产后出血预警评分表对产后出血预测效果的Meta分析
刘莹;黄群;韦小梅;叶文琴
【摘 要】Objective To evaluate the effects of postpartum hemorrhage (PPH )warning scoring tools. Methods Cochrane,JBI,Pubmed,Emba,CINAHL,CBM,CNKI and Wanfang databas were arched for articles from the establishment to June,2016.Prospective rearches focud on warning scoring tools of PPH were included.Two rearchers screened the literatures independently and data were extracted.The QUADAS was ud for asssment of the internal validity of the studies.Meta-analysis was conducted with Meta-Disc 1 .4 software.Results 9 studies were included,involving 6 scoring tools and a total number of 11934puerperae.There were 3 studies included,which adopted the scale developed by Chine PPH prevention and treatment group,for the meta-analysis of predictive accuracy.The Spearman correlation coefficient for analysis of diagnostic threshold was -0.500(P=0.667);the pooled SEN was 0.40(95%CI 0.36,0.45),with I2=91.1%(P<0.0001);the pooled SPE was 0.94(95%CI 0.93,0.94),with I2=93.8%(P<0.0001);the pool
ed DOR was 13.39(95%CI 7.04,25.49),with I2=77.3%(P=0.0124).Conclusions The accuracy of prent warning scoring system of PPH varies a lot.The scale developed by Chine PPH prevention and treatment group has a limited prediction compared with some modified scales,other accu-rate indicators can be ud in clinical trial.%目的 系统评价不同产后出血预测评分表对孕产妇产后出血的预测效果.方法 计算机检索Cochrane图书馆、JBI系统评价数据库、Pubmed、荷兰医学文摘数据库(the excerpta medica databa,Emba)、CINAHL、中国生物医学数据库(Chine biomedical literature databa,CBM)、中国期刊全文数据库(Chine national knowledge infrastructure,CNKI)及万方数据库,选择前瞻性地将产后出血预测评分表应用于孕产妇以预测产后出血发生率的相关研究文献.检索时间均从建库开始至2016年6月.由2位评价员独立筛选文献、提取资料,并采用质量评价工具(quality asssment of diagnostic accuracy studies,QUADAS)标准进行质量评价,使用Meta-Disc 1.4统计软件进行Meta分析.结果 共纳入9项研究,包含6个评分工具,总计11934例产妇.对全国产后出血防治组制定的评分表进行Meta分析,纳入了3项研究.其阈值检验的Spearman相关系数为-0.500(P=0.667);合并效应量灵敏度为0.40(95%CI 0.36,0.45),异质性I2=91.1%(P<0.0001);特异度为0.94(95%CI 0.93,0.94),异质性I2=93.8%(
P<0.0001);诊断比值比为13.39(95%CI7.04,25.49),异质性I2=77.3%(P=0.0124).结论 现有针对产后出血的预测评分表预测效果各不相同,全国产后出血防治组制定的评分表预测准确度有限,临床中可采用某些预测准确度更好的评分表.
【期刊名称】《解放军护理杂志》
【年(卷),期】2017(034)002
【总页数】5页(P13-17)
【关键词】产后出血;预警评分;预测效果;系统评价;Meta分析
【作 者】刘莹;黄群;韦小梅;叶文琴
【作者单位】第二军医大学附属长海医院 护理部,上海 200433;中国福利会国际和平妇幼保健院 护理部,上海 200003;第二军医大学附属长海医院 护理部,上海 200433;第二军医大学附属长海医院 护理部,上海 200433
【正文语种】中 文
【中图分类】R473.71
【 Abstract 】 Objective To evaluate the effects of postpartum hemorrhage(PPH)warning scoring tools.Methods Cochrane,JBI,Pubmed,Emba,CINAHL,CBM,CNKI and Wanfang databas were arched for articles from the establishment to June,2016.Prospective rearches focud on warning scoring tools of PPH were included.Two rearchers screened the literatures independently and data were extracted.The QUADAS was ud for asssment of the internal validity of the studies.Meta-analysis was conducted with Meta-Disc 1.4 software.Results 9 studies were included,involving 6 scoring tools and a total number of 11 934 puerperae.There were 3 studies included, which adopted the scale developed by Chine PPH prevention and treatment group,for the meta-analysis of predictive accuracy.The Spearman correlation coefficient for analysis of diagnostic threshold was -0.500(P=0.667);the pooled SEN was 0.40(95%CI 0.36,0.45),with I2=91.1%(P<0.0001);the pooled SPE was 0.94(95%CI 0.93,0.94),with I2=93.8%(P<0.0001);the pooled DOR was 13.39(95%CI 7.04,25.49),with I2=77.3%(P=0.0124).Conclusions The accuracy of prent warning scoring system of PPH varies
a lot.The scale developed by Chine PPH prevention and treatment group has a limited prediction compared with some modified scales,other accurate indicators can be ud in clinical trial.
【 Key words 】 postpartum hemorrhage;warning scoring;predictive effects;systematic review;Meta-analysis
目前,随着“二胎政策”的全面开放及以往“独生子女政策”导致的剖宫产后相关问题,增加了产后出血发生的可能性。而产后出血是造成孕产妇死亡的首位因素,现有研究[1-9]采用了一些产后出血的预警评分表来预测产后出血发生的概率,但各评估工具对产后出血的预测效果存在一定差异。选择真正可靠的评估工具,能够帮助医护人员筛选出高危人员并在分娩过程中尽早做好干预准备。为此,本研究针对孕产妇发生产后出血的预测评分工具进行系统回顾,总结可靠的评分工具,为临床中判断孕产妇发生产后出血的高危因素提供一定的理论依据。
1.1 文献检索策略 使用计算机检索计算机检索Cochrane图书馆、JBI系统评价数据库、Pubmed、荷兰医学文摘数据库(the excerpta medica databa,Emba)、CINAHL、中
国生物医学数据库(Chine biomedical literature databa,CBM)、中国期刊全文数据库(Chine national knowledge infrastructure,CNKI)及万方数据库。首先查找是否有产后出血预警评分的最新同类系统评价。若无,则进一步查找关于产后出血的预警及高危评分的原始研究文献,检索时间均从建库开始至2016年6月。此外,对于纳入文献的参考文献进行追溯,补充纳入符合标准的研究。分别以“postpartum hemorrhage AND early warning/prediction/warning system/warning scoring”及“产后出血、高危预警/预警系统/预测评分”为检索词检索相应的中英文数据库。
1.2 纳入与排除标准 纳入标准:(1)研究类型,选择期刊、学位及会议论文的中英文前瞻性队列研究。(2)研究对象,纳入医院分娩的孕产妇。(3)干预措施,在胎儿娩出前使用预警评分表对孕妇进行高危评分,同时在分娩过程中使用客观方法计算产后出血量。(4)结局指标,产后出血发生率、灵敏度(nsitivity,SEN)、特异度(specificity,SPE)、阳性预测值、阴性预测值等。若无全面指标,至少应包含SPE及SEN的结果,或能够通过原始数据计算得出两项指标结果。排除标准:回顾性研究、结局指标中缺乏评分表的内容或可靠性、未采用客观测量出血量方法的文献。
1.3 文献筛选与资料提取
1.3.1 文献筛选 由2名经过循证护理培训的研究人员分别按照纳入与排除标准对文献进行筛选。首先阅读文题,对符合要求的文章进一步阅读摘要、全文。若通读全文后文献的内容符合纳入排除标准,则予以纳入,否则予以舍弃。如在核对过程出现分歧,则由第3名专业人员协助判断。
1.3.2 资料提取 制定出文献信息提取表,主要内容有:(1)所纳入研究的基本信息,如第一作者、发表时间、研究地区;(2)被评估产妇的基本情况,如纳入例数、年龄、分娩方式、出血量测量方法。由于部分研究中包含了未评分组,故本研究仅采集评估组资料;(3)使用量表的结局指标,如产后出血发生率、预测结果列联表[即四格表,包括真阳性(true positive,TP)、假阳性(fal positive,FP)、假阴性(fal negative,FN)及真阴性(true negative,TN)]、SEN及SPE(无此结果报道的,根据四格表计算出相应数据)等。
1.4 纳入研究的文献质量评价 由参与上一步的两名研究人员按照诊断准确性研究的质量评价工具(quality asssment of diagnostic accuracy studies,QUADAS)标准对产后出血预测评分表进行评价,主要包括偏倚风险和临床适用性两个方面,针对病例选择、待评价的诊断方法、参考标准及研究流程分别进行评价[10],结果包含“高风险、不清楚及低风险”三个类别。