Delirium History and Preoperative Mild Neurocognitive Disorder: An Opportunity for Multidisciplinary Patient-Centered Care.

BACKGROUND Delirium is a well-established clinical phenomenon that remains largely underdiagnosed. In light of its association with diminished postoperative outcomes, recent efforts involve implementing preventive strategies and fostering early detection. This report highlights how multidisciplinary interventions can inform risk for delirium and the challenges that accompany identifying at-risk patients. CASE REPORT A 75-year-old male with a history of postoperative cognitive complications including delirium and mild cognitive impairment. He was attending an outpatient preoperative anesthesia clearance assessment prior to a planned removal for a left frontoethmoidal sinus mucocele. As part of clinical care, an in-house neuropsychologist completed a neurobehavioral exam to assess current cognitive status and guide perioperative cognitive care recommendations. Findings were consistent with mild neurocognitive disorder. CONCLUSIONS Given the patient's history and current status, he was listed as a high delirium risk. The team provided information on delirium and delirium risk factors, encouraged the patient to speak to his surgeon and also a geriatric specialist to assist with decision making. Due to their concern about delirium, the patient and his caregiver opted to postpone the left frontoethmoidal sinus mucocele removal.

Delirium Prediction using Machine Learning Models on Preoperative Electronic Health Records Data.

Electronic Health Records (EHR) are mainly designed to record relevant patient information during their stay in the hospital for administrative purposes. They additionally provide an efficient and inexpensive source of data for medical research, such as patient outcome prediction. In this study, we used preoperative Electronic Health Records to predict postoperative delirium. We compared the performance of seven machine learning models on delirium prediction: linear models, generalized additive models, random forests, support vector machine, neural networks, and extreme gradient boosting. Among the models evaluated in this study, random forests and generalized additive model outperformed the other models in terms of the overall performance metrics for prediction of delirium, particularly with respect to sensitivity. We found that age, alcohol or drug abuse, socioeconomic status, underlying medical issue, severity of medical problem, and attending surgeon can affect the risk of delirium.