Background: Successful sepsis treatment depends on early diagnosis. We aimed to develop and validate a system to predict sepsis and septic shock in real time using deep learning. Methods: Clinical data were retrospectively collected from electronic medical records (EMRs). Data from 2010 to 2019 were used as development data, and data from 2020 to 2021 were used as validation data. The collected EMRs consisted of eight vital signs, 13 laboratory data points, and three demographic information items. We validated the deep-learning-based sepsis and septic shock early prediction system (DeepSEPS) using the validation datasets and compared our system with other traditional early warning scoring systems, such as the national early warning score, sequential organ failure assessment (SOFA), and quick sequential organ failure assessment. Results: DeepSEPS achieved even higher area under receiver operating characteristic curve (AUROC) values (0.7888 and 0.8494 for sepsis and septic shock, respectively) than SOFA. The prediction performance of traditional scoring systems was enhanced because the early prediction time point was close to the onset time of sepsis; however, the DeepSEPS scoring system consistently outperformed all conventional scoring systems at all time points. Furthermore, at the time of onset of sepsis and septic shock, DeepSEPS showed the highest AUROC (0.9346). Conclusions: The sepsis and septic shock early warning system developed in this study exhibited a performance that is worth considering when predicting sepsis and septic shock compared to other traditional early warning scoring systems. DeepSEPS showed better performance than existing sepsis prediction programs. This novel real-time system that simultaneously predicts sepsis and septic shock requires further validation.

Background: Respiratory support is crucial for newborns with underdeveloped lung. The clinical outcomes of patients depend on the clinician’s ability to recognize the status underlying the presented symptoms and signs. With the increasing number of high-risk infants, artificial intelligence (AI) should be considered as a tool for personalized neonatal care. Continuous monitoring of vital signs is essential in cardiorespiratory care. In this study, we developed deep learning (DL) prediction models for rapid and accurate detection of mechanical ventilation requirements in neonates using electronic health records (EHR). Methods: We utilized data from the neonatal intensive care unit in a single center, collected between March 3, 2012, and March 4, 2022, including 1,394 patient records used for model development, consisting of 505 and 889 patients with and without invasive mechanical ventilation (IMV) support, respectively. The proposed model architecture includes feature embedding using feature-wise fully connected (FC) layers, followed by three bidirectional long short-term memory (LSTM) layers. Results: A mean gestational age (GA) was 36.61 ± 3.25 weeks, and the mean birth weight was 2,734.01 ± 784.98 g. The IMV group had lower GA, birth weight, and longer hospitalization duration than the non-IMV group (P < 0.05). Our proposed model, tested on a dataset from March 4, 2019, to March 4, 2022. The mean AUROC of our proposed model for IMV support prediction performance demonstrated 0.861 (95%CI, 0.853–0.869). It is superior to conventional approaches, such as newborn early warning score systems (NEWS), Random Forest, and eXtreme gradient boosting (XGBoost) with 0.611 (95%CI, 0.600–0.622), 0.837 (95%CI, 0.828–0.845), and 0.0.831 (95%CI, 0.821–0.845), respectively. The highest AUPRC value is shown in the proposed model at 0.327 (95%CI, 0.308–0.347). The proposed model performed more accurate predictions as gestational age decreased. Additionally, the model exhibited the lowest alarm rate while maintaining the same sensitivity level. Conclusion: Deep learning approaches can help accurately standardize the prediction of invasive mechanical ventilation for neonatal patients and facilitate advanced neonatal care. The results of predictive, recall, and alarm performances of the proposed model outperformed the other models.

Neonatal lupus erythematosus (NLE) is a rare disease caused by passively transmitted autoantibodies from the mother. NLE is a multi-organ system disease characterized by cutaneous, cardiac, hematological, hepatobiliary, and neurological manifestations. This study aimed to review the various symptoms and clinical manifestations in young infants with NLE and their mothers. We conducted a retrospective review of medical records of patients with NLE who were both examined and treated at Pusan National University Children’s Hospital between January 2009 and December 2020 and their mothers. Twenty-seven patients with NLE comprising 13 male patients (48.1%) and 14 female patients (51.9%) were included. The most common symptom was rash (40.7%), followed by fever (25.9%), arrhythmia (14.8%), splenomegaly (11.1%), and intrauterine growth retardation (7.4%). Seven patients with fever had various organ system manifestations, including cutaneous (100%), hematological (71.4%), hepatobiliary (57.1%), and central nervous system (CNS; 28.6%) manifestations. Two of the febrile patients had aseptic meningitis. Cutaneous, cardiac, hematological, hepatobiliary, and CNS involvement were noted in 44.4%, 18.5%, 51.9%, 40.7%, and 22.2% of the patients, respectively. Systemic lupus erythematosus (SLE) was the most common maternal disease (14/27, 51.9%). Ten mothers (37.0%) had not been diagnosed with any autoimmune disease until their babies were diagnosed. Among them, three were subsequently diagnosed with SLE, five were diagnosed with the Sjögren’s syndrome, and two of them still had no known diagnosis of any autoimmune disorder. Fever is a common symptom of NLE; thus, when there is no clear focus of fever in infants, NLE needs to be considered, especially in cases with skin rashes.

Background: The etiology of small for gestational age (SGA) is multifactorial and includes maternal/uterine-placental factors, fetal epigenetics, and genetic abnormalities. We evaluated the genetic causes and diagnostic effectiveness of targeted-panel sequencing (TES) or whole-exome sequencing (WES) in SGA infants without a known cause. Methods: A prospective study was conducted on newborn infants born with a birth weight of less than the 10th percentile for gestational age between January 2019 and December 2020 at the Pusan National University Hospital. We excluded infants with known causes of SGA, including maternal causes or major congenital anomalies or infections. SGA infants without a known etiology underwent genetic evaluation, including karyotyping, chromosomal microarray (CMA), and TES/WES. Results: During the study period, 82 SGA infants were born at our hospital. Among them, 61 patients were excluded. A total of 21 patients underwent karyotyping and chromosomal CMA, and aberrations were detected in two patients, including one chromosomal anomaly and one copy number variation. Nineteen patients with normal karyotype and CMA findings underwent TES or WES, which identified three pathogenic or likely pathogenic single-gene mutations, namely LHX3, TLK2, and MED13L. Conclusions: In SGA infants without known risk factors, the prevalence of genetic causes was 22% (5/21). The diagnostic yield of TES or WES in SGA infants with normal karyotype and CMA was 15.7% (3/19). TES or WES was quite helpful in identifying the etiology in SGA infants without a known cause.

We aimed to describe the clinical and microbiological characteristics of enterococcal bacteremia, as well as the effect of Enterococcus resistance against vancomycin on clinical outcomes in Korean children. We retrospectively reviewed the medical records of children diagnosed with enterococci isolated from blood cultures at Pusan National University Children's Hospital between December 2009 and November 2021. In total, 64 patients were enrolled in the study. The median age was 0 years (range 0-15), and 43 (67.2%) patients were male. Enterococcus faecalis (50%) was the most commonly identified bacterial strain. Significant underlying diseases were present in 60 patients (93.8%), and the source of bacteremia was identified in 36 patients (56.3%). Among these, intravascular device was the most common identifiable source. Fifty-six (87.5%) patients had previously received broad-spectrum antibiotics and 54 (84.4%) patients were nosocomial in origin. Twenty-nine (45.3%) strains were resistant to ampicillin, and 16 (25%) strains were resistant to vancomycin. All patients with vancomycin-resistant enterococci (VRE) had underlying disease (P = 0.199), and focus of bacteremia was significantly more frequent in VRE patients (P = 0.014). Of all the patients, after appropriate antibiotic treatment, five (7.8%) patients had recurrent enterococcal bacteremia, and seven (10.9%) patients were diagnosed with bacteremia, defined as other pathogens from blood culture. The 30-day mortality rate was 7.8%. Enterococcal bacteremia in children is usually nosocomial and occurs in children with serious underlying diseases. Because the number of enrolled patients and mortality were small in our study, it is difficult to identify whether the factor that determines prognosis in patients with enterococcal bacteremia is VRE or an underlying disease. Further studies with a large number of patients in a specific group are needed.

Background: nosocomial sepsis remains a significant source of morbidity and mortality in extremely low birth weight (ELBW) infants. Early and accurate diagnosis is very important, but it is difficult due to the similarities in clinical manifestation between the causative microorganisms. We tried to identify the differences between causative microorganisms in clinical and laboratory findings and to help choose antibiotics, when sepsis was suspected in ELBW infants. Methods: a retrospective study was conducted on preterm infants, born at less than 28 weeks of gestation, with a birth weight of less than 1000 g between January 2009 and December 2019. Clinical and laboratory findings of suspected sepsis, after the first 72 h of life, were assessed. We classified them into four groups according to blood culture results (gram positive, gram negative, fungal, and negative culture groups) and compared them. Results: a total of 158 patients were included after using the exclusion criteria, with 45 (29%) in the gram positive group, 35 (22%) in the gram negative group, 27 (17%) in the fungal group, and 51 (32%) in the negative culture group. There were no significant differences in mean gestational age, birth weight, and neonatal morbidities, except for the age of onset, which was earlier in the fungal group than other groups. White blood cell (WBC) counts were the highest in the gram negative group and the lowest in the fungal group. The mean platelet counts were the lowest in the fungal group. C-reactive protein (CRP) levels were the highest in the gram negative group, while glucose was the highest in the fungal group. Conclusions: in conclusion, we showed that there are some differences in laboratory findings, according to causative microorganisms in the nosocomial sepsis of ELBW infants. Increased WBC and CRP were associated with gram negative infection, while decreased platelet and glucose level were associated with fungal infection. These data may be helpful for choosing empirical antibiotics when sepsis is suspected.