Beiske Kornelia K, Sand Trond, Rugland Eyvind, Stavem Knut
*Department of Neurology and Clinical Neurophysiology, Medical Division, Akershus University Hospital, Lørenskog, Norway;†Department of Neuroscience, Norwegian University of Science and Technology, Norway;‡Department of Neurology and Clinical Neurophysiology, St. Olavs University Hospital, Norway;§Institute of Clinical Medicine, Campus Akershus University Hospital, University of Oslo, Norway;‖Department of Pulmonary Medicine, Akershus University Hospital, Lørenskog, Norway; and¶Health Services Research Unit, Akershus University Hospital, Lørenskog, Norway.
J Clin Neurophysiol. 2017 May;34(3):261-267. doi: 10.1097/WNP.0000000000000359.
Comparison of mean sleep latencies and number of sleep-onset rapid eye movement periods (SOREMPs) between modified multiple sleep latency test (MSLT) performed in the unattended home and in-hospital laboratory setting.
A randomized crossover single-blinded design. Thirty-four subjects referred to MSLT for suspected hypersomnia or narcolepsy were included. Participants were randomized to perform modified MSLT in the unattended home or in the hospital first. Scores in the two settings were compared using Wilcoxon signed-rank test or exact McNemar test. Agreement between home and hospital categorized mean sleep latency and number of SOREMPs was assessed using simple kappa (κ) and proportion agreement. Agreement between home and hospital mean sleep latency was assessed using a Bland-Altman plot and an intraclass correlation coefficient.
There was no difference between home and hospital assessment of mean sleep latency (P = 0.86). Two or more SOREMPs were found more frequently on modified MSLTs performed at home compared with those at the hospital (7 and 2, respectively; P = 0.025). Agreement was moderate for categorized sleep latency (κ = 0.53) and fair for categorized SOREMPs (κ = 0.39) in the 2 settings. Analysis of mean sleep latency using intraclass correlation coefficient showed a very good agreement between the two settings.
Group mean sleep latency for home modified MSLTs seems to be reliable compared with that for the attended sleep-laboratory setting. Higher rate of SOREMP in the unattended home suggests that napping in a familiar environment facilitates the transition into REM sleep. Further studies are needed to assess the normal limit, sensitivity, and specificity for SOREMP at home before the clinical utility of home-based napping can be determined.
比较在无人值守的家中和医院实验室环境下进行的改良多次睡眠潜伏期试验（MSLT）的平均睡眠潜伏期和睡眠起始快速眼动期（SOREMPs）数量。
采用随机交叉单盲设计。纳入34名因疑似发作性睡病或发作性睡病而接受MSLT检查的受试者。参与者被随机分配首先在无人值守的家中或医院进行改良MSLT。使用Wilcoxon符号秩检验或精确McNemar检验比较两种环境下的得分。使用简单kappa（κ）和比例一致性评估家中和医院分类的平均睡眠潜伏期和SOREMPs数量之间的一致性。使用Bland-Altman图和组内相关系数评估家中和医院平均睡眠潜伏期之间的一致性。
家中和医院对平均睡眠潜伏期的评估无差异（P = 0.86）。与在医院进行的改良MSLT相比，在家中进行的改良MSLT更频繁地发现两个或更多SOREMPs（分别为7个和2个；P = 0.025）。两种环境下分类睡眠潜伏期的一致性为中等（κ = 0.53），分类SOREMPs的一致性为一般（κ = 0.39）。使用组内相关系数对平均睡眠潜伏期进行分析表明，两种环境之间具有非常好的一致性。
与有人员值守的睡眠实验室环境相比，家中改良MSLT的组平均睡眠潜伏期似乎可靠。无人值守的家中SOREMP发生率较高表明，在熟悉的环境中小睡有助于向快速眼动睡眠过渡。在确定基于家中小睡的临床效用之前，需要进一步研究以评估家中SOREMP的正常限度、敏感性和特异性。
