Division of Gastroenterology and Hepatology, University of Michigan, Ann Arbor, MI, USA.
Institute for Healthcare Policy and Innovation, Ann Arbor, MI, USA.
Am J Gastroenterol. 2018 Apr;113(4):529-538. doi: 10.1038/ajg.2018.6. Epub 2018 Mar 13.
Minimal hepatic encephalopathy (HE) is common, characterized by deficits in reaction time and executive function, and strongly associated with disability and mortality. Point-of-care diagnostics performed without specialized skills or equipment are now available, albeit with limited data regarding their generalizability.
We systematically reviewed MEDLINE, EMBASE, Cochrane Library, and Scopus for diagnostic studies of MHE using broad search terms including HE and minimal, covert, or the names of published diagnostic modalities. We included tests that provide results during clinical visits without requiring neuropsychologists to administer and/or special equipment. These include the Inhibitory Control Test (ICT, n=16), EncephalApp Stroop (n=3), an algorithm based on the Sickness Impact Profile (n=2), and the Animal Naming Test (ANT, n=1).
The populations enrolled in the included study were highly selected, excluding patients with recent (6-months) alcohol or psychoactive medications use. Cutoffs for MHE for each test varied widely. For the ICT, the optimal cutoffs for MHE varied by 300%, whereas healthy control performance varied >400%. The optimal cutoffs for the EncephalApp also varied (by 50%). The gold standards for MHE varied substantially between studies, and clinical outcomes were never used to develop test cutoffs. Data comparing the performance of each modality are lacking. Longitudinal data are limited but suggest that good performance on the ICT, EncephalApp or ANT is associated with reduced risk of developing overt HE.
The point-of-care tests for MHE are promising tools. However, additional longitudinal studies are needed in clinically representative populations of at-risk patients with cutoffs validated based on the development of clinical outcomes.
轻微型肝性脑病（HE）较为常见，其特征为反应时间和执行功能受损，与残疾和死亡率密切相关。目前已有无需专门技能或设备即可进行的即时诊断检测，但关于其通用性的数据有限。
我们系统性地检索了 MEDLINE、EMBASE、Cochrane 图书馆和 Scopus，以获取使用广泛搜索词（包括 HE 和轻微、隐匿或已发表诊断方式的名称）进行的 MHE 诊断研究。我们纳入了在临床就诊时即可提供结果且无需神经心理学家进行管理和/或特殊设备的检测。这些检测包括抑制控制测试（ICT，n=16）、EncephalApp Stroop（n=3）、基于疾病影响概况的算法（n=2）和动物命名测试（ANT，n=1）。
纳入研究的患者人群高度选择，排除了近期（6 个月）使用酒精或精神活性药物的患者。每项检测的 MHE 截断值差异很大。对于 ICT，MHE 的最佳截断值变化幅度达 300%，而健康对照组的差异超过 400%。EncephalApp 的最佳截断值也有所不同（相差 50%）。MHE 的金标准在研究之间存在显著差异，且从未使用临床结局来制定检测截断值。缺乏比较每种方式性能的数据。纵向数据有限，但表明 ICT、EncephalApp 或 ANT 表现良好与发生显性 HE 的风险降低相关。
即时检测 MHE 的检测工具很有前景。然而，需要在有风险的患者中进行具有基于临床结局开发的截断值的代表性临床人群中进行更多的纵向研究。
