| Part One Establishment and analysis of prognostic model in children with sepsis based on pSOFA scoreObjective: To explore the risk factors of sepsis in children diagnosed based on the pediatric sequential organ failure assessment(pSOFA)score,and to establish and evaluate the related prognostic models.Method: The data of children with sepsis admitted to Hebei Children’s Hospital from January 1,2019 to December 31,2021 were retrospectively analyzed.According to the prognosis,they were divided into survival group and death group.Independent risk factors were analyzed by multivariate Logisitic regression,and prognostic model was established to analyze the predictive value for children with sepsis.Results: A total of 197 children with sepsis were included,with 160 in the survival group and 37 in the death group.The results of multivariate Logistic regression analysis showed that shock(OR=4.206,95%CI:1.789-9.891),febrile neutropenia(FN)(OR=3.605,95%CI: 1.374-9.459)and pSOFA score(OR=1.229,95%CI: 1.100-1.373)were independent risk factors for sepsis death in children.Model equations=-5.137+1.437× shock+1.282×FN+0.206×pSOFA score,the area under the ROC curve of the prediction model(0.824)is greater than that of shock(0.645),FN(0.658),and pSOFA score(0.729)alone,and the sensitivity and specificity of the model were 73.0% and 87.5%,respectively.The Bootstrap method was used to internally sample 1000 times for verification,and the calibration curve judged that the predicted value was more consistent with the actual value.Hosmer-Lemeshow test was used to judge the goodness of fit,which showed well calibration ability(P=0.344 > 0.05)of the model.Decision curve analysis showed that high-risk thresholds between 0.06 and 0.91 had higher net benefits,suggesting that the predictive model had higher clinical value.Conclusions: FN,shock and pSOFA score are independent risk factors for sepsis death in children.The prediction model has a good clinical predictive value and has practical significance for the prognosis evaluation and treatment guidance of children.Part Two Predictive value of the model and application value of hemodynamic monitoring in children with refractory septic shockObjective: To evaluate the predictive value of the model on the outcome of children with refractory septic shock(RSS),and to explore the non-invasive cardiac output monitoring system(NICaS)and pulse-indicated continuous cardiac output monitoring(Pi CCO)application value in children with RSS.Methods: The clinical data of children with RSS admitted to our hospital from January 1,2019 to June 30,2022 were collected,and the risk of death was evaluated by the model.Grouped according to the hemodynamic monitoring method adopted,those who performed Pi CCO and NICaS monitoring at the same time were group A,and those who only performed NICaS monitoring were group B.Stroke volume variation(SVV),cardiac index(CI),systemic vascular resistance index(SVRI),global ejection fraction(GEF)and granov goor Index(GGI)were monitored.Paired comparisons were made between NICaS and Pi CCO in group A,and comparisons were made between the data of groups A and B,and the data measured at 1h,6h,and 24 h after shock diagnosed in children with RSS were compared.Results: The sensitivity and specificity of the model for predicting death in children with RSS were 77.3% and 91.7%,respectively,and the area under the ROC curve was 0.898.The paired T-test of Pi CCO and NICaS data in group A showed that the SVRI and SVV measured by Pi CCO were greater than those of NICaS,but no significant difference(P>0.05);Bland-Altman analysis showed good consistency on the CI measured by Pi CCO and NICaS.Pearson correlation analysis showed a significant positive correlation between GGI and GEF.There was no significant difference in CI,SVRI and SVV between group A and B(P>0.05).By analysing the data measured at 1h,6h and 24 h after shock recognition,we found that the CI experienced the undulation of rising first,and then falling,CI(6h)was higher than CI(1h),but without significantly difference(P>0.05),CI(24h)of the whole group was significantly lower than CI(1h)(P<0.05).The SVRI fell first and rose later,the SVRI of 1h and 24 h were higher than that of 6h,and the differences were statistically significant,while SVRI(24h)had no significant change compared with SVRI(1h)(P>0.05).The SVV was gradually decreased,SVV(6h)and SVV(24h)were lower than that of SVV(1h),and the difference between SVV(24h)and SVV(1h)was statistically significant(P<0.05).Conclusions: The prediction model can effectively predict the risk of death in children with RSS.The results of CI,SVRI and SVV in children with RSS monitored by NICaS are in good agreement with Pi CCO.NICaS can provide simple,non-invasive,continuous and reliable hemodynamic measurement for children with RSS.Part Three mNGS for identifying pathogens in febrile neutropenic children with hematological diseaseObjective: To investigate the application value of metagenomic next-generation sequencing(mNGS)in children with hematological diseases presenting with febrile neutropenic(FN).Methods: Clinical data of hematological children with FN who admitted to our hospital from January 1,2018 to January 1,2021 were retrospectively analyzed.The pathogen recognition efficiency of mNGS was compared with traditional pathogen detection(TPD),and the impact on treatment options and prognosis were compared between mNGS positive group and negative group.Results: A total of 49 cases were included in the study,including 24 males and 25 females,ranging in age from 4 months to 13 years old,and 6died(n=6,12.2%).A total of 61 clinical samples(49 plasma samples and 12 alveolar lavage fluid samples)of 49 children were tested by TPD and mNGS,and 77 pathogenic strains were ultimately confirmed,of which 70 strains were detected by mNGS,19 strains by TPD,and 12 beads were co-positive for mNGS and TPD.Aspergillus and G-bacterias were the predominant strains in FN children who developed bloodstream infections.Among the 49 children with FN,mNGS detected 17 cases of simple infection,25 cases of two or more pathogenic infections,and 7 cases of no pathogens were detected;TPD detected 19 cases of simple infection,and no mixed infection was found.The detection rate of mNGS pathogens and the detection rate of more than 2 mixed pathogens were higher than that of TPD,with statistical difference(P<0.05).According to the mNGS results,the impact of the anti infective regimen was divided into two groups:(1)A1=39 cases,which were treated with new targeted anti infective drugs;A2=3 cases,those who have downgraded or stopped taking anti infective drugs;(2)B=7 cases,due to negative mNGS results,the anti infection regimen was not adjusted.The detection time of mNGS in group A was earlier than that in group B,and with lower mortality and drug-related adverse events(DRAE)than group B,and the results were statistically different(P<0.05).Conclusions: Children with hematological disorders complicated by FN after chemotherapy have low immunity and a high mortality rate.Early mNGS can effectively improve the efficacy of pathogen detection,and precise treatment after clarifying the pathogen can reduce mortality and avoid antibiotic abuse. |
PDF Full Text Request