Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022:193:171-198.
Epub 2022 Nov 28.

Meta-analysis of individualized treatment rules via sign-coherency

Jay Jojo Cheng  1 , Jared D Huling  2 , Guanhua Chen  1
Affiliations

Meta-analysis of individualized treatment rules via sign-coherency

Jay Jojo Cheng et al. Proc Mach Learn Res. 2022.

Abstract

Medical treatments tailored to a patient's baseline characteristics hold the potential of improving patient outcomes while reducing negative side effects. Learning individualized treatment rules (ITRs) often requires aggregation of multiple datasets(sites); however, current ITR methodology does not take between-site heterogeneity into account, which can hurt model generalizability when deploying back to each site. To address this problem, we develop a method for individual-level meta-analysis of ITRs, which jointly learns site-specific ITRs while borrowing information about feature sign-coherency via a scientifically-motivated directionality principle. We also develop an adaptive procedure for model tuning, using information criteria tailored to the ITR learning problem. We study the proposed methods through numerical experiments to understand their performance under different levels of between-site heterogeneity and apply the methodology to estimate ITRs in a large multi-center database of electronic health records. This work extends several popular methodologies for estimating ITRs (A-learning, weighted learning) to the multiple-sites setting.

Keywords: Causal inference; Individualized treatment rule; Meta-analysis; Personalized medicine.

PubMed Disclaimer

Figures

Figure 1:
Figure 1:
Meta-analysis adapts to the level of site heterogeneity compared to base learners.
Figure 2:
Figure 2:
Estimated ECHO study discharge times from using estimated treatment rules (lower is better).
Figure 3:
Figure 3:
Relative value function under low site heterogeneity.
Figure 4:
Figure 4:
Relative value function under moderate site heterogeneity.
Figure 5:
Figure 5:
Relative value function under high site heterogeneity.
Figure 6:
Figure 6:
Treatment rule accuracy compared to true potential outcomes under low site heterogeneity.
Figure 7:
Figure 7:
Treatment rule accuracy compared to true potential outcomes under moderate site heterogeneity.
Figure 8:
Figure 8:
Treatment rule accuracy compared to true potential outcomes under high site heterogeneity.
See this image and copyright information in PMC

References

    1. Abadi Martin, Chu Andy, Goodfellow Ian, H Brendan McMahan Ilya Mironov, Talwar Kunal, and Zhang Li. Deep learning with differential privacy. In Proceedings of the 2016 ACM SIGSAC conference on computer and communications security, pages 308–318, 2016.
    1. Bagdasaryan Eugene, Veit Andreas, Hua Yiqing, Estrin Deborah, and Shmatikov Vitaly. How to backdoor federated learning. In Artificial Intelligence and Statistics, pages 2938–2948. PMLR, 2020.
    1. Bang Heejung and Robins James M. Doubly robust estimation in missing data and causal inference models. Biometrics, 61 (4):962–973, 2005. - PubMed
    1. Ben-Michael Eli, Feller Avi, Hirshberg David A, and Zubizarreta José R. The balancing act in causal inference. arXiv preprint arXiv:2110.14831, 2021.
    1. Beygelzimer Alina and Langford John. The offset tree for learning with partial labels. In Proceedings of the 15th ACM SIGKDD international conference on Knowledge discovery and data mining, pages 129–138, 2009.

LinkOut - more resources