Related Experiment Videos
Modelling geographically referenced survival data with a cure fraction
Freda Cooner1, Sudipto Banerjee, A Marshall McBean
1Division of Biostatistics, School of Public Health, University of Minnesota, Minneapolis, Minnesota 55455-0392, USA.
Statistical Methods in Medical Research
|August 5, 2006
Summary
This study introduces a Bayesian cure model to analyze survival data with spatial patterns. It helps understand regional survival variations and patient prognosis in public health research.
Area of Science:
- Biostatistics
- Spatial Epidemiology
- Public Health
Background:
- Geographical Information Systems (GIS) enable spatial analysis of patient data.
- Detecting regional survival variations is crucial for public health.
- Cure models are needed for diseases with significant recovery rates.
Purpose of the Study:
- To propose a Bayesian modeling framework for spatial survival data.
- To incorporate cure models for diseases with potential recovery.
- To analyze areally referenced survival data accounting for spatial clustering.
Main Methods:
- Utilizing a Bayesian framework for spatial modeling.
- Applying a general class of cure models (Cooner et al.).
- Fitting models using standard Bayesian software (e.g., WinBUGS).
Main Results:
- The proposed framework models spatial associations in survival data.
- It accounts for cure proportions in the population.
- Offers an alternative to traditional proportional hazards models.
Conclusions:
- The Bayesian cure model framework effectively analyzes spatial survival data.
- It aids in understanding regional health disparities and disease prognosis.
- Provides a flexible tool for public health research.