Department of Allied Health Sciences, University of Connecticut, Storrs, CT, USA; Institute for Collaboration on Health, Intervention, and Policy, University of Connecticut, Storrs, CT, USA.
Department of Psychological Sciences and Brain Health Research Institute, Kent State University, Kent, OH, USA.
Drug Alcohol Depend. 2021 Jul 1;224:108726. doi: 10.1016/j.drugalcdep.2021.108726. Epub 2021 Apr 24.
Cognitive dysfunction is common in persons seeking medication for opioid use disorder (MOUD) and may hinder many addiction-related services. Brief but accurate screening measures are needed to efficiently assess cognitive dysfunction in these resource-limited settings. The study aimed to develop a brief predictive risk score tailored for use among patients in drug treatment.
The present study examined predictors of mild cognitive impairment (MCI), objectively assessed via the NIH Toolbox, among 173 patients receiving methadone as MOUD at an urban New England drug treatment facility. Predictors of MCI were identified in one subsample using demographic characteristics, medical chart data, and selected items from the Brief Inventory of Neuro-Cognitive Impairment (BINI). Predictors were cross-validated in a second subsample using logistic regression. Receiver operating curve (ROC) analyses determined an optimal cut-off score for detecting MCI.
A cognitive dysfunction risk score (CDRS) was calculated from patient demographics (age 50+, non-White ethnicity, less than high school education), medical and substance use chart data (history of head injury, overdose, psychiatric diagnosis, past year polysubstance use), and selected self-report items (BINI). The CDRS discriminated acceptably well, with a ROC curve area of 70.6 %, and correctly identified 78 % of MCI cases (sensitivity = 87.5 %; specificity = 55.6 %).
The CDRS identified patients with cognitive challenges at a level likely to impede treatment engagement and/or key outcomes. The CDRS may assist in efficiently identifying patients with cognitive dysfunction while requiring minimal training and resources. Larger validation studies are needed in other clinical settings.
认知功能障碍在寻求阿片类药物使用障碍（MOUD）药物治疗的人群中很常见，并且可能会阻碍许多与成瘾相关的服务。在资源有限的环境中，需要简短但准确的筛查措施来有效地评估认知功能障碍。本研究旨在为药物治疗患者开发一种针对使用的简短预测风险评分。
本研究检查了通过 NIH 工具包客观评估的在新英格兰市药物治疗机构接受美沙酮作为 MOUD 的 173 名患者中轻度认知障碍（MCI）的预测因素。使用人口统计学特征、病历数据和Brief Inventory of Neuro-Cognitive Impairment（BINI）中的选定项目，在一个子样本中确定 MCI 的预测因素。使用逻辑回归在第二个子样本中交叉验证预测因素。接收者操作特征（ROC）分析确定用于检测 MCI 的最佳截断分数。
从患者人口统计学特征（年龄 50 岁以上、非白种人、未接受过高中教育）、医疗和药物使用图表数据（头部受伤史、过量、精神科诊断、过去一年多种药物使用）以及选定的自我报告项目（BINI）计算出认知功能障碍风险评分（CDRS）。CDRS 具有可接受的区分能力，ROC 曲线下面积为 70.6%，正确识别了 78%的 MCI 病例（敏感性=87.5%；特异性=55.6%）。
CDRS 确定了认知能力挑战的患者，其水平可能会阻碍治疗的参与和/或关键结果。CDRS 可以在需要最小培训和资源的情况下，帮助有效地识别认知功能障碍患者。需要在其他临床环境中进行更大规模的验证研究。
