Abstract
The number of elderly patients with cervical trauma is increasing. Such patients are considered to be at high risk for delirium, which is an acute neuropsychological disorder that reduces the patient's capacity to interact with their environment due to impairments in cognition. This study aimed to establish a risk score that predicts delirium in elderly patients with cervical SCI and/or cervical fracture regardless of treatment type. This retrospective cohort study included 1512 patients aged ≥65 years with cervical SCI and/or cervical fracture. The risk factors for delirium according to treatment type (surgical or conservative) were calculated using multivariate logistic regression. A delirium risk score was established as the simple arithmetic sum of points assigned to variables that were significant in the multivariate analyses. Based on the statistical results, the delirium risk score was defined using six factors: old age (≥80 years), hypoalbuminemia, cervical fracture, major organ injury, dependence on pre-injury mobility, and comorbid diabetes. The score's area under the curve for the prediction of delirium was 0.66 (p < 0.001). Although the current scoring system must be validated with an independent dataset, the system remains beneficial because it can be used after screening examinations upon hospitalization and before deciding the treatment strategy.