Abstract
BACKGROUND: Mild cognitive impairment (MCI) is a major risk factor for delirium. Few studies with mid- to long-term follow-up periods have examined delirium, MCI, and the conversion of MCI to dementia. This prospective, longitudinal cohort study aims to assess the incidence of dementia after delirium onset in patients with MCI and to develop an artificial intelligence algorithm for predicting the conversion of MCI to dementia. METHODS: A 30-day evaluation for delirium will be performed for adults aged ≥ 65 years diagnosed with MCI who score 18-23 on the Korean Mini-Mental State Exam. Individuals exhibiting delirium at least once will be classified into the delirium group and the remainder into the non-delirium group. Over 3 years, patients will undergo follow-ups for survival and conversion to dementia at 30-day intervals to analyze risk factors for delirium and clinical outcomes. Differences between the two groups will be analyzed using the chi-square and independent t-tests. Kaplan-Meier survival and Cox proportional hazard regression will be used to assess the effects on 30-day, 60-day, 12-month, 24-month, and 36-month mortality, whereas linear regression will be used to assess the length of stay at long-term care facilities and medical costs. DISCUSSION: The results will emphasize the importance of preventing delirium in MCI patients, and preventive interventions can be strengthened. TRIAL REGISTRATION: NCT05113446 (ClinicalTrials.gov). Registered 9 November 2021 - Retrospectively registered.