回顾性测试可有效检测轻度认知障碍:108 项诊断研究的系统评价和荟萃分析。
Recall Tests Are Effective to Detect Mild Cognitive Impairment: A Systematic Review and Meta-analysis of 108 Diagnostic Studies.
机构信息
JC School of Public Health and Primary Care, The Chinese University of Hong Kong, Shatin, Hong Kong; Stanley Ho Big Data Decision Analytics Research Centre, The Chinese University of Hong Kong, Shatin, Hong Kong.
Stanley Ho Big Data Decision Analytics Research Centre, The Chinese University of Hong Kong, Shatin, Hong Kong.
出版信息
J Am Med Dir Assoc. 2017 Sep 1;18(9):807.e17-807.e29. doi: 10.1016/j.jamda.2017.05.016. Epub 2017 Jul 25.
BACKGROUND
Mild cognitive impairment (MCI) is a prevalent symptom associated with the increased risk of dementia. There are many cognitive tests available for detection of MCI, and investigation of the diagnostic performance of the tests is deemed necessary.
OBJECTIVE
This study aims to evaluate the diagnostic performance of different cognitive tests used for MCI detection.
DATA SOURCES
A list of cognitive tests was identified in previous reviews and from online search engines. Literature searches were performed on each of the cognitive tests in MEDLINE, Embase, and PsycINFO from the earliest available dates of individual databases to December 31, 2016. Google Scholar was used as a supplementary search tool.
STUDY SELECTION
Studies that were used to assess the diagnostic performance of the cognitive tests were extracted with inclusion and exclusion criteria. Each test's performance was compared with the standard diagnostic criteria. Bivariate random effects models were used to summarize the test performance as a point estimate for sensitivity and specificity, and presented in a summary receiver operating characteristic curve. Reporting quality and risk of bias were evaluated.
RESULTS
A total of 108 studies with 23,546 participants were selected to evaluate 9 cognitive tests for MCI detection. Most of the studies used the Mini-Mental State Examination (MMSE) (n = 58) and the Montreal Cognitive Assessment (MoCA) (n = 35). The combined diagnostic performance of the MMSE in MCI detection was 0.71 sensitivity [95% confidence interval (CI): 0.66-0.75] and 0.74 specificity (95% CI: 0.70-0.78), and of the MoCA in MCI detection was 0.83 sensitivity (95% CI: 0.80-0.86) and 0.75 specificity (95% CI: 0.69-0.80). Among the 9 cognitive tests, recall tests showed the best diagnostic performance with 0.89 sensitivity (95% CI: 0.86-0.92) and 0.84 specificity (95% CI, 0.79-0.89). In subgroup analyses, long- or short-delay recall tests have shown better performance than immediate recall tests.
CONCLUSIONS
Recall tests were shown to be the most effective test in MCI detection, especially for the population with symptoms of memory deterioration. They can be potentially used as the triage screening test for MCI in primary care setting. But when a patient shows cognitive impairments beyond memory deterioration, a more comprehensive test such as the MoCA should be used.
背景
轻度认知障碍(MCI)是一种与痴呆风险增加相关的常见症状。有许多认知测试可用于检测 MCI,因此有必要对这些测试的诊断性能进行研究。
目的
本研究旨在评估用于 MCI 检测的不同认知测试的诊断性能。
资料来源
在之前的综述和在线搜索引擎中确定了认知测试列表。对 MEDLINE、Embase 和 PsycINFO 中的每个认知测试进行了文献检索,检索时间为各数据库最早可用日期至 2016 年 12 月 31 日。Google Scholar 被用作补充搜索工具。
研究选择
使用纳入和排除标准提取用于评估认知测试诊断性能的研究。将每个测试的性能与标准诊断标准进行比较。使用双变量随机效应模型汇总测试性能,得出敏感性和特异性的点估计值,并以综合受试者工作特征曲线呈现。评估了报告质量和偏倚风险。
结果
共选择了 108 项研究,涉及 23546 名参与者,用于评估 9 种认知测试用于 MCI 检测。大多数研究使用了简易精神状态检查(MMSE)(n=58)和蒙特利尔认知评估(MoCA)(n=35)。在 MCI 检测中,MMSE 的综合诊断性能为 0.71 敏感性[95%置信区间(CI):0.66-0.75]和 0.74 特异性(95% CI:0.70-0.78),MoCA 的综合诊断性能为 0.83 敏感性(95% CI:0.80-0.86)和 0.75 特异性(95% CI:0.69-0.80)。在 9 种认知测试中,回忆测试的诊断性能最好,敏感性为 0.89(95% CI:0.86-0.92),特异性为 0.84(95% CI:0.79-0.89)。在亚组分析中,长或短延迟回忆测试的表现优于即时回忆测试。
结论
回忆测试被证明是 MCI 检测中最有效的测试,特别是对于有记忆恶化症状的人群。它们可作为初级保健环境中 MCI 的分诊筛选测试。但是,当患者的认知障碍超出记忆恶化范围时,应使用更全面的测试,如 MoCA。