Mothers of children with speech, communication, or developmental concerns in Germany may experience complex pathways through a support system marked by multiple institutional interfaces. This qualitative interview study examined how mothers experience the care and support pathway from the first signs of concern to the uptake or coordination of support services, and which conditions they describe as burdensome or relieving. It constitutes the interview component of the mixed-methods project "EGUBE - Eltern gut begleitet?". Eight semi-structured interviews from the ethically approved main study were included. Recruitment was open to parents and caregivers with parental responsibility irrespective of gender; however, the realised interview sample consisted exclusively of mothers. Interviews were conducted predominantly via videoconference, with one interview conducted by telephone; the children were school-aged or adolescents at the time of interview. Prior pilot interviews served only to refine the interview materials and were not included in the present analysis. Data were analysed using qualitative content analysis informed by Kuckartz and Rädiker, combining deductive and inductive elements. The broader health literacy perspective sensitised the analysis to questions of orientation, comprehensibility, navigation, coordination, communication, and organisational demands, while categories and subcategories were refined in close engagement with the interview material. A documented codebook and a case-by-category matrix supported transparent cross-case analysis. The findings reconstruct recurring nodal points along the pathway: first signs of concern, marked by mothers' observations and/or feedback from daycare or kindergarten; diagnostic assessment as an ambivalent turning point, sometimes experienced as "too early" or only weakly clarifying, but at other times as providing orientation when a problem was named or classified; high demands related to navigation and organisation, including appointments, applications, and information-seeking; fragmented interfaces between institutions and professions; and transitions-especially into school-as coordination-intensive phases. Communication was described as central to orientation, ranging from supportive explanations to experiences of devaluing attributions. Resources such as social networks, insurance status, and language- or translation-related accessibility were identified as factors shaping both access and burden. Explicit markers of emotional distress appeared particularly in situations of limited responsiveness, attribution, or uncertainty, whereas relief was associated with experiences of fit, support, and restored capacity to act. These findings point to functional requirements for supportive structures, including reliable orientation and navigation support, clear and structured communication, and improved coordination across institutional interfaces, without presupposing the effectiveness of any specific model.
OBJECTIVE: Procalcitonin (PCT) has moderate accuracy for bacteremia detection but is infrequently used in pediatric emergency departments, partly due to poor sensitivity when applied as a standalone threshold and guideline recommendations against its isolated use. We evaluated whether a machine learning (ML) approach that contextualizes PCT with additional clinical features has the potential to improve detection of gram-negative bacteremic organ dysfunction in children prior to culture results. METHODS: We conducted a retrospective analysis of 431 pediatric encounters across four sites where blood culture and PCT were co-ordered. Blood culture results were classified as gram-negative bloodstream infection (BSI), gram-positive BSI, contaminant, or negative using a priority-based NLP algorithm applied to free-text result strings. The primary outcome was gram-negative BSI with concurrent organ dysfunction ascertained by Phoenix-8 criteria excluding the immunologic domain. We evaluated PCT alone, a four-feature adult sepsis-aligned benchmarking model, and candidate multi-feature combinations identified through systematic univariate screening, using Random Forest (RF) and penalized logistic regression (LASSO) as co-primary algorithms with nested repeated stratified cross-validation and permutation importance analysis. Model discrimination was compared using paired fold-level AUROC testing and calibration was assessed using Platt scaling. RESULTS: Among 431 encounters with a numeric PCT result, 20 (4.6%) met the primary outcome of gram-negative BSI with Phoenix-8 organ dysfunction. PCT alone achieved AUROC 0.762. A four-feature adult sepsis-aligned model achieved AUROC 0.862; respiratory rate and systolic blood pressure contributed negligibly. Systematic screening identified platelet count and creatinine as optimal co-features; the three-feature model (PCT, platelet count, creatinine) achieved AUROC 0.884 (BCa 95% CI 0.871-0.982), consistent across RF and LASSO (0.874), with Brier score 0.038 below the null model (0.044), and positive likelihood ratio 12.56 (number needed to assess of 2.6). CONCLUSIONS: PCT alone showed limited performance. Embedding PCT within a 3-feature ML model improved discrimination over PCT alone (delta AUROC 0.122, p<0.001), with consistent performance across RF and LASSO. Unlike adult models, pediatric prediction relied on renal dysfunction (creatinine) rather than hemodynamics. These findings are hypothesis-generating and require prospective validation in adequately powered cohorts before clinical implementation.
Tom Velez, Maya Dewan, Oluwakemi Badaki-Makun et al.
PURPOSE: This study investigated the multiple components of flexibility in autistic children with and without co-occurring ADHD, and their associations with social functioning. We aimed to determine ADHD co-occurence influence on flexibility profiles and the links between flexibility and social outcomes. METHOD: 124 children aged 6-12 years were drawn from the Geneva Autism Cohort. Flexibility was assessed across three domains: restricted and repetitive behaviors (RRBs; EC2R), daily behavioral flexibility (BRIEF-2), and cognitive flexibility (MTT and IED: CANTAB, and CST: Early Years Toolbox). Socialization skills were measured using the Vineland-II. Group comparisons and correlation analyses were conducted across three groups: typically developing (TD), autism (ASD), and autism with co-occurring ADHD (ASD + ADHD). RESULTS: Both ASD and ASD + ADHD groups showed higher levels of RRBs compared to TD, with no difference between ASD and ASD + ADHD. Daily and cognitive flexibility difficulties were more pronounced in ASD + ADHD, particularly on tasks involving time pressure or implicit learning. BRIEF scores correlated strongly with RRBs in both ASD groups, while links with task-based flexibility were weak. Socialization difficulties were observed in both ASD groups and were strongly associated with caregiver-reported flexibility. Notably, the inverse correlation between RRBs and socialization was significant only in ASD, suggesting co-occuring ADHD may alter the regulatory function of RRBs. CONCLUSION: ADHD co-occurrence contributes to a distinct flexibility profile in autistic children, especially under implicit instructions or timed conditions. Flexibility difficulties were strongly related lower socialization scores, underscoring the need for tailored supports that address both executive and social domains across neurodevelopmental profiles.
Myopia is a global public health concern, especially among children and adolescents. This exploratory study aimed to develop a myopia risk prediction model for Tibetan adolescents using interpretable machine learning, while explicitly acknowledging the methodological limitations of non-cycloplegic refraction and self-reported data. We included 735 Tibetan adolescents. An XGBoost algorithm was used to build a binary classification model, with the SHAP method applied for feature interpretation. Data were split into 80% training and 20% test sets to analyze associations between behavioral, demographic factors, and myopia risk. Myopia was defined using non-cycloplegic autorefraction (uncorrected visual acuity < 5.0 [logMAR 0.0] and spherical equivalent ≤ - 0.50 D). In children aged 9-15 years, non-cycloplegic autorefraction is known to overestimate myopia due to accommodative tone, leading to outcome misclassification. Moreover, behavioral exposures were self-reported by children, introducing recall and desirability bias. Consequently, all findings are purely hypothesis-generating and require validation using cycloplegic refraction and objective exposure measures. Statistical analyses included independent samples t-tests, Mann-Whitney Utests, chi-square tests, and binary logistic regression. Older age was significantly associated with higher myopia risk (OR = 1.200, 95% CI: 1.069-1.347, P = 0.002). Male sex (OR = 0.583, 95% CI:0.427-0.796) and boarding at school (OR = 0.419, 95% CI: 0.250-0.702) were associated with lower myopia prevalence. Behavioral associated factors included irregular breakfast habits (OR = 2.882), reduced frequency of classroom seat rotation, and longer daily after-school study time. Machine learning analysis identified learning behaviors (seat rotation, study time, grade) as the most important predictors, followed by lifestyle factors (breakfast habits, screen time). However, due to the absence of cycloplegia and the cross-sectional design, these associations may be biased and should not be interpreted causally. In this exploratory cross-sectional study with major methodological limitations, several behavioral and demographic factors were associated with myopia status among Tibetan adolescents. The findings do not establish causality and are not suitable for clinical or public health recommendations. Future research must employ cycloplegic refraction, objective exposure monitoring (e.g., wearable light sensors, school records), and longitudinal designs before any reliable conclusions can be drawn.
PURPOSE: Millions of children under five, particularly in low- and middle-income countries, suffer from preventable anemia, making early detection critical for improving public health outcomes. This study proposes an interpretable machine learning framework for early prediction of childhood anemia using structured healthcare data. METHODS: The proposed approach integrates TabNet, XGBoost, and Multi-Layer Perceptron (MLP) within a stacked ensemble architecture, with logistic regression as a meta-learner. Hyperparameter optimization is performed using GridSearchCV and compared with RandomizedSearchCV, Optuna, Genetic Algorithms (GA), Particle Swarm Optimization (PSO), and Ant Colony Optimization (ACO). SHAP provided global feature importance, and LIME explained individual predictions. The system was trained and tested using the Tanzania Demographic and Health Survey (DHS) dataset. RESULTS: Experimental results on the Tanzania Demographic and Health Survey (TDHS 2022) dataset demonstrate that the proposed ensemble significantly outperforms individual models, achieving an accuracy of 98.5% and high precision, recall, and F1-scores across both classes. Among optimization strategies, GridSearchCV and Optuna provide the most consistent and optimal performance. SHAP-based global analysis identifies key predictors such as child age, wealth index, and breastfeeding status, while LIME offers instance-level explanations, enhancing model transparency and clinical interpretability. Comparative evaluation with baseline models and ablation analysis confirms the effectiveness of the stacked architecture and meta-learning strategy. External validation using NFHS (India) data shows close agreement (±1-2%) with observed anemia prevalence trends, indicating cross-regional generalizability. The study also discusses ethical considerations, fairness implications, and deployment strategies for integration into healthcare systems. CONCLUSIONS: The proposed framework is accurate, transparent, and adaptable. It supports early anemia detection and can assist clinical decision-making, especially in under-resourced regions, making it a valuable tool for global public health applications.
Ashima Kukkar, Julius Moinget Loibor, Mehmet Akif Cifci et al.
IMPORTANCE: Avoidant/restrictive food intake disorder (ARFID) is a feeding and eating disorder characterized by extremely restricted dietary variety and/or quantity resulting in significant physical health impairment and psychosocial dysfunction. ARFID frequently co-occurs with neurodevelopmental conditions, yet the extent to which this co-occurrence reflects shared genetic or environmental influences remains largely unknown, as few twin or genetic studies of ARFID have been conducted. OBJECTIVE: To examine the extent to which genetic and environmental influences contribute to the association between a broad ARFID phenotype and neurodevelopmental traits. DESIGN SETTING AND PARTICIPANTS: Population-based twin study using data from the Child and Adolescent Twin Study in Sweden, including 30,374 twins born 1992-2008. MAIN OUTCOMES AND MEASURES: A broad ARFID phenotype was identified using a composite measure derived from parent reports and national health registers between ages 6 and 12 years. Parents completed measures of neurodevelopmental traits at age 9 or 12 years, including autism (subdomains: social communication problems and restricted/repetitive behaviors), attention-deficit/hyperactivity disorder (ADHD, subdomains: inattention and impulsivity/hyperactivity), tic disorders, learning disorders, oppositional defiant disorder, conduct disorder, obsessive-compulsive disorder (OCD), sensory perception problems, and sleep problems. Phenotypic associations were estimated using polyserial correlations. Bivariate twin models decomposed variance and covariance into genetic and environmental components. RESULTS: Phenotypic correlations with the broad ARFID phenotype ranged from 0.18 (95% CI: 0.15-0.21) for OCD to 0.36 (95% CI: 0.33-0.38) for autism. Broad genetic correlations (rH; additive plus dominant genetic influences) ranged from 0.27 (95% CI: 0.21-0.33) for conduct disorder to 0.52 (95% CI: 0.44-0.60) for autism-restricted/repetitive behaviors. Genetic factors explained 77% to 95% of all phenotypic correlations. Non-shared environmental correlations were minimal to small, with the largest observed for autism (0.17; 95% CI: 0.08-0.26). CONCLUSIONS AND RELEVANCE: The broad ARFID phenotype shares substantial genetic influences with a number of neurodevelopmental traits. These findings suggest that the frequent co-occurrence of ARFID with neurodevelopmental traits largely reflects shared genetic influences rather than overlapping environmental influences, supporting the conceptualization of ARFID within a broader neurodevelopmental framework.
INTRODUCTION: Autistic children often face barriers to accessing timely and accurate hearing evaluations. Yet determining a child's hearing status is important for maximizing developmental outcomes when there are communication concerns. METHODS: This study evaluated the effectiveness of a standardized hearing assessment protocol for children aged 18 months to 8 years, with and without an autism diagnosis in their medical record. The protocol aimed to determine whether sufficient hearing data were available for children to proceed with speech-language or autism evaluations ("speech criteria") and/or be discharged from audiology care ("discharge criteria"). RESULTS: Retrospective analyses of 608 children's medical records from a single institution revealed that 93.9% of children met speech criteria and 56.6% met discharge criteria. Relative to children in the comparison group, autistic children were significantly less likely to meet speech or discharge criteria (all p-values < .05), required nearly twice as many encounters to do so, and relied more heavily on physiological measures (e.g., distortion product otoacoustic emissions and sedated auditory brainstem responses). Despite these differences, many autistic children successfully met criteria by the second encounter, demonstrating the protocol's overall effectiveness. DISCUSSION: Findings underscore the need for protocol refinement by incorporating tailored supports that address sensory sensitivities, communication differences, and learning profiles of autistic children. Future research should investigate strategies to optimize behavioral testing and design fast-track pathways based on physiological assessments. The protocol presented here has the potential to advance patient care and improve timely and accurate determination of hearing status in children undergoing a speech-language and/or autism evaluation.
Eriel J Confer, Alyssa M Fosnight, Lindsey R Wheeler et al.
OBJECTIVE: To evaluate the clinical effectiveness of the integrated labor-delivery-recovery-postpartum (LDRP) model among low-risk parturients at a tertiary hospital in China. METHODS: A total of 542 low-risk, full-term, singleton parturients voluntarily selected either the LDRP or traditional delivery mode. Outcomes between groups were compared using propensity score matching, multivariate regression, Bayesian analysis, and machine learning. RESULTS: The LDRP group had shorter labor duration (7.3 h vs. 9.8 h), and significantly lower cesarean section rates (1.2% vs. 4.6%) and neonatal transfer rates (5.4% vs. 16.0%; all P < 0.05). Postpartum hemorrhage rates were also lower in the LDRP group (1.2% vs. 6.0%), but this difference was not statistically significant after multivariate adjustment (P = 0.235). Women in the LDRP group more frequently reported feeling "calm and relaxed" (17.7% vs. 0.4%) but had higher latent-phase pain scores (2.0 vs. 1.0). Bayesian analysis confirmed that LDRP independently reduced neonatal transfer risk, though its protective effect against postpartum hemorrhage was insufficiently robust. CONCLUSION: The LDRP model effectively shortens labor, reduces cesarean section and neonatal transfer rates, and improves maternal psychological experience. Its independent protective effect on postpartum hemorrhage requires further validation.
UNLABELLED: Childhood obesity is rising globally. Yet, few studies have examined the microbiome and proteome in early childhood in relation to this outcome, and most are cross-sectional by design. Early-life factors in the ABIS birth cohort (n = 16,683) were associated with obesity up to age 26 (mean follow-up 25.3 years, range 23.7-26.5 years): psychosocial stressors, smoking, infections, and diet in the first year. We assessed biomarkers, including cord blood metabolome (n = 290) and proteome (n = 358), by liquid chromatography, mass spectrometry, and Olink. Gut microbial composition at age one (n = 1,743) was assessed using stool samples and 16S rRNA sequencing. In this prospective longitudinal cohort study, significant differences were found in infants with future obesity, including elevated angiopoietin-like 4 (ANGPTL4), follistatin, and hepatocyte growth factor (independently of maternal weight) and reduced isocaproic acid, tryptophan, and oleic acid, with prenatal mediation. Akkermansia, asaccharolytic bacteria (Phascolarctobacterium and Senegalimassiliensia), and equol-producers (Adlercreutzia and Slackia) were depleted. Machine learning models selecting 40 most predictive features showed long-term prediction from birth proteomics and bacterial taxa at age one (area under the curve [AUC] = 0.83 ± .05, n = 1,877) and additional metrics, for example, parental and child body mass index in the first 8 years (AUC = 0.89 ± .02, n = 1,877), suggesting durable biological encoding. Proteomic markers across folds included fibroblast growth factor 19, ANGPTL4, sulfotransferase family 2A member 1, and interleukin 20. These findings suggest clinically relevant biomarkers indicating early-life regulation of bile acid metabolism, lipid storage vs. oxidation, and immune-metabolic signaling and pathways to prospectively prevent childhood- and adult-onset obesity across a 26-year predictive gap. IMPORTANCE: Understanding the origins of obesity is critical for developing preventive strategies, and early life represents a particularly sensitive window. This study leverages a large, general-population cohort with prospectively collected data, including parental body mass index (BMI), cord blood proteomics, and the gut microbiome at age one, linked to obesity outcomes over 26 years. Using integrated machine learning models, we show that in addition to parental BMI, specific proteomic and microbial markers present in infancy can predict long-term obesity risk, highlighting the role of early metabolic programming. Several key markers point to bile acid signaling as a mechanism connecting early microbiome development with fat accumulation and insulin regulation. By identifying these early-life predictors long before obesity manifests, these results provide new insights into intergenerational risk and suggest measurable targets for preventing obesity and related metabolic disorders from the earliest stages of life.
Angelica P Ahrens, Raquel Dias, Tuulia Hyötyläinen et al.
BACKGROUND: The number of 1-year Accreditation Council for Graduate Medical Education (ACGME) fellowships continues to grow. The ACGME recommends a holistic curriculum with nonclinical areas, inclusive of educational sessions. Given the competing demands between clinical skill development, educational pursuits, and work-hour restrictions, we propose an andragogic curriculum using pediatric anesthesiology as the model fellowship. OBJECTIVE: The primary objective was to improve fellows' perceptions of their educational experience during their fellowship year after implementing an andragogic holistic curriculum. Secondary objectives assessed improvements in diversity, equity, and inclusion (DEI) training and resources. METHODS: This was a single-center educational improvement project completed at Lucile Packard Children's Hospital Stanford. Data were collected between 2014 and 2024. The new curriculum was introduced in 2021-2022 and involved 12 different teaching modalities rooted in andragogic principles. A statistical process control p-chart was used to analyze the primary outcome based on the ACGME annual program evaluation. Outcomes were analyzed using censored regression modeling or a t test, depending on the presence of ceiling effects. RESULTS: From 2014 to 2024, 58 of 60 pediatric anesthesiology fellows completed the ACGME survey. A break in the statistical process control p-chart for educational content scores occurred during 2021-2022, when the new curriculum was introduced. The mean difference was 0.89 (P<.001). Scores in DEI improved (mean difference 0.52; P=.03), and no difference was noted in resources (mean difference -0.13; P=.98). CONCLUSIONS: Introduction of an andragogic curriculum into a pediatric anesthesiology fellowship program was associated with more favorable perceptions of educational content and DEI training.
