Revisiting the High-Benefit Patient: Generic Machine Learning Inference for Intensive Blood Pressure Control
- 1University of Missouri-Columbia
Abstract
We applied the model-free GenericML framework using robust estimators (the Best Linear Predictor, Group Average Treatment Effects, and baseline profile classification) to reassess treatment effect heterogeneity (HTE) of intensive blood pressure (BP) using pooled data from two large-scale randomized control trials on BP controls. Among 10,712 participants with up to three years of follow-up, we estimated the differential risks of the primary cardiovascular outcome. Intensive BP control produced a modest average risk reduction (BLP {beta}1 = -0.0136), but evidence for HTE was not statistically significant (BLP {beta}2 p = 0.443; GATES top-bottom contrast p = 0.471). Although CLAN grouped participants with metabolically adverse profiles into the highest predicted-benefit stratum, their treatment response did not differ significantly from that of the lowest-benefit group. Overall, while machine learning can identify clusters of high-risk baseline phenotypes, group-level inference revealed no meaningful HTE, underscoring the need for caution when interpreting individualized treatment predictions.
Related
- Recurrent gene conversions and subcluster-specific variants are associated with Mycobacterium tuberculosis adaptation to intensive host pressure: a comparative genomic study
- Hospital-acquired pressure injury among COVID-19 patients admitted to the intensive care unit: Frequency and factors associated with injury stage
- Diagnostic challenges in distinguishing oral mucosal membrane pressure injuries from HSV-1-related oral mucosal lesions in intubated people in intensive care units: a scoping review protocol
- Revisiting Identification Thresholds of Specific Learning Disorders ( SLD ) in Gifted Children ( GC ). A Comparison of Three Groups: Twice‐Exceptional ( TE ) Children, GC Without SLD and Non‐ GC With SLD
- Revisiting brain asymmetry in schizophrenia: insights from machine learning subtypes from 1,761 patients
BibTeX
@misc{habibdoust2026revisiting,
title = {{Revisiting the High-Benefit Patient: Generic Machine Learning Inference for Intensive Blood Pressure Control}},
author = {A. Habibdoust and X. Song},
howpublished = {medRxiv},
year = {2026},
month = jun,
doi = {10.64898/2026.06.26.26356708},
url = {https://www.medrxiv.org/content/10.64898/2026.06.26.26356708v2},
note = {CC BY-NC},
}
RIS
TY - UNPB
TI - Revisiting the High-Benefit Patient: Generic Machine Learning Inference for Intensive Blood Pressure Control
AU - A. Habibdoust
AU - X. Song
PY - 2026
DA - 2026/06/26
PB - medRxiv
DO - 10.64898/2026.06.26.26356708
UR - https://www.medrxiv.org/content/10.64898/2026.06.26.26356708v2
AB - We applied the model-free GenericML framework using robust estimators (the Best Linear Predictor, Group Average Treatment Effects, and baseline profile classification) to reassess treatment effect heterogeneity (HTE) of intensive blood pressure (BP) using pooled data from two large-scale randomized control trials on BP controls. Among 10,712 participants with up to three years of follow-up, we estimated the differential risks of the primary cardiovascular outcome. Intensive BP control produced a modest average risk reduction (BLP {beta}1 = -0.0136), but evidence for HTE was not statistically significant (BLP {beta}2 p = 0.443; GATES top-bottom contrast p = 0.471). Although CLAN grouped participants with metabolically adverse profiles into the highest predicted-benefit stratum, their treatment response did not differ significantly from that of the lowest-benefit group. Overall, while machine learning can identify clusters of high-risk baseline phenotypes, group-level inference revealed no meaningful HTE, underscoring the need for caution when interpreting individualized treatment predictions.
ER -