摘要目的:探讨谵妄预测模型框架下风险等级预防干预在重症监护病房（ICU）机械通气患者中的应用价值。方法:选取武警海警总队医院2020年1月至2022年12月在ICU行机械通气的急性呼吸窘迫综合征患者124例进行随机对照研究，根据随机数字表法分为观察组、对照组各62例，对照组接受常规干预，观察组行谵妄预测模型框架下风险等级预防干预，两组干预时间4周。观察两组干预效果。结果:观察组机械通气时间［（149.97±29.86）h］、谵妄持续时间［（2.79±1.03）d］、谵妄发生率［3.23%（2/62）］、ICU住院时间［（7.57±2.14）d］及病死率［1.61%（1/62）］均显著短于、低于对照组的（170.78±34.28）h、（4.12±0.98）d、33.87%（21/62）、（10.21±2.85）d、19.35%（12/62）（ t=3.60、7.37，χ 2=19.29， t=5.83，χ 2=10.40，均 P < 0.001）；观察组入睡难度评分、睡眠深度评分、再入睡难度评分、夜间觉醒评分、整体睡眠质量评分均显著高于对照组（ t=-8.48、-4.57、-4.50、-5.26、-5.86，均 P < 0.001）；干预后，观察组治疗信心评分［（28.75±4.87）分］、家庭关怀度评分［（7.62±1.13）分］及干预满意度［91.94%（57/62）］均显著高于对照组［（28.75±4.87）分、（7.62±1.13）分、77.42%（48/62）］（ t=-7.79、-12.74，χ 2=5.03，均 P < 0.05）。 结论:谵妄预测模型框架下风险等级预防干预有助于降低患者的机械通气时间、谵妄持续时间及ICU住院时间，控制谵妄和病死率，改善患者的睡眠质量，提高治疗信心、家庭关怀度和干预满意度。

abstractsObjective:To investigate the application value of risk level prevention intervention based on delirium prediction models in ICU patients undergoing mechanical ventilation.Methods:A randomized controlled study was conducted on 124 patients with acute respiratory distress syndrome who underwent mechanical ventilation in the ICU at the Marine Police Corps Hospital of Chinese People's Armed Police Force from January 2020 to December 2022. They were divided into an observation group and a control group, with 62 patients in each group, using a random number table method. The control group received routine intervention, while the observation group received risk level prevention intervention based on delirium prediction models. The intervention lasted for 4 weeks. The intervention effects were observed in both groups.Results:In the observation group, the duration of mechanical ventilation was (149.97 ± 29.86) hours, the duration of delirium was (2.79 ± 1.03) days, the incidence of delirium was 3.23% (2/62), the length of ICU stay was (7.57 ± 2.14) days, and the mortality rate was 1.61% (1/62). All of these values were significantly shorter or lower than those in the control group [(170.78 ± 34.28) hours, (4.12 ± 0.98) days, 33.87% (21/62), (10.21 ± 2.85) days, 19.35% (12/62), t = 3.60, 7.37, χ2 = 19.29, t = 5.83, χ2 = 10.40, all P < 0.001]. The observation group scored significantly higher than the control group in terms of difficulty in falling asleep, sleep depth, difficulty in returning to sleep, nighttime awakenings, and overall sleep quality ( t = -8.48, -4.57, -4.50, -5.26, -5.86, all P < 0.001). After the intervention, the observation group also had significantly higher scores for treatment confidence [(28.75 ± 4.87) points], family care [(7.62 ± 1.13) points], and intervention satisfaction [91.94% (57/62) points] compared with the control group [(28.75 ± 4.87) points, (7.62 ± 1.13) points, 77.42% (48/62), t = -7.79, -12.74, χ2 = 5.03, all P < 0.05]. Conclusion:The risk level prevention intervention based on delirium prediction models helps to reduce the duration of mechanical ventilation, the duration of delirium, and the length of ICU stay. It also controls delirium and mortality rates, improves patients' sleep quality, and enhances treatment confidence, family support, and overall satisfaction with the intervention.