EUVIMEDObjevte EUVIMED
Open access

Improved risk prediction via cross‐domain calibration in a retrospective case‐control study

Ge Zhao iD, Yanyuan Ma, Yaqi Cao

Canadian Journal of Statistics · 2026-08-04

DOI: 10.1002/cjs.70058 ↗

At a glance

Journal article in Canadian Journal of Statistics, published online on 2026-08-04. Written by Ge Zhao and 2 co-authors. Institution: Department of Mathematics and Statistics, Portland State University, Portland, U.S. 45 references. The full text is freely available in PubMed Central. Funded by: NINDS NIH HHS.

AbstraktSubjectsBibliografické údajeCiteRelated studies

Abstract (original language)

Given a well‐calibrated standard model involving standard predictors, we propose a constrained maximum likelihood approach to incorporate the biomarker sampled from a different population without having to correctly model them to improve the prediction. The constraints ensure that the averaged risk with the biomarker included is as close as that from the standard model. Our method allows different distributions of the standard risk predictors between the target population and the source population of the study data, and relies on the study data to inform the relationship between the biomarker and the standard predictors. We develop the large‐sample theory of our method for parameters and measures for the predictive accuracy, and perform simulation studies to assess the finite‐sample performance of our method. We apply our method to analyze a case–control study of breast cancer to develop a model that includes both standard breast cancer risk predictors and the biomarker mammographic density.

Subjects

Author keywords

62P1090C30CalibrationCase-control dataConstrained optimizationPrimary 62–02Retrospective likelihoodSecondary 90–01

Authors and institutions

  • Department of Mathematics and Statistics, Portland State University, Portland, U.S.
  • Department of Statistics, Penn State University, University Park, U.S.
  • Department of Statistics, School of Science, Minzu University of China, Beijing, China.
  • Department of Biostatistics, Epidemiology and Informatics, University of Pennsylvania, Philadelphia, U.S.

ORCID profiles: Ge Zhao

Funding

  • NINDS NIH HHS

Cite

APA

Zhao, G., Ma, Y., & Cao, Y. (2026). Improved risk prediction via cross‐domain calibration in a retrospective case‐control study. Canadian Journal of Statistics. https://doi.org/10.1002/cjs.70058

Vancouver

Zhao G, Ma Y, Cao Y. Improved risk prediction via cross‐domain calibration in a retrospective case‐control study. Canadian Journal of Statistics. 2026. doi:10.1002/cjs.70058

BibTeX
@article{Zhao2026,
  author = {Zhao, Ge and Ma, Yanyuan and Cao, Yaqi},
  title = {Improved risk prediction via cross‐domain calibration in a retrospective case‐control study},
  journal = {Canadian Journal of Statistics},
  year = {2026},
  doi = {10.1002/cjs.70058}
}
Start for free →