after elective procedures were excluded. The performance of the scores was analyzed using discrimination, calibration and the standardized mortality ratio (SMR). Discrimination was assessed by the area under the Receiver Operating Characteristic (ROC) curve (AUC) and the Hosmer-Lemeshow test was used to measure calibration. Results: Two-hundred and nineteen patients were included, 21 (9,6%) of which died. The AUC with 95% confidence intervals (CI) indicated good discrimination by both scores, with results of 0,875 (95% CI: 0,782 – 0,967) and 0,896 (95% CI: 0,804 – 0,987) for PRISM and PRISM IV, respectively. PRISM overestimated mortality [SMR 0,488 (95% CI: 0,301 – 0,719)] and showed poor calibration (p < 0,001), while PRISM IV performed better in terms of mortality prediction [SMR 0,871 (95% CI: 0,538 – 1,283)] and calibration (p = 0,636). Conclusion: PRISM IV showed adequate calibration and discrimination, making it a more useful prognostic tool in this unit.
Descrição
Trabalho Final do Curso de Mestrado Integrado em Medicina, Faculdade de Medicina, Universidade de Lisboa, 2023
Palavras-chave
Pediatric Risk of Mortality (PRISM) Cuidados intensivos pediátricos Escalas de gravidade Prognóstico Mortalidade Pediatria