# BGC1: Survival and event history analysis

The universities of Stockholm, Oslo, Helsinki and Copenhagen
jointly organize a
Biostatistics Graduate Course Network. The
universities in Newcastle and Gent also take part in this
collaboration. Master- and Phd-students from the collaborating
universities as well as from other universities are welcome to take
part in the BGC-Network courses.
A course in *Survival and event
history analysis* will take place at the University of Oslo
19-23 March and 7-11 May 2012. Each of the two weeks the course starts Monday at 10.30 am and ends Friday at 2.30 pm.
(See below for a detailed program for the first week of the course.)

### General description of the course

The course gives a broad introduction to concepts and methods in
survival and event history analysis. Topics covered include: counting
process models for survival and event history data; basic results for
martingales and stochastic integrals; the Nelson-Aalen and
Kaplan-Meier estimators; the log-rank test and other non-parametric
tests; multi-state models; Cox's proportional hazard regression;
additive hazards regression; parametric survival models;
the frailty concept; frailty models, marginal models, and dynamic models for
multivariate survival data; dynamic path analysis and comments on causality; statistical computing using R. The methods
covered in the course have applications in epidemiology, clinical
medicine, demography, economics, insurance, sociology, and technical
reliability.
The goals for the course are:

- To understand the counting process formulation underlying modern survival and event history analysis, and be able to use it to derive heuristically the statistical properties of the most used estimators and test statistics.
- To understand the assumptions underlying the commonly used methods in survival and event history analysis, and be able to choose a suitable model formulation in a practical setting.
- To be able to use statistical software to analyse survival and event history data and to interpret the result of an analysis.

The course gives 7.5 ECTS credits, see here for more information.
(The BGC-Network cannot guarantee that these credits points are accepted by a specific masters- or PhD-program. That is up to each program to decide.)

### Course literature

Aalen, Borgan & Gjessing (2008): *Event History Analysis: A Process Point of View*. Springer-Verlag, New York

### Teachers

Lecturers at the course include Odd O. Aalen (Oslo), Ørnulf Borgan (Oslo), and Robin
Henderson (Newcastle)

### Schedule

The main schedule for the course is as follows:
- Deadline for application: 1 February
- Notification of acceptance: 6 February
- Section 1: Introductory reading at home university. A reading list is provided here.
- Section 2: Lectures 19-23 March at the University of Oslo.
Program.
- Section 3: Compulsory homework at the home university that counts 20% of the final mark.
The text for the homework is given
here.
The deadline for the project is Friday April 20th.
- Section 4: Lectures 7-11 May at the University of Oslo.
Program.
- Section 5: Project exam at the home university that counts 80% of the final mark.
The text for the project
exam is given
here.
The deadline for the project exam is Friday June 15th.

### Computing

For the practical exercises, the participants need to bring their own laptop with the statistical language R installed.

### Application and prerequisites

An application to take part in the course should be sent by e-mail to Ørnulf Borgan
(borgan@math.uio.no) as soon as possible and no later than 1
February, 2012,
and should contain the following informaton:
- Name
- Date of birth
- Email address
- University affiliation
- Information on attained degrees and courses
- Current study (master/PhD)
- Brief motivation for wanting to participate in the course

The number of participants at the course is limited to 25 students. In the (unlikely) case of overbooking students from the universities participating in the BGC-Network will be given priority.
As prerequisite for the BGC-Network courses a minimum of 80 ECTS
credits in mathematics and statistics is suggested, including
probability theory, statistical inference and the theory of linear
models. Experience of programming and software such MATLAB, R, S-Plus
and/or STATA is an advantage.

### Practical information

The course will take place in lecture room "Alfa-Omega" at the top floor of the
Kristen Nygaard's building (formerly the Informatics building) at the University of Oslo. Kristen Nygaard's building is shown on
this map of (parts of) the
university campus.
Information on how to get from Oslo airport
to downtown Oslo and from downtown Oslo to the University of Oslo and
Kristen Nygaard's building is given here.

The participants have to organize their own travel and accommodation.
Information on possible accommodation in Oslo is given below.

### Costs, travel and accommodation

The course is offered free of charge. But costs for student travel and accommodation are not covered by the
network programme, and each student should organize his or her own
travel and accommodation.
Information on possible accommodation in Oslo is given
here.

Contact information:
borgan@math.uio.no
08.02.2012