Causal Machine Learning for Patient-Level Intraoperative Opioid Dose Prediction from Electronic Health Records

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Andersena, Jonas Valbjørn, Karlsen, Anders Peder Højer, Olsen, Markus Harboe, Pedersen, Nikolaj Krebs
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866911102978228224
author Andersena, Jonas Valbjørn
Karlsen, Anders Peder Højer
Olsen, Markus Harboe
Pedersen, Nikolaj Krebs
author_facet Andersena, Jonas Valbjørn
Karlsen, Anders Peder Højer
Olsen, Markus Harboe
Pedersen, Nikolaj Krebs
contents This paper introduces the OPIAID algorithm, a novel approach for predicting and recommending personalized opioid dosages for individual patients. The algorithm optimizes pain management while minimizing opioid related adverse events (ORADE) by employing machine learning models trained on observational electronic health records (EHR) data. It leverages a causal machine learning approach to understand the relationship between opioid dose, case specific patient and intraoperative characteristics, and pain versus ORADE outcomes. The OPIAID algorithm considers patient-specific characteristics and the influence of different opiates, enabling personalized dose recommendations. This paper outlines the algorithm's methodology and architecture, and discusses key assumptions, and approaches to evaluating its performance.
format Preprint
id arxiv_https___arxiv_org_abs_2508_09059
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Causal Machine Learning for Patient-Level Intraoperative Opioid Dose Prediction from Electronic Health Records
Andersena, Jonas Valbjørn
Karlsen, Anders Peder Højer
Olsen, Markus Harboe
Pedersen, Nikolaj Krebs
Machine Learning
This paper introduces the OPIAID algorithm, a novel approach for predicting and recommending personalized opioid dosages for individual patients. The algorithm optimizes pain management while minimizing opioid related adverse events (ORADE) by employing machine learning models trained on observational electronic health records (EHR) data. It leverages a causal machine learning approach to understand the relationship between opioid dose, case specific patient and intraoperative characteristics, and pain versus ORADE outcomes. The OPIAID algorithm considers patient-specific characteristics and the influence of different opiates, enabling personalized dose recommendations. This paper outlines the algorithm's methodology and architecture, and discusses key assumptions, and approaches to evaluating its performance.
title Causal Machine Learning for Patient-Level Intraoperative Opioid Dose Prediction from Electronic Health Records
topic Machine Learning
url https://arxiv.org/abs/2508.09059