Objective: Pediatric tuberculosis (TB) is one of the top ten causes of death in children. Our study was to analyze influencing factors of multidrug-resistant tuberculosis (MDR-TB) and validation of whole-genome sequencing (WGS) used in children with drug-resistant TB (DR-TB). Methods: All Mycobacterium tuberculosis (Mtb) strains were isolated from patients aged below 18 years old of Children’s Hospital of Chongqing Medical University, China. A total of 208 Mtb isolates were tested for eight anti-TB drugs with phenotypic drug susceptibility test (DST) and for genetic prediction of the susceptible profile with WGS. The patients corresponding to each strain were grouped according to drug resistance and genotype. Influencing factors of MDR-TB and DR-TB were analyzed. Results: According to the phenotypic DST and WGS, 82.2% of Mtb strains were susceptible to all eight drugs, and 6.3% were MDR-TB. Using the phenotypic DSTs as the gold standard, the kappa value of WGS to predict isoniazid, rifampin, ethambutol, rifapentine, prothionamide, levofloxacin, moxifloxacin and amikacin was 0.84, 0.89, 0.59, 0.86, 0.89, 0.82, 0.88 and 1.00, respectively. There was significant difference in the distribution of severe TB, diagnosis, treatment and outcome between MDR and drug-susceptible group (P< 0.05). The distribution of severe TB and treatment between DR and drug-susceptible group was statistically different (P< 0.05). The results of binary logistic regression showed that Calmette–Guérin bacillus (BCG) vaccine is the protective factor for MDR-TB (OR=0.19), and MDR-TB is the risk factor for PTB and EPTB (OR=17.98). Conclusion: The BCG vaccine is a protective factor for MDR-TB, and MDR-TB might not be confined to pulmonary infection, spreading to extrapulmonary organs in children. MDR-TB had more severe cases and a lower recovery rate than drug-susceptible TB. WGS could provide an accurate prediction of drug susceptibility test results for anti-TB drugs, which are needed for the diagnosis and precise treatment of TB in children.

Human adenovirus type 14 variant p1 (HAdV-14p1) has been associated with severe respiratory infections worldwide. However, no cases of HAdV-14p1 infection have been reported in China since 2019. HAdV-14 was unexpectedly identified in 2024 in a pediatric patient hospitalized with bronchopneumonia in Chongqing, China. This study aims to elucidate the genetic characteristics of this strain and determine its phylogenetic relationship with previously circulating domestic and international strains. Whole-genome sequencing was performed on the HAdV-14 strain Chongqing2024-115, followed by comprehensive phylogenetic and genetic variation analyses using all publicly available global HAdV-14 genome sequences from the GenBank database. Genome-based phylogenetic analysis classified global HAdV-14 strains into two well-supported clades: the HAdV-14 prototype clade and the variant HAdV-14p1 clade. The latter was further subdivided into three distinct subclades (I–III). Chongqing2024-115 clustered closely with earlier HAdV-14p1 subclade III and isolates from Beijing (2012) and Gansu (2013), showing high sequence identity (99.7%–99.9%). Compared with the genome of the HAdV-14 prototype strain (de Wit, AY803294), all three HAdV-14p1 subclades shared 99 specific nucleotide variations, including a six-nucleotide deletion in the knob domain of the fiber gene (nt751–756). The identification of subclade-specific variations and mutations unique to strain Chongqing2024-115 indicates ongoing microevolution within the HAdV-14p1 clade. Our results confirm that the contemporary HAdV-14p1 strain shares a common genetic ancestry with earlier Chinese subclade III strains. Considering the association between HAdV-14p1 and severe disease, sustained surveillance is essential to understand better its prevalence, evolution, and impact on public health in China.

Background: Long-term combination anti-tuberculosis regimens may lead to various types and severities of adverse drug reactions, potentially resulting in treatment discontinuation, interruption, or prolonged therapy. Aim: Considering the lengthy duration of anti-tuberculosis combination regimens, it is crucial to monitor adverse drug reactions occurring during such treatment. This is the first hospital-based study conducted in western China to report adverse drug reactions and assess their prevalence and impact among children with tuberculosis receiving first-line anti-tuberculosis therapy. Methods: A retrospective study was conducted on registered tuberculosis cases at the Children's Hospital of Chongqing Medical University from 2014 to 2023. Data on demographic characteristics, treatment regimens, and adverse drug reactions were collected and evaluated. Statistical methods were used to analyze factors associated with adverse drug reactions. Results: Among 546 patients receiving anti-tuberculosis therapy, adverse drug reactions were reported in 476 (87.18%). Hyperuricemia was the most common adverse drug reaction (44.26%), followed by elevated liver enzymes without jaundice and elevated bile acids (10.38%) and leukopenia (8.31%). Most adverse drug reactions occurred within the first month of treatment (51.48%), whereas 33.44% persisted for >6 months. Pyrazinamide was the most frequently suspected causative drug (7.47%). Univariate analysis showed significantly higher adverse drug reaction rates among children aged 11–18 years, those with severe tuberculosis, those receiving the isoniazid + rifampicin + pyrazinamide + ethambutol + linezolid regimen, those on a recommended dose of rifampicin, and those receiving a high dose of pyrazinamide ( p < 0.05). Logistic regression identified age (odds ratio = 1.640, 95% confidence interval: 1.127–2.388, p = 0.010) and the isoniazid + rifampicin + pyrazinamide + ethambutol + levofloxacin regimen (odds ratio = 0.210, 95% confidence interval: 0.048–0.932, p = 0.040) as factors that reduced the odds of adverse drug reactions. Conclusions: A high occurrence rate of adverse drug reactions was observed in children undergoing anti-tuberculosis treatment. Timely recognition and management of adverse drug reactions is essential in this patient population.

Background: Significant hepatic injury (SHI) in children with chronic hepatitis B virus (HBV) infection requires timely identification, yet liver biopsy is invasive and not routinely feasible. This study aimed to develop a non-invasive predictive model for SHI. Methods: This retrospective analysis included treatment-naïve children with chronic HBV infection undergoing liver biopsy. Participants were randomly split into 70% training and 30% independent testing cohorts prior to preprocessing. Missing data imputation and least absolute shrinkage and selection operator (LASSO)-based feature selection were performed, and nine machine learning (ML) models were developed and optimized using nested cross-validation (CV). The optimal logistic regression (LR) model was used to construct and validate a predictive nomogram. Results: A total of 246 eligible children were included, of whom 147 (59.8%) had SHI. LASSO regression identified five predictors for model development. Among the evaluated ML models, LR showed stable performance, achieving an area under the receiver operating characteristic curve (AUC) of 0.799 (95% CI: 0.740– 0.858) in the training folds, 0.781 (95% CI: 0.585– 0.972) in the validation folds, and 0.707 (95% CI: 0.562– 0.851) in the independent testing cohort. The nomogram showed moderate calibration and better clinical utility compared with alanine aminotransferase (ALT) alone across a broad range of threshold probabilities. Conclusion: We developed a non-invasive predictive model based on routine clinical data to detect SHI in treatment-naïve children with chronic HBV infection, which exhibited moderate predictive power and marginal improvement over ALT alone. An infographic is divided into four parts (A–D): Clinical Data Collection, Model Selection, Model Presentation, and Clinical Application. A: Clinical data collection includes demographic and laboratory data. B: Model construction and validation includes variable selection using LASSO and VIF, with variables such as ALT and HBeAb. Model selection features classifiers such as LR and RF, highlighting LR. C: LR model construction and evaluation includes 147 children with SHI and 99 without SHI, followed by nomogram construction. D: Model application includes model evaluation and clinical decision-making.An infographic workflow of clinical prediction modeling for SHI in children with chronic HBV infection.