Data_Sheet_1_Predictive Factors for the Need of Tracheostomy in Patients With Large Vessel Occlusion Stroke Being Treated With Mechanical Thrombectomy.docx

Background: Patients with large vessel occlusion stroke (LVOS) eligible for mechanical thrombectomy (MT) are at risk for stroke- and non-stroke-related complications resulting in the need for tracheostomy (TS). Risk factors for TS have not yet been systematically investigated in this subgroup of stroke patients.Methods: Prospectively derived data from patients with LVOS and MT being treated in a large, academic neurological ICU (neuro-ICU) between 2014 and 2019 were analyzed in this single-center study. Predictive value of peri- and post-interventional factors, stroke imaging, and pre-stroke medical history were investigated for their potential to predict tracheostomy during ICU stay using logistic regression models.Results: From 635 LVOS-patients treated with MT, 40 (6.3%) underwent tracheostomy during their neuro-ICU stay. Patients receiving tracheostomy were younger [71 (62–75) vs. 77 (66–83), p < 0.001], had a higher National Institute of Health Stroke Scale (NIHSS) at baseline [18 (15–20) vs. 15 (10–19), p = 0.009] as well as higher rates of hospital acquired pneumonia (HAP) [39 (97.5%) vs. 224 (37.6%), p < 0.001], failed extubation [15 (37.5%) vs. 19 (3.2%), p < 0.001], sepsis [11 (27.5%) vs. 16 (2.7%), p < 0.001], symptomatic intracerebral hemorrhage [5 (12.5%) vs. 22 (3.9%), p = 0.026] and decompressive hemicraniectomy (DH) [19 (51.4%) vs. 21 (3.8%), p < 0.001]. In multivariate logistic regression analysis, HAP (OR 21.26 (CI 2.76–163.56), p = 0.003], Sepsis [OR 5.39 (1.71–16.91), p = 0.004], failed extubation [OR 8.41 (3.09–22.93), p < 0.001] and DH [OR 9.94 (3.92–25.21), p < 0.001] remained as strongest predictors for TS. Patients with longer periods from admission to TS had longer ICU length of stay (r = 0.384, p = 0.03). There was no association between the time from admission to TS and clinical outcome (NIHSS at discharge: r = 0.125, p = 0.461; mRS at 90 days: r = −0.179, p = 0.403).Conclusions: Patients with LVOS undergoing MT are at high risk to require TS if extubation after the intervention fails, DH is needed, and severe infectious complications occur in the acute phase after ischemic stroke. These factors are likely to be useful for the indication and timing of TS to reduce overall sedation and shorten ICU length of stay.

背景：符合机械血栓切除术（MT）条件的较大血管闭塞型卒中（LVOS）患者，存在因卒中及相关并发症而需进行气管造口（TS）的风险。尚未对这一卒中患者亚组的TS风险因素进行系统性的研究。方法：本单中心研究分析了2014年至2019年间在大型、学术性神经重症监护病房（neuro-ICU）接受治疗的LVOS和MT患者的前瞻性数据。本研究通过逻辑回归模型，调查了围手术期及术后因素、卒中影像学检查和术前病史的预测价值，以预测ICU住院期间是否需要气管造口。结果：在635名接受MT治疗的LVOS患者中，40名（6.3%）在神经ICU期间接受了气管造口。接受气管造口的患者年龄较轻[71岁（62–75岁）比77岁（66–83岁），p < 0.001]，基线时的美国国立卫生研究院卒中量表（NIHSS）评分较高[18分（15–20分）比15分（10–19分），p = 0.009]，以及医院获得性肺炎（HAP）发生率较高[39例（97.5%）比224例（37.6%），p < 0.001]，拔管失败[15例（37.5%）比19例（3.2%），p < 0.001]，败血症[11例（27.5%）比16例（2.7%），p < 0.001]，症状性颅内出血[5例（12.5%）比22例（3.9%），p = 0.026]和减压性颅骨切除术（DH）[19例（51.4%）比21例（3.8%），p < 0.001]。在多因素逻辑回归分析中，HAP（OR 21.26（CI 2.76–163.56），p = 0.003）、败血症[OR 5.39（1.71–16.91），p = 0.004]、拔管失败[OR 8.41（3.09–22.93），p < 0.001]和DH[OR 9.94（3.92–25.21），p < 0.001]仍然是TS最强的预测因子。从入院到TS的时间较长的患者，其在ICU的住院时间也较长（r = 0.384，p = 0.03）。从入院到TS的时间与临床预后（出院时的NIHSS评分：r = 0.125，p = 0.461；90天时的mRS评分：r = −0.179，p = 0.403）之间没有关联。结论：接受MT治疗的LVOS患者，若干预后拔管失败、需要DH，或在缺血性卒中急性期发生严重感染并发症，则高度需要TS。这些因素可能有助于TS的指征和时机选择，以减少总体镇静时间和缩短ICU住院时间。