Modeling correlated healthcare costs

Joanne K. Daggy, Joseph Thomas, Bruce A. Craig

Research output: Contribution to journalReview article

8 Scopus citations

Abstract

Accurate estimation and prediction of healthcare costs play crucial roles in decisions made by healthcare agencies on policy and resource allocation. Development of a cost model allows these decision-makers the opportunity to investigate the impact of different policies and/or allocations of resources. With increased subject-specific information, longitudinal studies and the breakdown of total costs into categories comes the need for healthcare cost models to account for correlation. In this article, we review the statistical models used to fit joint costs, emphasizing the use of copulas as a flexible and relatively straightforward approach to move from marginal to joint modeling.

Original languageEnglish (US)
Pages (from-to)101-111
Number of pages11
JournalExpert Review of Pharmacoeconomics and Outcomes Research
Volume11
Issue number1
DOIs
StatePublished - Feb 1 2011

    Fingerprint

Keywords

  • copulas
  • excess zeros
  • healthcare costs
  • heavy-tailed distributions
  • two-part models

ASJC Scopus subject areas

  • Pharmacology (medical)
  • Health Policy

Cite this