Related Experiment Videos
Guideline-Aligned Machine Learning for Predicting Ondansetron Administration at the End of Anaesthesia: Explainable
Tom Strube1, Leoni Weltermann1, Jonas Weber2
1Department Healthcare, Fraunhofer Institute for Software and Systems Engineering ISST, Dortmund, Germany.
Studies in Health Technology and Informatics
|May 23, 2026
Summary
Artificial Intelligence (AI) and clinical practice guidelines can conflict. A new Guideline-Aligned Machine Learning (GAML) model predicts ondansetron use for postoperative nausea and vomiting (PONV) with high accuracy.
Area of Science:
- Anesthesiology
- Medical Informatics
- Artificial Intelligence
Background:
- Clinical Practice Guidelines (CPGs) and Artificial Intelligence (AI) aim to improve clinical decisions but can offer conflicting advice.
- Postoperative Nausea and Vomiting (PONV) management presents a challenge where guideline adherence and decision support are crucial.
Purpose of the Study:
- To develop and evaluate a Guideline-Aligned Machine Learning (GAML) model for predicting ondansetron administration.
- To align AI-driven predictions with established clinical practice guidelines for PONV management.
Main Methods:
- Analysis of 16,240 anesthesia protocols to identify risk factors and PONV prophylaxis.
- Training and cross-validation of logistic regression, multinomial naïve Bayes, and CatBoost classifiers on 80% of the data.
- Evaluation of model performance on the remaining 20% using accuracy, precision, and recall metrics; SHAP plots for interpretability.
Main Results:
- The GAML model achieved high accuracy (90 ± 1%) in predicting ondansetron administration.
- Moderate precision (60 ± 5%) and recall (75 ± 4%) were observed across the evaluated machine learning models.
- SHAP decision plots provided visualization of predictor contributions, suggesting potential for interactive planning tools.
Conclusions:
- GAML demonstrates promise as an explainable AI tool for clinical decision support in anesthesia and PONV prophylaxis.
- Integrating AI with clinical guidelines can help resolve conflicting recommendations and enhance patient care.
- The developed model offers a foundation for explainable AI applications in perioperative medicine.
Related Concept Videos
01:15Impact of Pharmacokinetic–Pharmacodynamic Models: Regulatory Decisions
PK–PD modeling has significantly influenced FDA regulatory decisions, particularly drug approval, dosage optimization, and labeling. These models integrate pharmacokinetics (PK) and pharmacodynamics (PD) to predict drug behavior and effects, aiding in optimizing dosing regimens and enhancing the probability of clinical trial success.One notable example is Nesiritide (Natrecor®), a recombinant human brain natriuretic peptide for treating acute decompensated congestive heart failure (CHF).
01:24General Anesthesia: Overview
Anesthesia is a medical procedure that uses drugs for CNS suppression to enable painless surgeries and procedures. The selection of anesthetics is influenced by their pharmacokinetic properties, side effects, and patient characteristics. Various types of anesthesia include general, local, regional, spinal, and inhalational.
General anesthesia induces unconsciousness in the whole body, while the others target specific areas or sensations. It is administered to minimize adverse effects, maintain...
General anesthesia induces unconsciousness in the whole body, while the others target specific areas or sensations. It is administered to minimize adverse effects, maintain...
01:27Chemotherapy-Induced Nausea and Vomiting: 5-HT3 Receptor Antagonists
5-HT3 receptor antagonists, such as dolasetron, granisetron (Kytril), ondansetron (Zofran), and palonosetron (Axoli), are crucial in managing chemotherapy-induced nausea and vomiting (CINV) and postoperative nausea. These drugs selectively block 5-HT3 receptors in the visceral vagal and spinal afferent nerves, chemoreceptor trigger zone, and the vomiting center. They have a rapid onset of action and can be given as a single dose before chemotherapy. Ondansetron and granisetron, in particular,...
01:19Pharmacovigilance
Post-marketing surveillance is a critical component of pharmaceutical regulation, often uncovering unanticipated adverse drug reactions (ADRs) once a drug is widely used over an extended period.
This process, termed pharmacovigilance, aims to detect, evaluate, and minimize harmful effects related to medication use. The data collection for pharmacovigilance depends on spontaneous reporting systems, where healthcare professionals or patients voluntarily report suspected ADRs.
In some cases, there...
This process, termed pharmacovigilance, aims to detect, evaluate, and minimize harmful effects related to medication use. The data collection for pharmacovigilance depends on spontaneous reporting systems, where healthcare professionals or patients voluntarily report suspected ADRs.
In some cases, there...
01:26Pharmaceutical Poisoning: Treatment Strategies
Treatment strategies for poisoning are a critical aspect of emergency medicine, focusing on preventing the absorption of toxins and enhancing their elimination. When a poisoning incident occurs, the first response is to halt exposure and decontaminate the patient, particularly through gastrointestinal (GI) methods if the poison was ingested.Gastrointestinal Decontamination Techniques:Activated charcoal is the cornerstone of GI decontamination. It works through adsorption, binding the toxin to...
01:19One-Compartment Open Model for IV Bolus Administration: General Considerations
The one-compartment model is a pharmacokinetic tool that models the body as a single, uniform compartment, facilitating the understanding of drug distribution and elimination. This model is particularly beneficial for intravenous (IV) bolus administration, where the drug rapidly circulates throughout the body.
The drug's presence in the body is defined by an equation representing the difference between the rates of drug entry and exit. Key parameters—elimination rate constant, half-life,...
The drug's presence in the body is defined by an equation representing the difference between the rates of drug entry and exit. Key parameters—elimination rate constant, half-life,...