Jian Cheng, Xiao Dong, Yang Yang, Xiaohan Qin, Xing Zhou, Da Zhang
{"title":"Synergistic machine learning models utilizing ferroptosis-related genes for improved neuroblastoma outcome prediction.","authors":"Jian Cheng, Xiao Dong, Yang Yang, Xiaohan Qin, Xing Zhou, Da Zhang","doi":"10.21037/tp-24-323","DOIUrl":"10.21037/tp-24-323","url":null,"abstract":"<p><strong>Background: </strong>Neuroblastoma (NB) is a highly heterogeneous and common pediatric malignancy with a poor prognosis. Ferroptosis, an iron-dependent cell death pathway, may play a crucial role in NB tumor progression and immune response. This study aimed to investigate ferroptosis in NB to identify potential therapeutic targets and develop predictive models for prognosis and recurrence.</p><p><strong>Methods: </strong>Six datasets were accessed from the ArrayExpress database and Gene Expression Omnibus. Ferroptosis-related genes (FRGs) were selected from the FerrDb website. Unsupervised clustering, differential expression analysis, weighted correlation network analysis (WGCNA), and gene set enrichment analysis (GSEA) were adopted to investigate potential pathways associated with ferroptosis in NB and identify the key genes involved. We used the least absolute shrinkage and selection operator (LASSO) and multivariate Cox regression to develop the ferroptosis-related prognostic signatures (FRPS) while using machine learning (ML) algorithms to construct the recurrence model.</p><p><strong>Results: </strong>Ribosome and cell cycle may be the potential pathways for ferroptosis involved in NB, with <i>MYCN</i> and <i>RRM2</i> identified as key genes in this regulatory process. Five FRGs-<i>ATG7</i> (-1.009), <i>ELAVL1</i> (1.739), <i>PPARA</i> (0.493), <i>RDX6</i> (1.457), and <i>TERT</i> (0.247)-were screed out for the FRPS, which showed excellent predictive performance in comparison with other published NB signatures. Eight FRGs-<i>ALDH3A2</i> (48.597), <i>TERT</i> (23.398), <i>ULK2</i> (21.034), <i>AKR1C1</i> (20.699), <i>MFN2</i> (12.575), <i>SLC16A1</i> (12.342), <i>TF</i> (10.240), and <i>DDR2</i> (7.598)-were selected based on the importance scores to construct the recurrence model. Among the models, utilizing random forest (RF), XGboost, support vector machine (SVM), K-nearest neighbors (KNN), and linear discriminant analysis (LDA), the RF model exhibited the highest performance.</p><p><strong>Conclusions: </strong>We investigated the potential ferroptosis-related pathways and hub- FRGs in NB and developed prognosis and recurrence models, providing new potential targets for prognostic evaluation and treatment in NB patients.</p>","PeriodicalId":23294,"journal":{"name":"Translational pediatrics","volume":"13 12","pages":"2164-2182"},"PeriodicalIF":1.5,"publicationDate":"2024-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11732634/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143012450","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Stiljan Hoxha, Giulia Poretti, Jacopo Gardellini, Filippo Tonelli, Eleonora De Laurentis, Marco Parolo, Tiziano Menon, Antonia Maria Prioli, Giovanni Battista Luciani
{"title":"Venoarterial extracorporeal membrane oxygenation using magnetic levitation centrifugal pumps for fulminant myocarditis in infants, children and young adults.","authors":"Stiljan Hoxha, Giulia Poretti, Jacopo Gardellini, Filippo Tonelli, Eleonora De Laurentis, Marco Parolo, Tiziano Menon, Antonia Maria Prioli, Giovanni Battista Luciani","doi":"10.21037/tp-24-29","DOIUrl":"10.21037/tp-24-29","url":null,"abstract":"<p><strong>Background: </strong>Fulminant myocarditis (FM) is a potentially lethal disease with a wide spectrum of clinical presentation, thus making the diagnosis hard to depict. In cases where acute circulatory failure occurs venoarterial (VA) extracorporeal membrane oxygenation (ECMO) support is a valid management strategy, especially in the pediatric and adult patients. This study aims to report the results of VA ECMO for FM in our Institution.</p><p><strong>Methods: </strong>Between April 2009 and January 2021, 17 consecutive patients presenting with FM were supported using VA ECMO. We collected data dividing the population between infants, children and young adults. There were 8 male and 9 female patients, median age of 5.2 years [interquartile range (IQR), 2 months to 32 years] and median body weight of 16 kg (IQR, 3.8-56 kg). FM etiology was viral in 8 patients (47%), bacterial in 1 (6%), 2 giant cellular myocarditis (12%) and in 6 patients (35%) the etiology was unknown. Where it was possible also a cardiac biopsy was performed, usually during ECMO or vent implant. The endpoints of the study are: survival, incidence of ECMO-related complications, weaning rate from ECMO, recovery of cardiac function and the association between rescue ECMO (r-ECMO) and no-weaning/mortality. For the survival analysis we divided the population into three groups: infants (0-2 months; n=5), children (2-6 years; n=5) and young adults (24-40 years; n=7).</p><p><strong>Results: </strong>After a median duration of 168 hours (IQR, 120-240 hours), 13 patients were weaned from support (weaning rate 76 %), 2 (12%) underwent respectively cardiac transplantation and Bi-Vad Berlin Heart implantation and 2 (12%) died while on ECMO support. Bleeding occurred in 6 (35%) patients, infection in 8 (47%) and 5 (29%), all of them infants, required peritoneal dialysis for acute kidney injury. Overall mortality was 35% (6/17 patients): two patients died during ECMO support due to persistent cardiac failure, arrhythmias and bleeding; three patients died after ECMO weaning (2 due to recurrent ventricular failure and 1 for respiratory complications) and 1 died due to multi-organ failure (MOF) after Bi-VAD implantation. Overall survival at follow-up was 65%, with a statistically significant difference (P=0.05) between age groups: in the infant group was 20% (1/5), in the children group 60% (3/5) and in the adult group 100% (7/7). There was no association between r-ECMO and weaning failure (P=0.55) or hospital mortality (P>0.99). During a median follow-up of 76 months (IQR, 52-99 months), there were no late deaths, 1 patient presented a minor neurological sequela, while cardiac function had fully recovered in all late survivors.</p><p><strong>Conclusions: </strong>The present experience shows that VA ECMO is an effective bridge to myocardial function recovery in patients with FM, including those with circulatory collapse. The rate of hospital complications is decent when in light of th","PeriodicalId":23294,"journal":{"name":"Translational pediatrics","volume":"13 12","pages":"2233-2241"},"PeriodicalIF":1.5,"publicationDate":"2024-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11732621/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143012451","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Differential genes expression of immune tolerance induction in hemophilia A: an exploratory RNA-seq test from a Chinese hemophilia comprehensive care centre.","authors":"Jialu Zhang, Zekun Li, Guoqing Liu, Wanru Yao, Di Ai, Zhengping Li, Zhenping Chen, Runhui Wu","doi":"10.21037/tp-24-300","DOIUrl":"https://doi.org/10.21037/tp-24-300","url":null,"abstract":"<p><strong>Background: </strong>The production of inhibitors is a serious complication that can arise during coagulation factor replacement therapy for hemophilia A (HA). The primary therapeutic strategy to eliminate inhibitors is immune tolerance induction (ITI), which is known to be an extremely challenging, prolonged, and costly treatment. With the widespread use of RNA sequencing (RNA-seq) to analyze differentially expressed genes (DEGs) across various treatment outcomes, there is potential for predicting ITI outcomes. This study aims to use RNA-seq to test differently expressed genes in different outcomes of ITI treatment for HA patients with high-titer inhibitor (HAI), to explore its prediction possibility.</p><p><strong>Methods: </strong>RNA-seq was employed to screen and compare the DEGs between patients in the Success group and those in the Failure group, based on ITI clinical outcomes. DEGs were subjected to Gene Ontology (GO) analysis and enrichment analysis of Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways.</p><p><strong>Results: </strong>Thirteen analyzable HAI cases were collected, comprising seven in the Success group and six in the Failure group. Blood samples were taken before and after ITI. RNA-seq was applied to all samples to screen for expressed genes. In the Success group, a total of 4,967 messenger RNA (mRNA) transcripts were differentially expressed between pre-ITI and post-ITI, with 2,865 being up-regulated and 2,102 down-regulated. In the Failure group, 515 mRNA transcripts were expressed either before or after ITI, showing up-regulation in 68.7% (354/515) and down-regulation in 31.3% (161/515).</p><p><strong>Conclusions: </strong>The increased expression of genes which related to immune system activation suggests a possibly favorable therapeutic outcome of ITI. Future studies should test with a larger cohort to validate these findings.</p>","PeriodicalId":23294,"journal":{"name":"Translational pediatrics","volume":"13 12","pages":"2110-2117"},"PeriodicalIF":1.5,"publicationDate":"2024-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11732629/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143012419","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Kun-Quan Tang, Miao-Yuan Li, Qin-Yuan Guo, Juan Wang, Huo-Qing Lu, Bei Wang
{"title":"Impact of Re-Du-Ning enema treatment on the intestinal microflora in pediatric patients with hand, foot, and mouth disease.","authors":"Kun-Quan Tang, Miao-Yuan Li, Qin-Yuan Guo, Juan Wang, Huo-Qing Lu, Bei Wang","doi":"10.21037/tp-24-257","DOIUrl":"https://doi.org/10.21037/tp-24-257","url":null,"abstract":"<p><strong>Background: </strong>Hand, foot, and mouth disease (HFMD) is a prevalent infectious condition in children. This study aimed to assess the regulatory effects of Re-Du-Ning on the intestinal microflora of pediatric patients with HFMD.</p><p><strong>Methods: </strong>Fecal samples were collected from children affected by HFMD, who were diagnosed at the traditional Chinese medicine pediatrics outpatient and emergency departments of Liuzhou Women and Children's Healthcare Hospital, as well as from healthy children undergoing physical examinations at the same hospital during the same period. DNA was extracted from these samples and subjected to 16S ribosome DNA amplicon sequencing. The sequenced data were categorized, quantified, and compared. Analyses involved creating relative abundance bar graphs, constructing unweighted pair-group method with arithmetic mean clustering trees, and generating heatmaps of clustering to evaluate the variations in abundance and diversity across different groups. The analysis of molecular variance and <i>t</i>-test were used to analyze structural differences in microbial flora between groups, and linear discriminant analysis was used to identify significant differences in microbial genera between the groups.</p><p><strong>Results: </strong>A total of 67 fecal samples were collected from children with HFMD (13 in the intravenous group, 40 in the enema group) and from healthy children (14 in the healthy group). When compared with the healthy group, the intestinal microflora diversity and similarity were highest after enema treatment, although the microbial structure exhibited significant changes (weighted_unifrac, P<0.05). The composition of species relative abundance was comparable between the healthy group and the post-enema group.</p><p><strong>Conclusions: </strong>Re-Du-Ning enema treatment regulated the intestinal microflora in these children, significantly increasing the abundance of probiotics like <i>Bifidobacteria</i> and reducing the abundance of opportunistic pathogens like <i>Enterobacter</i>.</p>","PeriodicalId":23294,"journal":{"name":"Translational pediatrics","volume":"13 12","pages":"2118-2133"},"PeriodicalIF":1.5,"publicationDate":"2024-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11732627/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143012436","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Yingyi Qi, Yao Zhang, Xiaoxiao Liu, Yun Li, Rongxiu Zheng
{"title":"Preliminary comparison of net gain in final adult height of girls with early menarche treated with or without gonadotropin-releasing hormone agonist.","authors":"Yingyi Qi, Yao Zhang, Xiaoxiao Liu, Yun Li, Rongxiu Zheng","doi":"10.21037/tp-24-348","DOIUrl":"https://doi.org/10.21037/tp-24-348","url":null,"abstract":"<p><strong>Background: </strong>Early menarche is associated with both physical and psychosocial problems. Based on psychological and physical health considerations, for girls with early menarche, some parents and physicians may elect to use gonadotropin-releasing hormone agonists (GnRHa) to delay menstruation. This study aimed to explore the effects of GnRHa treatment on the final height of girls with early menarche and build the models to predict the final adult height (FAH).</p><p><strong>Methods: </strong>Girls who experienced menarche between the ages of 8 and 10 years and were diagnosed with idiopathic central precocious puberty (ICPP) at Tianjin Medical University General Hospital between July 2017 and August 2023 were included in this study. Participants were divided into two groups based on treatment strategy: GnRHa-treated and GnRHa-untreated groups. Laboratory parameters including growth factors and basal gonadotropins were tested at diagnosis. The heights and weights of the participants were measured every three months. Bone radiographs of the left hand and wrist were assessed by professional appraisers to determine the bone age (BA), which was measured every 6 months after diagnosis.</p><p><strong>Results: </strong>Clinical data of 176 girls who experienced early menarche were retrospectively analyzed. For participants in the GnRHa-treated group (n=87), growth velocity (GV) showed significant differences between the first 6 months and second 6 months of treatment (P=0.01; 5.82±2.3 <i>vs.</i> 4.79±2.31 cm, respectively). The height standard deviation score (SDS) and BA (P<0.001) decreased during treatment. The predicted adult height was higher at the end of treatment, but was not statistically different from that at diagnosis (P=0.73). In the linear regression analysis, no significant relationships were observed between GnRHa treatment and net gain (NG) in final height [Model A, adjusted for BA and chronological age (CA) at baseline: P=0.43; Model B, adjusted for Model A plus HtSDS-BA, HtSDS, and BMISDS: P=0.65; Model C, adjusted for Model B plus LH, FSH, and IGF-1: P=0.82]. The generalized additive model (GAM) for NG in final height in GnRHa-treated participants included three independent risk factors: LH/FSH [estimated degrees of freedom (edf) =5.36, P=0.02], GV (edf =4.11, P=0.007), and the bone maturation ratio (BMR) (edf =4.79, P=0.02). GAM performed better than multivariate linear (stepwise) regression in predicting the FAH in GnRHa-treated girls with early menarche.</p><p><strong>Conclusions: </strong>For girls who experienced menarche between the ages of 8 and 10 years, GnRHa treatment suppressed GV and skeletal maturity. The GAM provides a theoretical basis for pediatric endocrinologists in deciding whether to apply GnRHa treatment, determining the time to withdraw GnRHa treatment, and predicting the FAH of girls with early menarche.</p>","PeriodicalId":23294,"journal":{"name":"Translational pediatrics","volume":"13 12","pages":"2204-2213"},"PeriodicalIF":1.5,"publicationDate":"2024-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11732639/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143012449","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Birth weight and ponderal index percentiles for twins based on sex and chorionicity in a center of Guangdong Province, China.","authors":"Qian Liu, Jiying Wen, Liyuan Fang, Yanlin Huang, Tengzi Rao, Xiaomei Shi, Jing Wu","doi":"10.21037/tp-24-308","DOIUrl":"10.21037/tp-24-308","url":null,"abstract":"<p><strong>Background: </strong>Intrauterine growth restriction (IUGR) which is judged based on birth weight and gestational age, is associated with increased neonatal mobility and mortality and also has a further impact on physical and mental health during later in life. Using the birth weight percentile for singletons to assess twins might not accurately reflect the growth status of the twins; this could potentially lead to an incorrect evaluation of growth-restricted children. For a more precise assessment of twin newborns, it is beneficial to utilize twin-specific birth weight percentile curves and ponderal index (PI) curves that consider factors such as birth order and sex. The aim of this study is to establish a contemporary up-to-date method that provides sex-specific percentiles for birth weight and the PI of twin neonates, categorized by gestational age and differentiated by chorionicity within a defined cross-sectional study.</p><p><strong>Methods: </strong>We retrospectively analyzed the birth weight and PI percentile of 3,433 twins born at a gestational age between 25 and 40 weeks based on sex and chorionicity, between 2015 and 2020 in a single center in southern China. The smoothed percentile curves were drawn via generalized additive models for location, scale and shape (GAMLSS).</p><p><strong>Results: </strong>A total of 3,433 live-born twins were included in this study and consisted of 366 monochorionic girls, 405 monochorionic boys, 1,237 dichorionic girls and 1,425 dichorionic boys. Compared with dichorionic twins, the mean birth weight of monochorionic twins was significantly lower and the gestational week of delivery was earlier. The calculated 50th percentile birth weight values for twin boys were higher than those for girls based on the same chorionicity, except for those male monochorionic twins who were born before week 31 of gestation.</p><p><strong>Conclusions: </strong>The birth weight and PI percentile charts for neonates, categorized by sex and chorionicity, serve as valuable tools for assessing the wellbeing of twin infants. Utilizing these charts could reduce the overdiagnosis of small for gestational (SGA) age in normally growing twins.</p>","PeriodicalId":23294,"journal":{"name":"Translational pediatrics","volume":"13 12","pages":"2221-2232"},"PeriodicalIF":1.5,"publicationDate":"2024-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11732626/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143012147","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Long-term changes in the gut microbiota before and after bismuth quadruple therapy or concomitant therapy in children with peptic ulcers receiving <i>Helicobacter pylori</i> eradication.","authors":"Chunmeng He, Yuhuan Wang, Zifei Tang, Junping Lu, Ying Zhou, Ying Huang","doi":"10.21037/tp-24-207","DOIUrl":"10.21037/tp-24-207","url":null,"abstract":"<p><strong>Background: </strong><i>Helicobacter pylori</i> (<i>H. pylori</i>) is a gram-negative spiral bacterium that can cause peptic ulcers. The effects of bismuth quadruple therapy and concomitant therapy on fecal ecology among children with peptic ulcers remain largely unknown. According to guidelines, these patients should be treated. Therefore, the aim of this study is to investigate the short- and long-term effects of bismuth quadruple therapy and concomitant therapy on gut microbiota in children with peptic ulcer disease (PUD).</p><p><strong>Methods: </strong>Forty-two patients with PUD were enrolled and collected their fecal samples in Shanghai. A 16S ribosomal RNA (rRNA)-based analysis was conducted. The impacts of bismuth quadruple therapy and concomitant therapy on the gut microbiota were compared before eradication and at 2, 6 and 52 weeks (12 months) after eradication.</p><p><strong>Results: </strong>Whatever bismuth quadruple therapy and concomitant therapy, the diversity and richness of gut microbiota markedly decreased and fecal microbiota structure markbly changed at week 2. At week 6, the composition a returned to the baseline level; nevertheless, the diversity and richness gradually returned to baseline until week 52. Compared to those in the healthy control group, the relative abundances of 90% (18/20) of the bacterial genera differed at week 2, 15% (3/20) of the bacterial genera differed at week 6, and 10% (2/20) of the bacterial genera were not restored at week 52.</p><p><strong>Conclusions: </strong>Both bismuth-containing therapy and concomitant therapy resulted in transient changes of the gut microbiota in Chinese children, and these changes nearly returned to pre-eradication levels 1 year after eradication, suggesting the long-term safety of <i>H. pylori</i> therapy in children with peptic ulcers.</p>","PeriodicalId":23294,"journal":{"name":"Translational pediatrics","volume":"13 11","pages":"1947-1958"},"PeriodicalIF":1.5,"publicationDate":"2024-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11621894/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142802253","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"A 6-year-old girl with gout: a case report and another unusual presentation.","authors":"Shengyou Yu, Junbo Liang, Qi Ren","doi":"10.21037/tp-24-354","DOIUrl":"10.21037/tp-24-354","url":null,"abstract":"<p><strong>Background: </strong>Gout is the most common inflammatory arthritis in adults, and although gout has been trending younger in recent years, children with gout before the age of 10 years are still rare cases, especially asymptomatic children with gout, making gout in children easy to misdiagnose. This study aimed to describe the clinical characteristics, laboratory tests, imaging findings, and treatment outcomes of a 6-year-old girl with gout.</p><p><strong>Case description: </strong>This case presents a detailed report of a 6-year-old girl diagnosed with primary gout, a condition that is exceedingly rare in pediatric patients, especially those asymptomatic. The patient exhibited a swelling on her right foot that was initially misdiagnosed multiple times before being correctly identified as gout. This patient experienced a marked improvement in their condition following the administration of the prescribed treatment. Furthermore, subsequent follow-up assessments have confirmed the absence of any recurrence. The case highlights the diagnostic challenges faced due to atypical presentations and lack of pediatric-specific diagnostic criteria for hyperuricemia and gout. In this study, we provided a comprehensive discussion on the genetic and metabolic factors contributing to gout, as well as the importance of early and accurate diagnosis to prevent long-term complications.</p><p><strong>Conclusions: </strong>This case report provides valuable insights into the rare occurrence of gout in children, emphasizing the importance of early and accurate diagnosis and treatment for primary gout in children, emphasizing the need for heightened awareness and accurate diagnostic criteria among pediatric clinicians.</p>","PeriodicalId":23294,"journal":{"name":"Translational pediatrics","volume":"13 11","pages":"2077-2087"},"PeriodicalIF":1.5,"publicationDate":"2024-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11621891/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142802224","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Xuefeng Tan, Xiufang Zhang, Jie Chai, Wenjuan Ji, Jinling Ru, Cuilin Yang, Wenjing Zhou, Jing Bai, Yueling Xiong
{"title":"Constructing a predictive model for early-onset sepsis in neonatal intensive care unit newborns based on SHapley Additive exPlanations explainable machine learning.","authors":"Xuefeng Tan, Xiufang Zhang, Jie Chai, Wenjuan Ji, Jinling Ru, Cuilin Yang, Wenjing Zhou, Jing Bai, Yueling Xiong","doi":"10.21037/tp-24-278","DOIUrl":"10.21037/tp-24-278","url":null,"abstract":"<p><strong>Background: </strong>The clinical characteristics of neonatal sepsis (NS) are subtle and non-specific, posing a serious threat to the lives of newborn infants. Early-onset sepsis (EOS) is sepsis that occurs within 72 hours after birth, with a high mortality rate. Identifying key factors of NS and conducting early diagnosis are of great practical significance. Thus, we developed a robust machine learning (ML) model for the early prediction of EOS in neonates admitted to the neonatal intensive care unit (NICU), investigated the pivotal risk factors associated with EOS development, and provided interpretable insights into the model's predictions.</p><p><strong>Methods: </strong>A retrospective cohort study was conducted. This includes 668 newborns (EOS and non-EOS) admitted to the NICU of Bozhou People's Hospital from January to December 2023, excluding 72 newborns born more than three days ago and 166 newborns with medical record data missing more than 30%. Finally, 430 newborns (EOS and non-EOS) were included in the study. Clinical case data were meticulously analyzed, and the dataset was randomly partitioned, allocating 75% for model training and the remaining 25% for test. Data preprocessing was meticulously performed using R language, and the least absolute shrinkage and selection operator (LASSO) regression was implemented to select salient features, mitigating the risk of overfitting. Six ML models were leveraged to forecast the incidence of EOS in neonates. The predictive performance of these models was rigorously evaluated using the receiver operating characteristic (ROC) curve and precision-recall (PR) curve. Furthermore, the SHapley Additive exPlanations (SHAP) framework was employed to provide intuitive explanations for the predictions made by the Categorical Boosting (CatBoost) model, which emerged as the top performer.</p><p><strong>Results: </strong>The ROC area under the curve (ROCAUC) of six ML models, CatBoost, random forest (RF), eXtreme Gradient Boosting (XGBoost), multilayer perceptron (MLP), support vector machine (SVM), logistic regression (LR) all exceeded 0.900 on the test set. Especially the CatBoost model exhibited superior performance, with favorable outcomes in calibration, decision curve analysis (DCA), and learning curves. Notably, the ROCAUC attained 0.975, and the area under the PR curve (PRAUC) reached 0.947, signifying a high degree of predictive accuracy. Utilizing the SHAP method, seven key features were identified and ranked by their importance: respiratory rate (RR), procalcitonin (PCT), nasal congestion (NC), yellow staining (YS), white blood cell count (WBC), fever, and amniotic fluid turbidity (AFT).</p><p><strong>Conclusions: </strong>By constructing a precision-oriented ML model and harnessing the SHAP method for interpretability, this study effectively identified crucial risk factors for EOS development in neonates. This approach enables early prediction of EOS risk, thereby facilitating timely","PeriodicalId":23294,"journal":{"name":"Translational pediatrics","volume":"13 11","pages":"1933-1946"},"PeriodicalIF":1.5,"publicationDate":"2024-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11621883/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142802237","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}