Development and validation of a nomogram for predicting model for delirium after stroke.

Abstract

Purpose: To explore the risk factors of delirium in patients with stroke and develop a nomogram model to predict the occurrence of delirium.

Methods: Convenience sampling was used to select 502 patients with stroke admitted to a tertiary hospital with a neurology specialty in Tianjin from December 2023 to June 2024, who were categorized into the delirium group (n=141) and the non-delirium group (n=361) using the ICU Patient Ambiguity of Consciousness Assessment Scale. We explored the independent risk factors for the occurrence of delirium through univariate and multifactorial logistic regression analyses, established a risk prediction model, developed a nomogram, and validated the model both internally and externally.

Results: Multifactorial logistic regression analysis revealed that age (OR=1.038), abnormal vision (OR=2.740), post stroke infection (OR=3.486), National Institute of Health Stroke Scale score (NIHSS) (OR=4.175),whether restrained (OR=3.440) were independent risk factors for the development of delirium. The consistency index of the nomogram model for the occurrence of delirium in stroke patients was 0.920, with a sensitivity of 83.00% and a specificity of 90.00%.

Conclusions: This study has developed and validated a predictive nomogram for identifying delirium in patients with stroke. It can help healthcare professionals quickly identify high-risk patients for post-stroke delirium, providing a basis for further developing personalized prevention strategies and intervention measures for post-stroke delirium.