What question did this study set out to answer?

Systematically evaluate and compare the statistical properties and performance of 13 subgroup identification methods used to guide precision medicine.

June 9, 2019Wiley Interdisciplinary Reviews Data Mining and Knowledge Discovery

Subgroup identification for precision medicine: A comparative review of 13 methods

View Full Paper

Ask AI

Bookmark

Authors

WLWei‐Yin LohUniversity of Wisconsin–Madison LCLuxi CaoZhejiang Provincial People's Hospital PZPeigen ZhouUniversity of Wisconsin–Madison

Discussion

Member takes

Implication

Comparative review reveals significant trade-offs across 13 subgroup identification methods in clinical datasets, highlighting widespread statistical limitations in precision medicine pipelines.

Key Points

Systematically evaluate and compare the statistical properties and performance of 13 subgroup identification methods used to guide precision medicine.
Evaluated publicly available software for 13 subgroup identification methods across simulated and real-world patient datasets.
Benchmarked methods using seven standardized statistical criteria: variable selection bias, false discovery rate, predictive variable recovery, treatment effect estimation bias, subgroup size, true treatment effect, and subgroup stability.
Many of the 13 evaluated methods perform poorly on at least one critical statistical metric.
Methods display substantial variability and trade-offs in selection bias, false discovery rates, and the replicability of identified patient subgroups.

Cite This Study

Loh et al. (2019) studied this question.

synapsesocial.com/papers/69d8fa3a2c87b79b92d189cc https://doi.org/10.1002/widm.1326

View Full Paper

Ask AI

Bookmark

Implication

Comparative review reveals significant trade-offs across 13 subgroup identification methods in clinical datasets, highlighting widespread statistical limitations in precision medicine pipelines.

Subgroup identification for precision medicine: A comparative review of 13 methods

Discussion

Member takes

Implication

Key Points

Cite This Study

Subgroup identification for precision medicine: A comparative review of 13 methods

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider