Modelo de Cox para eventos cardiovasculares recorrentes em pacientes sob diálise com covariáveis medidas no tempo

• Autor(a) principal: Mota, Thiago Santos [UNESP]
• Data de Publicação: 2012
• Outros Autores: Silveira, Liciana Vaz de Arruda [UNESP], Antunes, Aline de Araujo [UNESP]
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://jaguar.fcav.unesp.br/RME/fasciculos/v30/v30_n1/indice_v30_n1.php; http://hdl.handle.net/11449/140350
• topic: Marginal modeling; AIC; AG model and PWP model; Modelagem marginal; AIC; Modelo AG e modelo PWP

The aim is to evaluate the effect of the covariates measured over time for the occurrence of the (stroke) cardiovascular events recurring in patients under dialysis, consisting of 145 hemodialysis patients from the clinics hospital of the Medicine College, UNESP – Botucatu, SP, starting in 2008. The covariates considered in this study were clinical, nutritional, laboratory and dialysis. As the data treat the recurring events of the same type, the marginal modeling employing, in particular, the models AG (Andersen and Gill, 1982) and PWP (PRENTICE et al., 1981). The covariates were considered significant at 10% in the two models, using the selection criteria Collett (1994). It also presented the risk ratio of patients in relation to each covariate in both models and the AKAIKE information criteria (AIC). For the AG model the result of AIC was 499.9 and in the PWP model it was 425.68. All the results obtained in the R software. From the results, the PWP model was chosen because it had value lowest of the AIC and by the residuals analysis it showed good fit to the data. This study showed the importance of the marginal modeling as a way to model recurring events of the same type, and the utilization of the models mentioned in data which necessarily follow an order.