Kwok Heemun, Coult Jason, Blackwood Jennifer, King Julia A, Kudenchuk Peter, Rea Thomas
Department of Emergency Medicine, University of Washington Seattle WA United States.
Department of Medicine, University of Washington Seattle WA United States.
Resuscitation. 2025 Apr;209:110520. doi: 10.1016/j.resuscitation.2025.110520. Epub 2025 Jan 27.
Prior studies have proposed defibrillator biosignal algorithms which characterize cardiac arrest rhythm and physiologic status. We evaluated whether a novel, individualized resuscitation strategy that integrates multiple ECG and impedance-based algorithms could reduce CPR interruptions and better align rescuer actions with patient-specific physiology.
In a retrospective cohort of ventricular fibrillation out-of-hospital cardiac arrests, observed rescuer actions (rhythm analysis, shock delivery, pulse checks, and drug therapy) were compared to hypothetical actions recommended by the proposed individualized strategy. Misdirected drug therapy was defined as either (1) epinephrine when the algorithm predicted a spontaneous pulse or (2) antiarrhythmic during predicted bradyasystole. Clinically avoidable actions included delivering a shock without restoring return of spontaneous circulation (ROSC) and interrupting CPR for pulse assessment when no spontaneous pulse was present.
Of 390 cases, 46% survived to hospital discharge. The individualized treatment strategy achieved comparable shock decision accuracy (95% sensitivity, 98% specificity) to observed care while decreasing median CPR interruption for shock from 12 to 6 s. The individualized strategy also identified 17% of 597 epinephrine and 9% of 248 antiarrhythmic administrations as misdirected. Following 1334 ventricular fibrillation shocks, the frequency of post-shock ROSC was 4% when its predicted probability was low versus 22% when not. During 1088 pulse checks, pulse was present in 5% when predicted probability of spontaneous pulse was low versus 35% when not.
An individualized resuscitation strategy could improve CPR interruption, medication administration, shock delivery, and pulse assessment. Prospective evaluation is required to assess clinical benefit.
先前的研究提出了除颤器生物信号算法，用于表征心脏骤停节律和生理状态。我们评估了一种整合多种基于心电图和阻抗的算法的新型个体化复苏策略是否可以减少心肺复苏（CPR）中断，并使救援人员的行动更好地与患者特定的生理状况相匹配。
在一组院外心室颤动心脏骤停的回顾性队列中，将观察到的救援人员行动（节律分析、电击除颤、脉搏检查和药物治疗）与所提出的个体化策略推荐的假设行动进行比较。错误的药物治疗定义为：（1）算法预测有自主脉搏时使用肾上腺素；（2）预测为缓慢性心搏停止时使用抗心律失常药物。临床上可避免的行动包括在未恢复自主循环（ROSC）时进行电击除颤，以及在无自主脉搏时中断CPR进行脉搏评估。
在390例病例中，46%存活至出院。个体化治疗策略在电击决策准确性方面（敏感性95%，特异性98%）与观察到的治疗效果相当，同时将因电击导致的CPR中断中位数从12秒降至6秒。个体化策略还识别出597次肾上腺素给药中的17%和248次抗心律失常药物给药中的9%是错误的。在1334次心室颤动电击后，当预测的电击后ROSC概率较低时，其发生率为4%，而当预测概率不低时为22%。在1088次脉搏检查中，当预测的自主脉搏概率较低时，脉搏出现的比例为5%，而当预测概率不低时为35%。
个体化复苏策略可以改善CPR中断、药物给药、电击除颤和脉搏评估。需要进行前瞻性评估以评估临床益处。
