吸入一氧化氮治疗早产儿低氧性呼吸衰竭的Meta分析

目的 总结国内、外吸入一氧化氮（iNO）治疗早产儿低氧性呼吸衰竭（HRF）的研究结果，采用Meta分析方法综合评价iNO治疗早产儿HRF的临床效果，为临床应用提供指导。方法 制定原始文献的纳入标准、排除标准及检索策略，检索PubMed、EMBASE、Ovid、Springer和中国期刊全文数据库等，获得iNO治疗早产儿HRF的临床文献。使用国际Cochrane中心推荐的方法进行文献质量评估后，采用Review Manager 4.2软件对满足纳入标准的有关iNO治疗早产儿HRF的RCT研究进行Meta分析。选取iNO治疗组和对照组用氧时间、住院时间、机械通气时间、住院期间的病死率作为近期观察指标；支气管肺发育不良（BPD）、颅内出血或脑白质软化（PVL）、早产儿视网膜病（ROP）的发生率、1岁时随访脑瘫、Bayley 精神发育指数（MDI）或精神运动发育系数（PDI）＜2个标准差（s）的发生率为终点疗效观察指标，得出合并后疗效的优势比（OR）及其95％CI进行定性、定量综合评估。结果 共检索出883篇文献，对符合标准的9篇RCT文献进行Meta分析，漏斗图检验未发现发表偏倚。Meta分析结果显示治疗后的近期观察指标：3项RCT研究：用氧时间:iNO治疗组（n=259）低于对照组（n=259），P＜0.05；机械通气时间：iNO治疗组（n=259）和对照组（n=259）差异无统计学意义；住院时间：iNO治疗组（n=259）低于对照组（n=259），P＜0.05；6项住院期间患儿病死率的RCT研究:iNO治疗组（n=1 084）和对照组（=1 096）差异无统计学意义。治疗后的终点疗效观察指标：5项BPD发生率的RCT研究:iNO治疗组（n=618）低于对照组（n=621），P＜0.05；5项颅内出血（所有级别）发生率的RCT研究：iNO治疗组（n=787）和对照组（n=806）差异无统计学意义；4项颅内出血（3级或4级）发生率的RCT研究：iNO治疗组（n=767）发生率低于对照组（n=784），P＜0.05；5项颅内出血（3级或4级）或PVL发生率的RCT研究：iNO治疗组（n=1 021）和对照组（n=1 012）差异无统计学意义；2项阈值ROP发生率的RCT研究：iNO治疗组（n=608）和对照组（n=605）差异无统计学意义；4项ROP发生率的RCT研究：iNO治疗组（n=474）和对照组（n=463）差异无统计学意义；1项需要手术治疗ROP发生率的RCT研究：iNO治疗组（n=294）和对照组（n=288）差异无统计学意义；3项随访至1岁时脑瘫发生率的RCT研究：iNO治疗组（n=168）和对照组（n=184）差异无统计学意义；2项Bayley MDI＜2s发生率的RCT研究：iNO治疗组（n=156）和对照组（n=110）差异无统计学意义；2项Bayley PDI＜2s发生率的RCT研究：iNO治疗组（n=155）和对照组（n=167）差异无统计学意义。结论 现有的RCT研究结果还不支持iNO作为早产儿HRF的常规治疗方法，仍需要大样本、多中心RCT研究和远期神经系统发育情况随访来探讨iNO治疗早产儿HRF的确切疗效。

Meta-analysis of inhaled nitric oxide for hypoxaemic respiratory failure in preterm infants

Objective Inhaled nitric oxide（iNO） has been used in preterm infants with hypoxic respiratory failure (HRF). However, there is controversy about its therapeutic effect. In this article, The studies of iNO for HRF summarized and the effect of iNO for HRF was assessed in preterm infants using meta-analysis method. Methods Medical electronic databases and other sources were searched without language restriction.According to inclusion and exclusion cretria,articles were evaluated ( Shekelle PG, et al).RCTs were assessed according to the Juni assessment.Software Rev Man 4.2 was used to assess control trials, and other studies were reviewed and descrbied.Results Nine RCTs with 6 blind trials and 3 non-blind trials were selected from 883 articles. Publication bias was not found through funnel polt analysis. There were 3 RCTs that compared the oxygen use and the duration of ventilation, the length of hospitalization, including 259 patients in iNO and 259 in the control group. Compared to the control group, iNO could reduce the oxygen use (WMD＝－10.24，95％CI：－19.62,－0.86，P＜0.05); There was no difference in the duration of ventilation between iNO and the control group (WMD＝－4.87，95％CI：－10.68,0.94， P＞0.05); Compared to the control group, iNO could reduce the length of hospitalization (WMD＝－8.31，95％CI：－16.23,－0.38，P＜0.05). There were 6 RCTs that compared the death before discharge, including 1084 patients in iNO and 1096 in the control group. There was no difference in the deaths before discharge between iNO and the control group (OR＝0.95，95％CI：0.77,1.17，P＞0.05). There were 5 RCTs that compared the incidence of BPD among survivors at 36 weeks, including 618 patients in iNO and 621 in the control group. Compared to the control group, iNO could reduce the incidence of BPD (OR＝0.75，95％CI：0.58,0.95，P＜0.05). There was no difference in the incidence of intraventricular hemorrhage (all grades) between iNO and the control group (OR＝0.79，95％CI：0.62,1.00，P＝0.05). There were 4 RCTs that compared the incidence of intraventricular hemorrhage (grade 3 or 4), including 767 patients in iNO and 784 in the control group. Compared to the control group, iNO could reduce the incidence of intraventricular hemorrhage (grade 3 or 4) (OR＝0.84，95％CI：0.61,1.16，P＜0.05). There were 5 RCTs that compared the intraventricular hemorrhage (grade 3 or 4) or periventicular leukomacia, including 1021 patients in iNO and 1012 in the control. There was no difference in the incidence of intraventricular hemorrhage (grade 3 or 4) or periventicular leukomacia between iNO and the control group (OR＝0.95，95％CI：0.68,1.22，P＞0.05). There were 2 RCTs that compared the incidence of threshold ROP, including 608 patients in iNO and 605 in the control group. There was no difference in the incidence of threshold ROP between iNO and the control group (OR＝0.98，95％CI：0.72,1.34，P＞0.05). There were 4 RCTs that compared the incidence of ROP, including 474 patients in iNO and 463 in the control group. There was no difference in the incidence of ROP between iNO and the control group (OR＝1.08，95％CI：0.75,1.57，P＞0.05). There was 1 RCT that compared the incidence of ROP requiring surgery, including 294 patients in iNO and 288 in the control group. There was no difference in the incidence of ROP requiring surgery between iNO and the control group (OR＝1.05，95％CI：0.72,1.53，P＞0.05). There were 3 RCTs that compared the incidence of cerebral palsy, including 168 patients in iNO and 184 in the control group. There was no difference in the incidence of cerebral palsy between iNO and the control group (OR＝1.08，95％CI：0.75,1.57，P＞0.05). There were 2 RCTs that compared the incidence of Bayley MDI＜2SD, including 156 patients in iNO and 110 in the control group. There was no difference in the incidence of Bayley MDI＜2SD between iNO and the control group (OR＝0.84，95％CI：0.50,1.42，P＞0.05). There were 2 RCTs that compared the incidence of Bayley PDI＜2SD, including 155 patients in iNO and 167 in the control group. There was no difference in the incidence of Bayley PDI＜2SD between iNO and the routine therapy group (OR＝0.94，95％CI：0.57,1.58，P＞0.05). Conclusions The currently published evidence from randomised trials does not support the routine use of iNO in preterm infants with HRF. Further multicentre and large-scale RCTs are still needed to be performed to evaluate the potential benefit of iNO as routine therapy in preterm infants with HRF, and these trials need to be designed to evaluate not only the short term effects, but should be powered to evaluate neurondevelopmental outcome at a minimum of two years of age. Key words Nitric oxide；Infant；Premature；Respiratory insufficiency；Meta-analysis；Randomized controlled trial