Truace
Subscribe

Machine learning for estimating and comparing clinical rules for treating diarrheal illness with antibiotics

Acute diarrheal disease is a leading cause of death in children under age 5, disproportionately impacting children in low-resource settings. Although many cases could respond to antibiotic treatment, the benefits of widely prescribing antibiotics must be weighed against the risks of antimicrobial resistance. These challenges motivate development of individualized treatment guidelines for diarrheal disease. In this study, we utilize a framework for creation and evaluation of individualized treatment rules that le…

By the Truvace desk2 min read1 sourcePeer-reviewedTRV-2026-1680· cite & verify
Machine learning for estimating and comparing clinical rules for treating diarrheal illness with antibiotics
Navy and Marine Corps Medical News April 2010 by U.S. Navy. Bureau of Medicine and Surgery. Public domain

In brief

Researchers developed a framework that creates individualized rules for deciding which children with diarrheal illness should receive antibiotics. The rules draw on diagnostic and clinical information and are built with nested cross-validation, ensemble machine learning, and doubly-robust estimation.

The work matters because diarrheal disease is a leading killer of young children, yet widespread antibiotic use risks driving antimicrobial resistance. A rule that targets treatment could balance those pressures. The abstract reports a simulation and an application to trial data, but gives no results on health outcomes.

Main points

  1. Acute diarrheal disease is a leading cause of death in children under 5, especially in low-resource settings.
  2. The rules deliberately cap the proportion of children treated to limit overtreatment and antimicrobial resistance.
  3. The method compares rules built on different sets of input covariates to estimate the value of measuring extra biomarkers.
  4. It was tested in a realistic simulation study and applied to data from the ABCD trial.

The gain

Individualized treatment rules built with ensemble machine learning can use clinical and diagnostic data to recommend antibiotics while capping the share of children treated, and can quantify what extra biomarkers add.

The rundown

The study starts from a tension in pediatric care: many diarrhea cases could respond to antibiotics, but prescribing them widely carries the risk of antimicrobial resistance. This motivates guidelines that tailor treatment to the individual child.

Unlike many treatment-rule pipelines, this approach builds in a limit on how many children are treated, based on expected clinical benefit. It also offers a way to compare rules that use different input variables, which could show whether additional biomarker testing is worthwhile.

For estimation, the authors use nested cross-validation with ensemble machine learning and doubly-robust estimation. They report that the method gives appropriate inference in simulation and apply it to the AntiBiotics for Children with severe Diarrhea (ABCD) trial.

Sources

  1. Peer-reviewedThe International Journal of Biostatistics2026-10-09

The debate