急性主动脉夹层发生低氧血症患者 危险因素的Meta 分析

A meta-analysis of risk factors for hypoxemia in patients with acute aortic dissection

摘要:

目的 通过Meta 分析明确主动脉夹层患者发生低氧血症的危险因素。方法 检索中英文数据库及 纳入文献的参考文献，由2 位评价人员按照标准独立筛选文献，提取资料并进行文献质量评价，应用RevMan 5.0 分析软件对资料进行Meta 分析。结果 从27 篇文献筛选7 篇符合纳入标准，样本量为1 175 例，主动脉夹 层发生低氧血症患者为376 例，主动脉夹层发生非低氧血症患者为799 例，共纳入危险因素63 个，仅体重指数 （BMI）>25 kg/m2、动脉血氧分压／吸人氧浓度（PaO2/FiO2）≤ 300 mmHg、体外循坏时间、机械通气延长、 ICU 留置时间合并效应量差异有统计学意义，是主动脉夹层发生低氧血症的预测因素；介入治疗是主动脉夹 层的保护因素。结论 该研究表明，BMI>25 kg/m2、PaO2/FiO2 ≤ 300 mmHg、体外循坏时间、机械通气延长、 ICU 留置时间是主动脉夹层并发低氧血症的危险因素。术后低氧血症是主动脉夹层的常见并发症，其危险因素 受多种因素的影响，仍需更大量前瞻性、高质量的研究进行分析。

Abstract:

Objective To identify the risk factors of hypoxemia in patients with acute aortic dissection by a meta-analysis. Methods Electronic literature databases in Chinese and English and references of included studies were searched. Those studies were independently screened by two evaluation persons in accordance with the standard, the information was extracted, the quality of studies was assessed in term of study design, meta-analysis was conducted by RevMan 5.0. Results Totally 7 studies were identified from 27 papers including 1,175 cases. There were 376 cases of hypoxemia and 799 cases of non-hypoxemia in the patients with aortic dissection. A total of 63 risk factors were involved; only BMI >25 kg/m2, PaO2/FiO2 ratio ≤ 300 mmHg, CPB time, prolonged mechanical ventilation and ICU stay time were identified as the predictive factors for hypoxemia in the patients with acute aortic dissection; and interventional therapy was identified as the protective factor for aortic dissection. Conclusions Our results suggest that BMI >25 kg/m2, PaO2/FiO2 ratio ≤ 300 mmHg, CPB time, prolonged mechanical ventilation, and ICU stay time are the risk factors for hypoxemia in the patients with acute aortic dissection. Postoperative hypoxemia is a common complication in the patients treated by aortic dissection surgery. Its risk factors are affected by many factors. Therefore, a large number of prospective and high-quality studies are still needed.