Choosing a coverage probability for forecasting the incidence of cancer

Derek S. Young, Terence M. Mills

Research output: Contribution to journalArticlepeer-review

2 Scopus citations

Abstract

Loddon Mallee Integrated Cancer Service plays a key role in planning the delivery of cancer services in the Loddon Mallee Region of Victoria, Australia. Such planning relies on the accuracy of forecasting the incidence of cancer. Perhaps more importantly is the need to reflect the uncertainty of these forecasts, which is usually carried out through prediction intervals. Standard confidence levels (e.g., 90% or 95%) are typically employed when forecasting the incidence of cancer, but decision-theoretic approaches are available to help choose an optimal coverage probability by minimizing the combined risk of the interval width and noncoverage of the interval. We proceed with the decision-theoretic framework and discuss some general strategies for defining candidate loss functions for forecasting the incidence of cancer, such as the data we analyze for the Loddon Mallee Region.

Original languageEnglish
Pages (from-to)4104-4115
Number of pages12
JournalStatistics in Medicine
Volume33
Issue number23
DOIs
StatePublished - Oct 15 2014

Bibliographical note

Publisher Copyright:
Copyright © 2014 John Wiley & Sons, Ltd.

Keywords

  • cancer statistics
  • coverage probability
  • decision theory
  • prediction intervals
  • strategic planning

ASJC Scopus subject areas

  • Epidemiology
  • Statistics and Probability

Fingerprint

Dive into the research topics of 'Choosing a coverage probability for forecasting the incidence of cancer'. Together they form a unique fingerprint.

Cite this