Abstract:

Objective To conduct a scoping review on assessment tools for ocular complications in critically ill patients，clarify the evaluation content，application scenarios，and psychometric properties of existing tools，and analyze their clinical applicability and methodological quality，thereby providing a reference for subsequent tool development，optimization，and evidence-based clinical practice. Methods Following the scoping review framework，a systematic search was conducted across PubMed，Embase，Web of Science，CINAHL，Scopus，CNKI，Wanfang Database，VIP Database，and Chinese biomedical database. The search period was from the establishment of the databases to December 9，2025. A total of 2 researchers independently analyzed the included literature. Results A total of 10 articles were included，involving 2 longitudinal studies，3 cohort studies，2 cross-sectional studies，and 3 tool development studies. 7 ocular complication assessment tools were extracted，with core assessment dimensions including local ocular conditions，disease and treatment-related factors，and nursing operations，etc.Several tools demonstrated good reliability and validity，with Cronbach’s α coefficients ranging from 0.767 to 0.862，and both Item-level Content Validity Index and Scale-level Content Validity Index exceeding 0.800. At the same time，4 risk prediction models were identified，mostly based on Logistic or Cox regression，and common predictors were eyelid closure insufficiency，consciousness disorders，and length of hospital stay，etc. The models exhibited good discriminative ability，with Area Under the Curve（AUC） ranging from 0.860 to 0.925；the prediction model for exposure keratopathy in neurosurgical critically ill patients demonstrated the optimal performance（AUC 0.925）. Conclusion Existing assessment tools for ocular complications in critically ill patients have certain clinical applicability，but most tools face issues such as single research design，small sample sizes，and insufficient validation. Future efforts should focus on further validation and improvement of these tools.

Key words: Intensive Care Unit, Eye Complication, Evaluation Tool, Scope Review, Nursing Care

摘要：

目的 对重症患者眼部并发症评估工具进行范围综述，明确现有工具的评估内容、应用场景及测量学属性，分析其临床适用性及方法学质量，为后续工具的开发、优选及临床循证实践提供参考依据。 方法 基于范围综述框架，系统检索PubMed、Embase、Web of Science、CINAHL、Scopus、中国知网、万方数据库、维普数据库、中国生物医学文献数据库等中英文数据库，检索时限为建库至2025年12月9日。由2名研究者对纳入的文献进行综合分析。结果 共纳入10篇文献，包括纵向研究2篇、队列研究3篇、横断面研究2篇及工具开发类研究3篇。共提取7种眼部并发症评估工具，其核心评估维度包括眼部局部状况、疾病与治疗相关因素及护理操作等；部分工具信效度良好，其Cronbach’s α系数为0.767~0.862，条目水平内容效度指数及量表水平内容效度指数均>0.800。同时，识别出4个风险预测模型，多基于Logistic或Cox回归构建，常见预测因子为眼睑闭合不全、意识障碍、住院时间等；模型区分度较好，受试者操作特征曲线下面积为0.860~0.925，其中神经重症患者暴露性角膜炎预测模型性能最优，受试者操作特征曲线下面积为0.925。 结论 目前针对重症患者眼部并发症的评估工具具有一定的临床适用性，但多数工具存在研究设计单一、样本量有限、验证不充分等问题，未来应进一步对现有工具进行验证和完善。

关键词: 重症监护病房, 眼部并发症, 评估工具, 范围综述, 护理