4种疾病危重评分方法预测小早产或极低出生体质量儿死亡风险比较

PREDICTION OF MORTALITY RISK IN EXTREMELY PREMATURE OR VERY LOW BIRTH WEIGHT INFANTS:A COMPARISON AMONG FOUR METHODS

目的探讨4种疾病危重评分方法对NICU内小早产或极低出生体质量儿死亡风险评估的价值。方法对2010年1~8月我院NICU救治的74例胎龄<31周或出生体质量<1 500 g早产儿行新生儿危重病例评分法、新生儿临床危险指数(CRIB)、新生儿临床危险指数-Ⅱ(CRIB-Ⅱ)及新生儿多器官功能障碍(NEOMOD)评分。结果新生儿危重病例评分法显示,死亡组评分明显低于非死亡组(t=9.028,P<0.01);具备单项指标越多,死亡率越高(χ2=6.317,P<0.01)。死亡组CRIB评分明显高于非死亡组(t=-5.245,P<0.01);≥6分组病死率较≤5分组高(χ2=20.23,P<0.01)。死亡组CRIB-Ⅱ评分明显高于非死亡组(t=-5.413,P<0.05);≥6分组病死率较≤5分组高(χ2=11.11,P<0.01)。死亡组NEOMOD评分明显高于非死亡组(t=-10.092,P<0.01);≥6分组病死率较≤5分组高(χ2=41.26,P<0.01)。CRIB、CRIB-Ⅱ、NEOMOD与新生儿疾病危重评分均呈高度负相关(r=-0.774^-0.386,P<0.05)。对新生儿疾病危重评分、CRIB、CRIB-Ⅱ及NEOMOD进行ROC曲线分析,4种评分方法对预测小早产或极低出生体质量儿死亡风险均有显著意义(P<0.01);其曲线下面积(AUC)分别为0.923(SE 0.033)0、.814(SE 0.051)0、.807(SE 0.051)与0.949(SE 0.022)。结论 4种评分方法均适合我国新生儿临床实际,对小早产或极低出生体质量儿死亡危险度有很好的预测价值,评分间有很好的相关性。新生儿危重病例评分及NEOMOD评分对小早产或极低出生体质量儿的死亡风险预测的准确性更高。

ObjectiveTo compare the value of four methods in predicting risk of mortality in extremely preterm or very low birth weight infants.MethodsSeventy-four premature neonates-fetal age31 weeks or birth weight 1 500 g-were remedied in neonatal intensive care unit of our hospital from January to May,2010.A score was conducted for them by using neonatal critical illness score（NCIS）,clinical risk index for babies（CRIB）,clinical risk index for babies-Ⅱ（CRIB-Ⅱ） and neonatal multiple organ dysfunction score（NEOMOD）.ResultsThe NCIS indicated that the score in fatality group was lower than that in non-fatality group（t=9.028,P0.01）,the more single index to have,the higher mortality would be（χ2=6.317,P0.01）.The CRIB score in fatality group was higher than that in non-fatality group（t=-5.413,P0.01）,the case fatality of CRIB ≥6 score group was higher than that of ≤5 score group（χ2=20.23,P0.01）.The CRIB-Ⅱscore in fatality group was higher than that in non-fatality group（t=-5.413,P0.01）,the case fatality was higher in CRIB-Ⅱ ≥6 score group than ≤5 score group（χ2=11.11,P0.01）.The NEOMOD score in fatality group was higher than that in non-fatality（t=-10.092,P0.01）,the case fatality in NEOMOD≥6 score group was higher than that in ≤5 score group（χ2=41.26,P0.01）.CRIB,CRIB-Ⅱ and NEOMOD were negatively correlated with NCIS（r=-0.774——0.386,P0.05）.Receiver operating characteristics（ROC） curve analysis showed the areas under curve of NCIS,CRIB,CRIB-Ⅱ and NEOMOD were 0.923（SE 0.033）,0.814（SE 0.051）,0.807（SE0.051） and 0.949（SE 0.022）,respectively,indicating the use of the four methods for predicting mortality risk in premature or very low birth weight infants was of significance（P0.01）.ConclusionThe four evaluation methods discussed suit the neonatal clinical actuality in our country,which have a good value for prediction of mortality risk in these infants.An all-right correlation exists among the four assessment systems,of which,the prediction of NCIS and NEOMOD is more accurate.