The applications of high-flow nasal cannulas in pediatric intensive care units in Taiwan.

BACKGROUND: /purpose: The use of high-flow nasal cannulas (HFNC) in patients admitted to the pediatric intensive care unit (PICU) has gradually increased worldwide; however, details on clinical efficacy remain limited in Taiwan. Therefore, we explored the clinical characteristics and outcomes of pediatric patients using HFNC in the PICU. METHODS: Medical records were retrospectively collected from pediatric patients (aged <18 years) who received HFNC support from December 2021 to January 2023 in the PICU of a medical center. Outcome parameters included treatment failure (defined as increased respiratory support to advanced non-invasive ventilators or intubations), duration of support from HFNC, and changes in clinical parameters after initiating HFNC. RESULTS: A total of 261 episodes of HFNC use were included, with a failure rate of 24.5% and a median support length of 4 days. Multivariable analysis showed that infant age (adjusted odds ratio [aOR]: 2.1, p = 0.02) and accompanying complex chronic disease (aOR: 4.4, p = 0.014) were risk factors for treatment failure and a diagnosis of asthma or bronchiolitis had a lower hazard of treatment failure (aOR: 0.29, p = 0.025) than other diagnoses did. Improvements in clinical parameters, including pulse rate, respiratory rate, SpO2, and CO2 levels, were observed 24 h after the initiation of HFNC. CONCLUSION: The application of HFNC in the PICU in Taiwan is effective but should be performed with care in infants with accompanying complex chronic diseases. In addition to low treatment failure, HFNC utilizations stabilized the clinical parameters of children with asthma/bronchiolitis within one day.

Machine learning models to evaluate mortality in pediatric patients with pneumonia in the intensive care unit.

OBJECTIVES: This study aimed to predict mortality in children with pneumonia who were admitted to the intensive care unit (ICU) to aid decision-making. STUDY DESIGN: Retrospective cohort study conducted at a single tertiary hospital. PATIENTS: This study included children who were admitted to the pediatric ICU at the National Taiwan University Hospital between 2010 and 2019 due to pneumonia. METHODOLOGY: Two prediction models were developed using tree-structured machine learning algorithms. The primary outcomes were ICU mortality and 24-h ICU mortality. A total of 33 features, including demographics, underlying diseases, vital signs, and laboratory data, were collected from the electronic health records. The machine learning models were constructed using the development data set, and performance matrices were computed using the holdout test data set. RESULTS: A total of 1231 ICU admissions of children with pneumonia were included in the final cohort. The area under the receiver operating characteristic curves (AUROCs) of the ICU mortality model and 24-h ICU mortality models was 0.80 (95% confidence interval [CI], 0.69-0.91) and 0.92 (95% CI, 0.86-0.92), respectively. Based on feature importance, the model developed in this study tended to predict increased mortality for the subsequent 24 h if a reduction in the blood pressure, peripheral capillary oxygen saturation (SpO2), or higher partial pressure of carbon dioxide (PCO2) were observed. CONCLUSIONS: This study demonstrated that the machine learning models for predicting ICU mortality and 24-h ICU mortality in children with pneumonia have the potential to support decision-making, especially in resource-limited settings.

The impact of the coronavirus disease 2019 epidemic and national public restrictions on Pediatric Intensive Care Units in Taiwan.

BACKGROUND/PURPOSE: Coronavirus disease 2019 (COVID-19) pandemic challenges pediatric health globally by limited medical accessibility. In response to COVID-19 epidemic in Taiwan, public restrictions were applied and the Level 3 alert was announced from May to July in 2021 for local outbreak. This study aims to analyze patients' clinical features and outcomes in the pediatric intensive care unit (PICU) during the COVID-19 epidemic with the Level 3 alert in Taiwan. METHODS: Medical records were retrospectively collected in patients admitted to the PICU of National Taiwan University Children's Hospital from May to July 2021 (Level 3 alert) and May to July 2019 and 2020 (control periods). Clinical characteristics and outcomes were compared between patients in the period with the Level 3 alert and control periods. RESULTS: During the study period, PICU monthly admissions significantly decreased in the Level 3 alert period and were negatively correlated with monthly newly confirmed COVID-19 cases. Patients admitted during the Level 3 alert were older, had higher disease severity, lower proportion of cardiovascular disease, and higher proportion of hematology-oncology diseases than those in the control group. After adjusting for the above factors, admission during Level 3 alert was an independent factor for higher mortality rate and prolonged length of stay (>14 days) in the PICU. CONCLUSION: During the COVID-19 epidemic with strict public restrictions, critically ill patients admitted to the PICU decreased but had increased disease severity, prolonged length of stay in the PICU, and higher mortality, reflecting the impact of quarantine and limited medical access.

Factors associated with postoperative respiratory conditions and critical outcomes on pediatric liver transplantation: A single-center experience.

BACKGROUND: Orthotopic liver transplantation (OLT) is an established therapeutic option for pediatric end-stage liver disease (PELD). The postoperative respiratory conditions of OLT recipients may be associated with subsequent clinical outcomes including length of stay (LOS) in the pediatric intensive care unit (PICU). This study aimed to characterize the postoperative respiratory conditions, associated factors, and outcomes after pediatric OLT. METHODS: Clinical data of children receiving OLT from July 2014 to July 2020 were retrospectively collected. Postoperative respiratory conditions were defined as time to extubation, significant pleural effusion, and initial postoperative PaO2/FiO2 ratio. Logistic and multiple regressions were applied to analyze the associations among clinical factors, postoperative respiratory conditions, and clinical outcomes. RESULTS: Twenty-two patients with median age of 1.4-year-old (range: 25 days to 12 years old) were analyzed. Mortality within 28 days was 4.5% and median LOS in the PICU was 18 days. Of 22 patients, 11 patients (50.0%) were extubated over 24 hours after surgery, and 8 patients (36.4%) required drainage for pleural effusions. Longer LOS in the PICU were noted in patients extubated over 24 hours (p = 0.008), complicated with significant pleural effusions (p = 0.02) after surgery, and having low initial postoperative PaO2/FiO2 (<300 mmHg) (p = 0.001). Among clinical factors, massive intraoperative blood transfusion (>40 mL/kg) was significantly associated with prolonged intubations, significant pleural effusions, low initial postoperative PaO2/FiO2, and prolonged LOS in the PICU (>14 days). The initial postoperative PaO2/FiO2 significantly depended on age, disease severity (PELD score), and whether the patient received massive intraoperative blood transfusion. CONCLUSION: Pediatric patients of OLT with poor postoperative respiratory conditions including low initial PaO2/FiO2 ratio, extubation over 24 hours or significant pleural effusions have longer LOS in the PICU, and the requirement of massive intraoperative transfusion was a risk factor for both poor postoperative respiratory conditions and prolonged LOS in the PICU.

Haploinsufficiency of RCBTB1 is associated with Coats disease and familial exudative vitreoretinopathy.

Familial exudative vitreoretinopathy (FEVR) belongs to a group of genetically and clinically heterogeneous disorders in retinal vascular development. To date, in approximately 50% of patients with FEVR, pathogenic mutations have been detected in FZD4, LRP5, TSPAN12, NDP and ZNF408. In this study, we identified two heterozygous frameshift mutations in RCBTB1 from three Taiwanese cases through exome sequencing. In patient-derived lymphoblastoid cell lines (LCLs), the protein level of RCBTB1 is approximately half that of unaffected control LCLs, which is indicative of a haploinsufficiency mechanism. By employing transient transfection and reporter assays for the transcriptional activity of ß-catenin, we demonstrated that RCBTB1 participates in the Norrin/FZD4 signaling pathway and that knockdown of RCBTB1 by shRNA significantly reduced nuclear accumulation of ß-catenin under Norrin and Wnt3a treatments. Furthermore, transgenic fli1:EGFP zebrafish with rcbtb1 knockdown exhibited anomalies in intersegmental and intraocular vessels. These results strongly support that reduced RCBTB1 expression may lead to defects in angiogenesis through the Norrin-dependent Wnt pathway, and that RCBTB1 is a putative genetic cause of vitreoretinopathies.

Genomic differences between cultivated soybean, G. max and its wild relative G. soja.

BACKGROUND: Glycine max is an economically important crop and many different varieties of soybean exist around the world. The first draft sequences and gene models of G. max (domesticated soybean) as well as G. soja (wild soybean), both became available in 2010. This opened the door for comprehensive comparative genomics studies between the two varieties. RESULTS: We have further analysed the sequences and identified the 425 genes that are unique to G. max and unavailable in G. soja. We further studied the genes with significant number of non-synonymous SNPs in their upstream regions. 12 genes involved in seed development, 3 in oil and 6 in protein concentration are unique to G. max. A significant number of unique genes are seen to overlap with the QTL regions of the three traits including seed, oil and protein. We have also developed a graphical chromosome visualizer as part of the Soybean Knowledge Base (SoyKB) tools for molecular breeding, which was used in the analysis and visualization of overlapping QTL regions for multiple traits with the deletions and SNPs in G. soja. CONCLUSIONS: The comparisons between genome sequences of G. max and G. soja show significant differences between the genomic compositions of the two. The differences also highlight the phenotypic differences between the two in terms of seed development, oil and protein traits. These significant results have been integrated into the SoyKB resource and are publicly available for users to browse at http://soykb.org/GSoja.