Wednesday, January 18, 2017

Survival Analysis

Survival Analysis

Academic Year 2016/2017

Learning outcomes

The aims of this course are to introduce the basic concepts of survival analysis and to explain and illustrate how survival analysis is applied to biomedical and social data 

Course contents

Introduction to the analysis of survival data in biomedical research, censoring and truncation. Estimation of the integrated survivorship functions with non parametric methods: Kaplan-Meier product-limit and lifetable methods.
Hazard function: definition ed estimate. Relationship between hazard, survivorship and density funcion in survival data analysis.
Comparison of survivorship functions: log-rank and Wilcoxon rank sum test.
Semiparametric Cox regression model: definition, assumptions, estimation and hypothesis tests; interpretation of estimated coefficients and residual analysis.
Extensions of the proportional hazard model: stratification and time-varying coefficients.
Parametric regression models. The exponential regression model. Weibull, Log-normal, log-logistic and generalized gamma regression models.
Accelerated failure time data.
Competing risk models. Frailty model.
Longitudinal data analysis.
Case studies

Readings/Bibliography

D. COLLETT, Modelling survival data in medical research, Chapman & Hall, 2003
D. W. HOSMER, S. LEMESHOW, S. MAY, Applied Survival Analysis: Regression Modeling of Time to Event Data, Wiley, New York, 2008.
E.T.LEE, Statistical Methods for Survival Data Analysis, Wiley, New York, 2003

Teaching methods

Traditional lessons
Seminar
Computer session

Assessment methods

The exam aims at testing the student's achievement of the following learning outcomes:
- deep knowledge of the statistical methods described and discussed during the lectures
- ability to use these methods in the analysis of survival data
- ability to use the obtained results for the quantitative interpretation of the studied data.
The exam consists of a lab test that aims to evalutate the ability to conduct statistical analysis of survival data. The student should positively pass this lab test in order to sustain the oral exam. The validity of the lab test is limited to the exam session and the following one.
The oral exam integrate the lab test in order to evaluate the achievment of the learning outcome. The final evaluation, given in marks out of 30, represents a mean of the results obtained in both the exams.

Teaching tools

Computer sessions

Office hours

See the website of Miglio Rossella

No comments:

Post a Comment