通过聚类分析刻画特发性正常压力脑积水特征:辅助临床怀疑与诊断
本研究对178例确诊特发性正常压力脑积水(iNPH)患者纳入6项临床、功能与人口学定量变量,采用主成分分析(PCA)提取两个维度,覆盖总变异的61.8%,并通过聚类分析描述两组患者特征,以辅助临床怀疑与诊断。
要点
- PCA提取两个维度覆盖总变异61.8%,第一维度主要反映临床与功能变量,第二维度主要反映人口学特征。
- 聚类分析得出两组:第1组更年长、残疾更重、功能表现差且症状重;第2组稍年轻、教育程度更高、体能更好且临床表现更细微。
- 作者结论认为,利用定量变量和聚类分析刻画iNPH患者特征有助于指导内科与外科干预。
核心结论
- 利用定量变量和聚类分析对iNPH患者进行分型,可帮助识别其不同特征,并辅助指导内科与外科干预。
研究设计与样本
- 研究纳入178例确诊iNPH患者,考虑6项关于临床、功能和人口学方面的定量变量;采用主成分分析(PCA)定义解释现象变异的主要维度,随后描述两个iNPH患者聚类。
主要发现
- 178例患者纳入分析,PCA产生两个维度,覆盖总变异的61.8%。
- 第一维度主要基于临床变量(mRS、iNPHGs)和功能变量(TUG、Tinetti),第二维度主要代表人口学模式(年龄和教育)。
- 聚类分析描绘出两个主要患者组:第1组由更年长、更残疾、功能表现差且症状高度明显的个体组成;第2组患者稍年轻、教育程度更高、体能更好,临床方面更细微。
临床怎么用
- 作者认为,使用定量变量和聚类分析对iNPH患者进行特征刻画,可帮助识别这些患者的不同特征,并辅助指导内科与外科干预。
局限性
- 原文未报告局限性。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
PURPOSE: Idiopathic Normal Pressure Hydrocephalus (iNPH) is a neurological syndrome defined by gait disturbance, cognitive impairment and urinary incontinence. However, its clinical presentation can vary widely due to overlapping syndromes and common comorbidities in older adults. This study aims to provide practical guidance to aid in the clinical suspicion and support the diagnostic and therapeutic processes for these patients. METHODS: Six quantitative variables regarding clinical, functional, and demographic aspects were considered for a large sample of patients with diagnosed iNPH. Principal component analysis (PCA) was adopted to define the main dimensions explaining the variability of the phenomenon. Then, two clusters of iNPH patients were described. RESULTS: 178 patients were included in the analysis. The PCA produced two dimensions covering 61.8% of the total variability. The first one relied mainly on both clinical (mRS, iNPHGs) and functional (TUG, Tinetti) variables, while the second one was represented mainly on the demographic pattern (age and education). Cluster analysis depicted two main groups of patients. Cluster n.1 is composed of individuals who are older, more disabled, with poor functional performances, and highly symptomatic. Cluster n.2 patients are slightly younger, more educated, fitter, and with more nuanced clinical aspects. CONCLUSIONS: Profiling iNPH patients using quantitative variables and cluster analysis can help identify distinct characteristics of these patients, aiding in the guidance of both medical and surgical interventions.
引用
- Mazza L, Giannini G, Pirina A, et al. Profiling iNPH features through cluster analysis: an aid for clinical suspicion and diagnosis. Acta Neurochir (Wien). 2024. PMID: 39298012. doi:10.1007/s00701-024-06271-z
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。