Desempenho da nova versão do Paediatric Index of Mortality (PIM 3) em uma avaliação independente no sul do Brasil

Detalhes bibliográficos
Ano de defesa: 2016
Autor(a) principal: Fucks, Aline Acatrolli lattes
Orientador(a): Garcia, Pedro Celiny Ramos lattes
Banca de defesa: Não Informado pela instituição
Tipo de documento: Dissertação
Tipo de acesso: Acesso aberto
Idioma: por
Instituição de defesa: Pontifícia Universidade Católica do Rio Grande do Sul
Programa de Pós-Graduação: Programa de Pós-Graduação em Medicina/Pediatria e Saúde da Criança
Departamento: Faculdade de Medicina
País: Brasil
Palavras-chave em Português:
Área do conhecimento CNPq:
Link de acesso: http://tede2.pucrs.br/tede2/handle/tede/6969
Resumo: Objetive: To evaluate the performance of the latest version of the Pediatric Index of Mortality 3 (PIM 3) and compare it with the Pediatric Index of Mortality 2 (PIM 2) Intensive Care Unit Hospital of Pediatric São Lucas of the Catholic University of Rio Grande do South (PICU HSL / PUCRS). Methods: A prospective cohort study, longitudinal, descriptive observational census. The study included all patients aged 1 month to 18 years who were hospitalized in the Pediatric ICU HSL / PUCRS day 1 January 2015 to 31 December 2015. We excluded patients with less than 8 hours admission to the PICU with order no resuscitation or diagnosis of brain death in the first four hours of admission. Interventions: None. Results: Three hundred and ninety-nine children were included in the study during the period. We collected the data necessary to calculate the PIM 2 and PIM 3 in the first hour of admission to the PICU. The overall mortality rate was 3.5% and mortality index Standardized (SMR) was 1.13 (CI 95% 0,61-1,89), z = - 0,52 for the PIM 2 and 1.61 (CI 95% 0,88-2,70), z = - 2,15 to PIM 3. In relation to discrimination, there was a larger area under the ROC curve similar to PIM 2 and PIM 3 with 0.934 values (CI 95% 0.87-0.98) and 0.937 (CI 95% 0.88 to 0.98), respectively. The similarity between the mortality observed in the sample with expected mortality by calculating the PIM 2 in Hosmer-Lemeshow fit test showed a chi-square of 7.22 (p = 0.205). As for the PIM-3 showed a chi squared 12.88 (p = 0.025). Conclusion: Compared to the general population mortality by standardized mortality rate (SMR) we showed that the PIM 2 shows a predicted mortality very close to the observed mortality with no significant difference between them. The PIM 3 underestimates the mortality significantly. PIM 2 showed good calibration, not rejecting H0, while the PIM 3 does not possess good calibration, rejecting H0. By using the area under the ROC curve, both as PIM 2 PIM 3 showed a good discriminative power.