Machine Learning–Based Prediction of Subsyndromal Delirium in Intensive Care Patients

Activity Summary

Required reading for all learners: Implicit Bias impacts patient outcomes

Selecting an appropriate model for early detection of ICU patients at high risk of subsyndromal delirium can help medical staff develop tailored intervention plans, improving patient outcomes and reducing healthcare costs. This article presents the development of four machine learning models to predict the risk of subsyndromal delirium in the ICU. Eight predictors of subsyndromal delirium were identified by least absolute shrinkage and selection operator and multivariate logistic regression analysis.

Objectives

Identify the 8 independent predictors of subsyndromal delirium in ICU patients, including RCSQ score and CRRT/ECMO receipt.
Describe the performance of the random forest model (AUC=0.885, F1=0.629) as the optimal model for subsyndromal delirium prediction.
Analyze modifiable clinical interventions (e.g., sleep optimization, reduced restraints) to lower subsyndromal delirium risk in ICU patients.

Continuing Education Disclosure Statement

Criteria for Awarding Contact Hours

Learners must complete the entire activity and the associated evaluation AND read Implicit Bias impacts patient outcomes. No partial credit will be awarded.

Accreditation

American Association of Critical-Care Nurses is accredited as a provider of nursing continuing professional development by the American Nurses Credentialing Center's Commission on Accreditation.
Provider approved by the California Board of Registered Nursing, Provider number CEP 1036, for {contactHours} contact hours.

Disclosures

The Nurse Planner has determined that no individuals with the ability to control content of this activity have relevant relationships with ineligible companies.

Activities with pharmacotherapeutic credit are to assist the APRN in fulfilling their education requirements for licensure and certification renewals.

Refund Policy

Continuing Education Activities are nonrefundable.