Rapid cognitive assessment tools for screening of mild cognitive impairment in the preoperative setting: A systematic review and meta-analysis

Jason Tran,Tristen Nimojan,Aparna Saripella,David F Tang-Wai,Nina Butris,Paras Kapoor,Linor Berezin,Marina Englesakis,Frances Chung

DOI: https://doi.org/10.1016/j.jclinane.2022.110682

Abstract:Importance: Mild cognitive impairment (MCI) is a high-risk precursor to dementia, post-operative delirium, and prolonged hospitalization. There is a need for preoperative rapid cognitive screening tools. Study objective: To evaluate the predictive parameters of rapid MCI screening tools in different clinical settings for preoperative application. Design: Systematic review and meta-analyses searching Medline, and other databases from inception to May 26, 2021. The Preferred Reporting Items for Systematic Review and Meta-analyses (PRISMA) guidelines were followed for data curation and quality assessment. Title and abstract screening were conducted independently via Rayyan. Data was curated through a random-effects model and statistical analysis used R-software. Setting: Community, memory clinic, emergency, long-term care, and in-patient settings. There were no studies in the preoperative setting. Patients: Twenty-three studies with 9973 patients (≥ 60 years old) undergoing rapid MCI screening. Intervention: Rapid (≤ 5 min) MCI screening tools. Measurements: Pooled predictive parameters (sensitivity, specificity) of screening tests. Main results: Eighteen screening tools, compared to neuropsychological tests, were identified. The overall prevalence of MCI among the Rapid Cognitive Screen (RCS), Six-item Screener (SIS), Mini-Cog, and Clock Drawing Test (CDT) studies were 24.6%, 28.3%, 40.9%, and 20.7%, respectively. RCS has 82% sensitivity and 79% specificity in detecting MCI. SIS has 61% sensitivity and 89% specificity. Mini-Cog has 52% sensitivity and 80% specificity. CDT has 56% sensitivity and 59% specificity. Seven other index tools had high sensitivities of 97%-82% and specificities of 90%-73% but were studied only once. Conclusion: No rapid screening tools had been validated in the surgical population. In other populations, RCS may be a promising screening tool for MCI with stronger sensitivity and specificity than Mini-Cog, SIS, and CDT. CDT alone is ineffective for MCI detection. Further validation in the preoperative setting is required to determine the efficacy of these screening tools.