Bibliographic citations
This is an automatically generated citacion. Modify it if you see fit
Flores, C., (2011). Modelo de regresión de Cox usando splines [Tesis, Universidad Nacional Mayor de San Marcos]. https://hdl.handle.net/20.500.12672/211
Flores, C., Modelo de regresión de Cox usando splines [Tesis]. PE: Universidad Nacional Mayor de San Marcos; 2011. https://hdl.handle.net/20.500.12672/211
@misc{renati/491036,
title = "Modelo de regresión de Cox usando splines",
author = "Flores Flores, Claudio Jaime",
publisher = "Universidad Nacional Mayor de San Marcos",
year = "2011"
}
Title: Modelo de regresión de Cox usando splines
Authors(s): Flores Flores, Claudio Jaime
Advisor(s): Bravo Quiroz, Antonio
Keywords: Análisis de regresión - Modelos matemáticos; Análisis de supervivencia (Biometría); Cáncer - Perú - Estadísticas
OCDE field: https://purl.org/pe-repo/ocde/ford#1.01.03
Issue Date: 2011
Institution: Universidad Nacional Mayor de San Marcos
Abstract: En muchos estudios clínicos es muy frecuente el uso de modelo de riesgos proporcionales de Cox; el cual asume riesgos proporcionales y restringe a que el logaritmo de la razón de riesgo sea lineal en las covariables, lo cual en muchos casos no se verifica. En este sentido, una forma funcional no lineal del efecto de las covariables puede ser aproximada por una función spline. En este trabajo, se presenta la metodología del modelo de regresión de Cox usando splines, particularmente regresión splines y P-splines, para aproximar la forma funcional no-lineal de los efectos de las covariables en la función de riesgo. Como una aplicación, se analiza los datos de pacientes con LNH para determinar los factores pronósticos para la supervivencia global. Los resultados muestran que el efecto de las covariables contínuas como hemoglobina, leucocitos, linfocitos y DHL presentan una forma funcional no lineal en el logaritmo de la razón de riesgo.
-- Palabras claves: Modelo de Cox, regresión splines, P-splines, LNH.
-- In many clinical studies, Cox proportional hazard model is very common to use, it assumes proportional hazard and restricts the log hazard ratio to be linear in the covariates; these asumptions can not be verified. In this way, a nonlinear functional form of the covariates effect can be approximated by a spline function. In this paper, we present the methodology and an application of Cox model using splines, particularly regression splines and P-splines, to approximate the nonlinear functional form of the effect of covariates on the hazard function. As an application, we analyse data from patients with NHL to determine prognostic factors for overall survival. These results show that the effect of continuous covariates as: hemoglobin, leukocytes, lymphocytes and LDH have a nonlinear form with the log hazard ratio. -- Keywords: Cox model, regression splines, P-spline, NHL.
-- In many clinical studies, Cox proportional hazard model is very common to use, it assumes proportional hazard and restricts the log hazard ratio to be linear in the covariates; these asumptions can not be verified. In this way, a nonlinear functional form of the covariates effect can be approximated by a spline function. In this paper, we present the methodology and an application of Cox model using splines, particularly regression splines and P-splines, to approximate the nonlinear functional form of the effect of covariates on the hazard function. As an application, we analyse data from patients with NHL to determine prognostic factors for overall survival. These results show that the effect of continuous covariates as: hemoglobin, leukocytes, lymphocytes and LDH have a nonlinear form with the log hazard ratio. -- Keywords: Cox model, regression splines, P-spline, NHL.
Link to repository: https://hdl.handle.net/20.500.12672/211
Discipline: Estadística
Grade or title grantor: Universidad Nacional Mayor de San Marcos. Facultad de Ciencias Matemáticas. Escuela Académico Profesional de Estadística
Grade or title: Licenciado en Estadística
Register date: 20-Aug-2013
This item is licensed under a Creative Commons License