Department of Neurology, Mayo Clinic, 200 1st Street SW, Rochester, MN, 55902, USA.
Department of Neurology, Concord Repatriation General Hospital, Concord, NSW, Australia.
Neurocrit Care. 2023 Jun;38(3):600-611. doi: 10.1007/s12028-022-01601-4. Epub 2022 Sep 19.
Although median nerve somatosensory evoked potentials are routinely used for prognostication in comatose cardiac arrest survivors, myogenic artifact can reduce inter-rater reliability, leading to unreliable or inaccurate results. To minimize this risk, we determined the benefit of neuromuscular blockade agents in improving the inter-rater reliability and signal-to-noise ratio of SSEPs in the context of prognostication.
Thirty comatose survivors of cardiac arrest were enrolled in the study, following the request from an intensivist to complete an SSEP for prognostication. Right and left median nerve SSEPs were obtained from each patient, before and after administration of an NMB agent. Clinical histories and outcomes were retrospectively reviewed. The SSEP recordings before and after NMB were randomized and reviewed by five blinded raters, who assessed the latency and amplitude of cortical and noncortical potentials (vs. absence of response) as well as the diagnostic quality of cortical recordings. The inter-rater reliability of SSEP interpretation before and after NMB was compared via Fleiss' κ score.
Following NMB administration, Fleiss' κ score for cortical SSEP interpretation significantly improved from 0.37 to 0.60, corresponding to greater agreement among raters. The raters were also less likely to report the cortical recordings as nondiagnostic following NMB (40.7% nondiagnostic SSEPs pre-NMB; 17% post-NMB). The SNR significantly improved following NMB, especially when the pre-NMB SNR was low (< 10 dB). Across the raters, there were three patients whose SSEP interpretation changed from bilaterally absent to bilaterally present after NMB was administered (potential false positives without NMB).
NMB significantly improves the inter-rater reliability and SNR of median SSEPs for prognostication among comatose cardiac arrest survivors. To ensure the most reliable prognostic information in comatose post-cardiac arrest survivors, pharmacologic paralysis should be consistently used before recording SSEPs.
虽然正中神经体感诱发电位常用于昏迷性心脏骤停幸存者的预后评估，但肌源性伪迹会降低评分者间的可靠性，导致结果不可靠或不准确。为了最大程度地降低这种风险，我们确定了神经肌肉阻滞剂在改善预后正中神经体感诱发电位评分者间可靠性和信噪比方面的益处。
根据一名重症监护医生的要求，对 30 名昏迷性心脏骤停幸存者进行了正中神经体感诱发电位检查以进行预后评估。在使用神经肌肉阻滞剂前后，从每位患者的右侧和左侧正中神经获取体感诱发电位。回顾性审查临床病史和结局。在使用神经肌肉阻滞剂前后，将体感诱发电位记录随机化并由五名盲法评分者进行评估，他们评估皮质和非皮质电位的潜伏期和振幅（与无反应相比）以及皮质记录的诊断质量。使用 Fleiss'κ评分比较使用和不使用神经肌肉阻滞剂前后体感诱发电位解释的评分者间可靠性。
使用神经肌肉阻滞剂后，皮质体感诱发电位解释的 Fleiss'κ评分从 0.37 显著提高到 0.60，表明评分者之间的一致性更高。评分者报告皮质记录不可诊断的可能性也更低（使用神经肌肉阻滞剂前为 40.7%不可诊断的体感诱发电位，使用后为 17%）。使用神经肌肉阻滞剂后信噪比显著提高，尤其是在使用神经肌肉阻滞剂前信噪比较低（<10dB）时。在所有评分者中，有 3 名患者在使用神经肌肉阻滞剂后，体感诱发电位从双侧缺失变为双侧存在（无神经肌肉阻滞剂时的潜在假阳性）。
神经肌肉阻滞剂可显著提高昏迷性心脏骤停幸存者正中神经体感诱发电位的评分者间可靠性和信噪比。为了在昏迷性心脏骤停后幸存者中获得最可靠的预后信息，应在记录体感诱发电位前一致使用药物性瘫痪。
