Department of Psychiatry, University of KwaZulu-Natal, Durban, South Africa.
Dement Geriatr Cogn Disord. 2013;36(1-2):119-35. doi: 10.1159/000350768. Epub 2013 Jul 12.
The effectiveness of dementia screening depends on the availability of suitable screening tools with good sensitivity and specificity to confidently distinguish normal age-related cognitive decline from dementia. The aim of this study was to evaluate the discriminant validity of 7 screening measures for dementia.
A sample of 140 participants aged ≥60 years living in a residential facility for the aged were assessed clinically and assigned caseness for dementia using the Diagnostic and Statistical Manual of Mental Disorders, 4th edition, text revised diagnostic criteria. Sensitivity and specificity of a selection of the following screening measures were tested using receiver operating characteristic (ROC) analysis for individual and combined tests: the Mini-Mental State Examination (MMSE), Six-Item Screener (SIS), Subjective Memory Complaint, Subjective Memory Complaint Clinical (SMCC), Subjective Memory Rating Scale (SMRS), Deterioration Cognitive Observee (DECO) and the Clock Drawing Test (CDT).
Using ROC analyses, the SMCC, MMSE and CDT were found to be 'moderately accurate' in screening for dementia with an area under the curve (AUC) >0.70. The AUCs for the SIS (0.526), SMRS (0.661) and DECO (0.687) classified these measures as being 'less accurate'. At recommended cutoff scores, the SMCC had a sensitivity of 90.9% and specificity of 45.7%; the MMSE had a sensitivity of 63.6% and a specificity of 76.0%, and the CDT had a sensitivity of 44.4% and a specificity of 88.9%. Combining the SMCC and MMSE did not improve their predictive power except for a modest increase when using the sequential rule.
The SMCC is composed of valid screening questions that have high sensitivity, are simple to administer and ideal for administration at the community or primary health care level as a first level of 'rule-out' screening. The MMSE can be included at a second stage of screening at the general hospital level and the CDT in specialist clinical settings. Sequential use of the SMCC and MMSE will improve the specificity of the former and the sensitivity of the latter.
痴呆症筛查的有效性取决于是否有合适的筛查工具，这些工具需要具有良好的敏感性和特异性，以便能够有信心地区分正常的与年龄相关的认知衰退与痴呆症。本研究旨在评估 7 种痴呆症筛查工具的判别效度。
对居住在养老院的 140 名年龄≥60 岁的参与者进行临床评估，并使用《精神障碍诊断与统计手册》第 4 版文本修订版诊断标准进行痴呆症病例分配。使用受试者工作特征（ROC）分析对以下筛选措施中的一部分进行个体和组合测试的敏感性和特异性测试：简易精神状态检查（MMSE）、六项目筛选器（SIS）、主观记忆抱怨、主观记忆抱怨临床（SMCC）、主观记忆评定量表（SMRS）、认知观察下降者（DECO）和时钟绘图测试（CDT）。
使用 ROC 分析发现，SMCC、MMSE 和 CDT 在筛查痴呆症方面具有“中度准确性”，曲线下面积（AUC）>0.70。SIS（0.526）、SMRS（0.661）和 DECO（0.687）的 AUC 分类为“准确性较低”。在推荐的截断分数下，SMCC 的敏感性为 90.9%，特异性为 45.7%；MMSE 的敏感性为 63.6%，特异性为 76.0%，CDT 的敏感性为 44.4%，特异性为 88.9%。除了使用顺序规则时略有增加外，组合使用 SMCC 和 MMSE 并不能提高其预测能力。
SMCC 由有效的筛查问题组成，具有高敏感性、易于管理，非常适合在社区或初级保健水平作为“排除”筛查的第一道防线。MMSE 可在综合医院水平的筛查的第二阶段使用，CDT 可在专科临床环境中使用。SMCC 和 MMSE 的顺序使用将提高前者的特异性和后者的敏感性。
