Related Experiment Videos
Simple benchmark for mesothelioma projection for Great Britain.
María Dolores Martínez-Miranda1, Bent Nielsen2, Jens Perch Nielsen3
1Department of Statistics and Operations Research, University of Granada, Granada, Spain.
Occupational and Environmental Medicine
|June 2, 2016
Summary
Forecasting mesothelioma mortality is crucial. A simple response-only model predicts peak mortality around 2100 deaths in 2017, performing as well as complex models.
Area of Science:
- Epidemiology
- Biostatistics
Background:
- Forecasting future mesothelioma mortality is of significant interest.
- Available data primarily includes death records, lacking direct asbestos exposure measures.
Purpose of the Study:
- To compare the predictive performance of two Poisson models for mesothelioma mortality.
- To evaluate a response-only model against a dose-response model for forecasting.
Main Methods:
- Comparison of a response-only Poisson model (age-cohort specification) with a multinomial dose-response model.
- Epidemiologically motivated frequencies were used for the multinomial model.
Main Results:
- The response-only model predicted a 5% higher peak mortality compared to the dose-response model.
- Out-of-sample comparisons indicated slightly better performance for the response-only model.
Conclusions:
- Mesothelioma mortality is projected to peak around 2100 deaths in 2017 for males born in cohorts up to 1966 and under 90 years old.
- The response-only model serves as an effective and simple benchmark for mesothelioma mortality forecasting.