JAMA Pediatrics

ImportanceWhole genome sequencing (WGS) is increasingly used to diagnose severely ill children, yet the long-term impact of a genetic diagnosis on healthcare utilization and resource allocation remains poorly understood. ObjectiveTo determine the influence of a genetic diagnosis via WGS on long-term healthcare utilization metrics in severely ill children. DesignA retrospective cohort study using data from the Next Generation Children study (2016-2020) with record linkage and analysis of primary care records conducted between 2022 and 2024. SettingA multicenter study involving primary care and hospital records linked via the UK National Health Research Institute (NIHR) Rare Disease Bioresource, Cambridge, UK. ParticipantsA referred sample of 270 severely ill children who underwent WGS. Exposure(s)Receipt of a genetic diagnosis (87/270; 32%) compared to those who remained undiagnosed (183/270; 68%) following WGS. Main Outcome(s) and Measure(s)Comparison of 36 healthcare utilization parameters, including hospitalizations, primary care prescriptions, and diagnostic tests. ResultsAmong the 270 children analyzed, those receiving a genetic diagnosis (n=87) exhibited significantly higher overall healthcare utilization compared to undiagnosed peers (n=183). This included increased hospital admissions and outpatient visits, particularly for neurodevelopmental and seizure-related conditions. Diagnosed children received a higher volume of neurological, gastrointestinal, and nutritional prescriptions. The most pronounced differences in utilization were observed in children initially diagnosed in neonatal (NICU) or pediatric (PICU) intensive care settings. While genetic diagnosis was not associated with reduced healthcare costs during the study period, it was linked to more targeted, condition-specific medical care. Conclusions and RelevanceWGS diagnosis facilitates the integration of specialist care and the alignment of healthcare resources with the specific needs of children with complex disorders. These findings suggest that while costs may not decrease immediately, a diagnosis enables more precise and targeted clinical management. Key PointsO_ST_ABSQuestionC_ST_ABSDoes a genetic diagnosis through whole genome sequencing influence long-term healthcare utilization in severely ill children? FindingsIn this cohort study of 270 children, those who received a genetic diagnosis demonstrated significantly greater overall healthcare utilization, including more hospitalizations and targeted prescriptions, compared with undiagnosed children. MeaningA genetic diagnosis facilitates the integration of specialized, condition-specific care, helping to align healthcare resources with the individual needs of children with complex disorders.

BackgroundPediatric long COVID is associated with substantial symptom burden, yet evidence-based pharmacologic treatments remain limited. Low-dose naltrexone (LDN) has been proposed as a potential symptomatic therapy, but data in pediatric populations is lacking. MethodsWe conducted a retrospective analysis of pediatric and young adult patients ([≤]25 years) with a clinical diagnosis of long COVID who were prescribed LDN between July 2020 and July 2025 at three multidisciplinary pediatric long COVID programs in the United States. Deidentified clinical data were extracted from medical records. Outcomes included symptom prevalence, dosing practices, treatment continuation or discontinuation, adverse effects, and available patient-reported quality-of-life measures (PedsQL and PROMIS(R)). FindingsThe study included 62 patients (mean age, 15.6 years [range, 8-23]; 53.2% male and 46.8% female). Fatigue was nearly universal (98.4%), followed by headaches (87.1%), brain fog (74.2%), dizziness/lightheadedness (67.7%), anxiety (66.1%), and post-exertional malaise (56.5%). LDN-treated patients demonstrated a higher prevalence of neurocognitive and autonomic symptoms, compared to general clinic cohorts. Most patients (71.0%) reported no adverse effects; the most common were vivid dreams (9.7%) and insomnia (9.7%). At follow-up, 66.1% of patients remained on LDN. Medication discontinuation was attributed to perceived lack of benefit (43.8%) or side effects (25.0%). Baseline quality-of-life measures at initiation showed marked impairment: PedsQL Physical Health (M=38.0, SD=20.9) and Multidimensional Fatigue (M=35.7, SD=15.8) scores were low. PROMIS scores indicated reduced physical functioning (M=36.8, SD=8.7) and cognitive functioning (M=40.8, SD=7.6), with elevated fatigue (M=68.0, SD=10.4) and pain interference (M=58.6, SD=8.2) relative to population norms. The study was not designed to assess efficacy. InterpretationLDN was primarily prescribed to patients with prominent fatigue, neurocognitive symptoms, and autonomic dysfunction, and was generally well tolerated. These findings provide descriptive evidence of real-world prescribing practices and support the need for clinical trials to systematically evaluate LDNs efficacy in pediatric long COVID.

PurposeGenetic diseases often present and are first diagnosed in the neonatal intensive care unit (NICU). Accurate identification of neonates with genetic diagnoses (GDs) in electronic health records (EHR) would enable a more complete understanding of their phenotypic spectrum, advancing care and personalized medicine. Prior research has used International Classification of Diseases (ICD) billing codes as proxies for GDs, though their accuracy for detecting confirmed GDs is uncertain. We evaluate the ICD codes for neonates with confirmed GDs and compare ICD billing code patterns between neonates with and without GD in two independent NICU cohorts. MethodsRetrospective analysis of patients admitted to the Boston Childrens Hospital (BCH) level IV NICU (1,344 neonates) and Nationwide Childrens Hospital (NCH)s neonatal network (33,315 neonates, mixed Level III/IV). For both cohorts, GDs captured by phecodes, aggregates of ICD codes, were compared with confirmed GDs. Two separate phenome-wide association studies (PheWAS) compared phecode patterns between neonates with GDs and those without, adjusting for sex, age at admission, gestational age, and NICU length of stay. ResultsGenetic phecodes were able to correctly identify 43.5% of neonates that received a GD in the BCH or NCH NICUs. Among 719 individuals with two or more genetic phecodes at BCH or NCH, 566 (78.72%) had a true GD. The BCH PheWAS analysis revealed a statistically significant positive association with atrioventricular septal defects and a negative association with bronchopulmonary dysplasia. The NCH pheWAS revealed 179 significantly associated phecodes, including many congenital anomalies. ConclusionThe use of ICD codes to identify NICU infants with GDs is neither sensitive nor accurate, though phecode analysis demonstrated stronger accuracy than sensitivity. Our data highlight clinical features of NICU infants more commonly seen in those that receive a GD (congenital heart defects) and those that are not (BPD). Our results can help to better predict and identify NICU neonates that receive a GD.

BackgroundOrganization, time management, and planning (OTMP) difficulties are associated with academic underachievement. OTMP skills training programs are effective in reducing OTMP deficits and improving academic performance. A randomized controlled trial of Homework, Organization, and Planning Skills (HOPS) for students ages 11-14 (1) found it to be effective with medium to large effects. In that study, HOPS was provided by counselors employed by the research team. This study is a replication examining HOPS under more authentic conditions when providers are employed by schools serving enrolled students. The primary aim is to evaluate HOPS offered by school providers in relation to treatment-as-usual/waitlist (TAU/WL). To respond to limited school resources post-COVID-19, HOPS is also provided by research team members, creating the opportunity to replicate the findings from the prior trial (1) and explore differential effectiveness when HOPS is implemented by school vs. research providers. MethodsStudents in about 30 schools serving students ages 11-14 will be enrolled. Schools are randomly assigned to HOPS vs. TAU/WL on a 2:1 ratio. Students assigned to HOPS schools are randomly assigned to a school vs. research provider on a 1:1 basis. Providers receive two hours of training and additional assistance on request. Child outcomes related to OTMP skills, homework, and academic performance are assessed at post-treatment, 6-month (from baseline) follow-up, and 12-month follow-up. HOPS sessions are video recorded for fidelity coding. Potential effect modifiers include student ADHD, oppositional defiant, and internalizing symptoms, and family socioeconomic level. Analyses will use mixed effects modeling. The goal of the study is to enroll 135 participants, yielding a minimal detectable effect size of 0.50, within the expected range based on prior research. DiscussionThe study is unique in examining intervention implementation and effectiveness when intervention is provided under authentic practice conditions. Trial RegistrationThis study was registered with clinicaltrials.gov (NCT04465708).

BackgroundCaregiver skills training programmes are well-researched in the fields of autism and intellectual disability, but children with motor disorders such as cerebral palsy remain underrepresented despite their high prevalence. These caregivers face unique challenges, and group programmes may provide family-centred care through information provision, problem-solving and peer support. MethodsSystematic searches of five databases (CINAHL, Medline, Embase, PsychINFO and ERIC) were conducted for interventional studies of group programmes aiming to improve the skills, confidence and wellbeing of caregivers of children with neurodisability focusing on motor disorders. Data were extracted on study and intervention characteristics and outcomes. Risk of bias was assessed, effect sizes calculated, and results summarised descriptively using forest plots. ResultsOf 6093 studies identified, 21 studies met inclusion criteria (nine randomised-controlled trials, two quasi-experimental and ten pre-post designs). Most reported on programmes developed in resource-constrained settings and addressed caregiver skills, coping strategies, or health-promoting behaviours. Outcomes were grouped according to caregiver wellbeing, caregiver skills and confidence, and social support and family functioning. Child outcomes were reported separately. Most caregiver outcomes showed positive effects, though most studies had high risk of bias due to self-reported outcomes and lack of blinding of intervention allocation and outcome measurement. DiscussionGroup-based training programmes show promise for improving caregiver skills and wellbeing. Clinicians and stakeholders in high-income countries may learn from these innovations in low-resource settings. Future research should strengthen protocol reporting, address attrition, control for confounding factors, and establish a core set of caregiver-reported outcomes to better capture programme impact. Systematic review registrationPROSPERO registration CRD42024595002

PurposeWhile genomic testing is integral to pediatric inborn errors of immunity (IEI) care, few studies have examined strategies to support its optimal delivery. This study aimed to characterize a pediatric IEI cohort and assess the impact of implementing a mainstream model-of-care (MoC). Materials/MethodsComprehensive chart audit was conducted for patients ([&le;]18y) who received IEI genomic testing in Queensland, Australia, from 2017-2025. Descriptive analyses captured demographic and clinical characteristics, genomic testing and results, and management outcomes. Inferential analyses assessed changes in genomic practices pre-MoC (<2021) and post-MoC ([&ge;]2021). Results322 patients met eligibility criteria (n=481 genomic test). Diagnostic yield (27.6%) varied by testing indication, with the highest rate among phagocytic defects (n=4/4;100%) and severe combined immunodeficiency (n=8/10;80%). Very-early-onset inflammatory bowel disease had the lowest diagnostic yield (n=3/68;4.4%), prompting changes to testing criteria. Molecular diagnosis resulted in management changes for 90.5% patients. Genomic testing was widely used pre-MoC (n=251 genomic tests). All outcomes significantly improved pre-and post-MoC (p<0.05): duplicate testing decreased (13.9% to 0%); variants of uncertain significance reduced (37.7% to 7.1%); informed consent documentation increased (70.5% to 88.4%); and diagnostic yield increased (16.2% to 27.4%). ConclusionTargeted interventions are needed to support delivery of genomic testing and strengthen service effectiveness.

PurposeCaregivers of children with complex neurodisability frequently experience high caregiving demands, social isolation, unmet support needs, and reduced wellbeing. This paper explores caregivers perceptions of the impact of "Encompass", a ten-modular, community-based group support programme for caregivers of children under five with complex neurodisability, co-facilitated by an expert parent. Materials and methodsThis study formed part of a pilot and feasibility study conducted in two socially disadvantaged, ethnically diverse urban areas in the United Kingdom. Outcome measures were collected pre-intervention, post-intervention and at three-month follow-up to explore caregiver wellbeing, empowerment, activation, and quality of life. Semi-structured qualitative interviews were conducted within three months of programme completion. Interview data were analysed using deductive coding informed by the "Encompass" programme theory alongside inductive analysis to explore mechanisms and unanticipated benefits. Results and conclusionsSeven participating caregivers described improved wellbeing, increased confidence in caring for their child, navigating services, advocating for their family and engaging in the community. Peer support, shared learning and expert parent facilitation were key identified mechanisms of impact. Data from outcome measures showed patterns of improvement post-intervention, with less consistent eYects at follow-up. Findings confirmed the key change mechanisms, informing future iterations and other caregiver group programmes. Trial RegistrationClinicalTrials.gov Identifier: NCT06310681

BackgroundHome disciplinary practices shape childrens health and development. Yet, comprehensive, up-to-date global evidence on their levels, trends, and socioeconomic and regional inequalities remains limited. This study provides the first global prevalence estimates of both violent and non-violent forms of discipline, examining regional disparities, variations by child and family characteristics, and changes over time. MethodsWe drew from 176 nationally-representative Multiple Indicator Cluster Surveys and Demographic and Health Surveys, collected between 2005 and 2023 across 83 low- and middle-income and 5 high-income countries (N= 1,544,000 1-14y-olds). We estimated weighted prevalence estimates for all types of discipline (exclusively or only non-violent, physical and severe physical punishment, emotional violence, exclusively or only physical punishment, exclusively or only emotional violence, both physical and emotional violence). Disparities by child age, sex, residence, maternal education, household wealth, and world regions were computed. We also assessed changes over time for countries with multiple surveys. ResultsOnly 19.1% of children experienced exclusively non-violent discipline; 55.0% and 12.7% physical and severe physical punishment; and 64.0% emotional violence. Violent discipline was highest among 6-9y-olds, in Sub-Saharan Africa, and in poorer households. Sex differences were more limited. Use of only non-violent discipline slightly increased in 26 countries, while physical and emotional violence decreased in 33 and 31 countries, respectively. Yet, in some countries, violent discipline increased over time. ConclusionsDespite policy efforts to increase its use, exclusive non-violent discipline remains low, and violent methods are widespread. Targeted and context-specific interventions for specific age groups and poorer households curb violence exposure at home.

UK-based quantitative research on the health and education outcomes of Unaccompanied Asylum-Seeking Children (UASC) remains limited, especially at national level. Linked administrative data provide an unprecedented opportunity to study these outcomes among UASC. This paper lays a foundation for further research, particularly examining the influence of socio-demographic, legal and environmental factors on UASCs health and educational outcomes. We described the UASC population with a first recorded episode of local authority care between 1st April 2005 and 31st March 2021 in ECHILD, which gathers national records for England, by age, gender, ethnicity, region, and placement type. We calculated linkage rates between the social care and educational dataset, estimating how many UASC were recorded as being enrolled in state-funded schools. We also assessed how many of those linked to the school dataset was linked to National Health Service (NHS) datasets. Finally, we explored how linkage rates between social care, education, and NHS datasets vary by socio-demographic factors and placement type. There were 37,170 UASC recorded in the ECHILD of which 32,570 (88%) were male and 24,290 (65%) aged 16 - 17 years. We found 7,740 (21%) UASC recorded as being enrolled in state funded schools, of whom 6,690 (88%) were also linked to NHS data. The linkage rate for UASC in the social care to health datasets was therefore 19%. Of those 16-17 years at entry in social care, 4% (1,060/24,290) were recorded as enrolled in school compared to 50% (6,390/12,880) under 16 years. Linkage to the school, and subsequently to the NHS dataset, wholly depends on enrolled state-funded education, excluding College and Sixth-form education. Despite this limitation, we characterised a national cohort of 6,890 UASC in England whose social care, education, and health outcomes can be examined.

BackgroundParental genetics matters for childrens behavioural difficulties, but the extent to which this is due to direct genetic transmission versus environmentally mediated indirect genetic effects remains unclear. MethodsWe studied eight European birth cohorts with over 33,000 family-based trio samples. We analysed polygenic scores (PGSs) for 13 mental health and neurodevelopmental conditions and their composite indices (PC1 and mean) representing general neuropsychiatric liabilities, as well as educational attainment (EA) and alcohol and cigarette use, from children (PGSc), mothers (PGSm), and fathers. Child internalising, externalising, and total difficulties reported by mothers and/or fathers were examined at preschool and school ages. We then conducted multivariate meta-analyses to combine cohort-level results. FindingsWe observed several direct genetic effects on externalising difficulties, while indirect genetic influences were mainly identified for internalising difficulties. Specifically, child PGSs for attention-deficit/hyperactivity disorder (ADHD) and EA predicted higher and lower levels, respectively, of child externalising and total difficulties (all pFDR<0{middle dot}001; for school-aged externalising difficulties, PGSc-ADHD: {beta}=0{middle dot}121 [95% CI 0{middle dot}091 to 0{middle dot}151], pFDR<0{middle dot}0001; PGSc-EA: {beta}=-0{middle dot}095 [95% CI -0{middle dot}127 to -0{middle dot}063], pFDR<0{middle dot}0001), whereas maternal PGSs for major depressive disorder (MDD) and general neuropsychiatric liabilities were associated with internalising and total difficulties across parental raters and child ages (all pFDR<0{middle dot}05; for school-aged internalising difficulties, PGSm-MDD: {beta}=0{middle dot}049 [95% CI 0{middle dot}017 to 0{middle dot}081], pFDR=0{middle dot}016; PGSm-PC1: {beta}=0{middle dot}056 [95% CI 0{middle dot}022 to 0{middle dot}091], pFDR=0{middle dot}011). No statistically significant effects from paternal PGSs were identified. InterpretationIn this multi-cohort study, findings across multiple traits, raters, and ages supported several direct genetic effects of ADHD and EA on child externalising difficulties and indirect genetic effects on internalising difficulties, especially maternal depression and general neuropsychiatric liabilities. These suggest that child internalising difficulties are not solely driven by direct genetic transmission. More comprehensive research is needed to better understand the mechanisms involved, and ultimately how to ameliorate child behavioural difficulties. FundingEU, ERC, RCN, RCF, UKRI, SERI, DFG Research in contextO_ST_ABSEvidence before this studyC_ST_ABSIndirect genetic effects (IGEs) refer to the influence of parental genotypes on offspring outcomes beyond direct genetic effects (DGEs), for example via environmental pathways. While IGEs on offspring cognitive traits are well-established for educational attainment, evidence for IGEs of parental liabilities to mental health and neurodevelopmental conditions remains limited. To assess the current state of evidence, we conducted a systematic search of published studies applying trio-based polygenic score (PGS) designs to child and adolescent mental health outcomes. We identified 141 primary studies in MEDLINE, Embase, PsycInfo, and Web of Science, by 6 March 2025, after removing duplicates; following screening, 12 studies met inclusion criteria (see supplement for a full description including results). Ten out of the 12 studies focused on externalising outcomes, with little or inconsistent support for IGEs. When observed, IGEs were mainly driven by maternal liabilities to autism, educational attainment, and cognitive performance on child outcomes. The current evidence was too limited and heterogeneous to synthesize findings quantitatively, therefore a qualitative synthesis was conducted. Many studies were statistically underpowered, and the observed IGEs were in all cases sample-specific. There were no published multi-cohort studies. Added value of this studyWe integrated information across over 33,000 mother-father-child trios from eight European cohorts, investigating 18 PGSs from parents and children, using maternal and paternal ratings of offsprings internalising, externalising, and total difficulties as outcomes at both preschool and school age. We mainly observed DGEs on externalising difficulties, consistent with previous studies. Some evidence of IGEs was found for internalising and total difficulties. IGEs were often found to be maternally driven, with the most robust evidence across ages and raters emerging for maternal depression and general neuropsychiatric liabilities. Implications of all the available evidenceThe current evidence suggests that childrens behavioural difficulties, especially internalising difficulties, may be partly driven by the environment shaped by maternal neuropsychiatric liabilities. Ours and previous findings highlight a pressing need for more comprehensive studies across different cohorts, raters, outcomes, and time points to understand the true extent of IGEs in the intergenerational transmission of mental health.

BackgroundHelmet use is a proven safety measure that reduces the risk of head injury among cyclists and e-scooter riders. Despite legal requirements for pedal bikes and e-bikes in Australia, compliance varies, particularly among users of electric vehicles. The growing popularity of e-bikes and e-scooters in urban areas presents new public health challenges, yet observational data on helmet use, behavioural determinants, and the effectiveness of safety interventions remain limited. AimPhase 3 of the Helmet Use in Canberra study aims to characterise head injury presentations associated with cycling and e-scooter use and examine their association with helmet use and injury severity. MethodsDe-identified emergency department records from The Canberra Hospital will be retro-spectively analysed for presentations involving cycling or e-scooter-related head injuries during the study period. Extracted variables will include age, sex, vehicle type, documented helmet use, injury diagnosis, severity indicators, and date/time of presentation. Descriptive statistics will summarise injury patterns, while regression analyses will evaluate associations between helmet use and injury severity, controlling for demographic and contextual factors. Sensitivity analyses will address missing helmet data and subgroup differences by vehicle type, age, and gender. Expected ResultsIt is hypothesised that lower helmet use will correlate with higher rates and greater severity of head injury presentations. Findings will provide a population-level perspective on helmet effectiveness, inform local injury prevention strategies, and guide public safety interventions. Trial RegistrationAustralian and New Zealand Clinical Trials Registry (ANZCTR) [ACTRN12626000245392]

BackgroundThe epidemiology of suspected pediatric meningoencephalitis has shifted in the era of conjugate vaccines and multiplex PCR diagnostics, with viral pathogens now predominating over bacterial causes. Updated epidemiologic data are essential to adapt diagnostic and therapeutic algorithms to current clinical practice. MethodsThis retrospective single-center study included children and adolescents <18 years who underwent lumbar puncture with cerebrospinal fluid multiplex PCR for suspected central nervous system infection at a tertiary-care pediatric hospital in Germany between 2016 and 2024. Clinical, laboratory, and outcome data were extracted from electronic medical records. Cerebrospinal fluid was analyzed using the BioFire(R) FilmArray(R) Meningitis/Encephalitis Panel. Statistical analyses included descriptive statistics, nonparametric group comparisons, receiver operating characteristic analyses. ResultsAmong 1,198 included children, definite bacterial meningitis was diagnosed in 13 (1.1%), definite viral meningitis in 80 (6.7%), aseptic meningitis of unknown etiology in 131 (11.0%), confirmed/probable encephalitis in 53 (4.4%), and possible encephalitis in 34 (2.8%). Bacterial meningitis accounted for 5.8% of all meningitis cases. A causative pathogen was identified in all bacterial meningitis cases, most commonly Streptococcus pneumoniae (n = 7). Enterovirus (n = 52) and parechovirus (n = 9) predominated in viral meningitis, whereas an infectious etiology was identified in only 13 of 53 confirmed/probable encephalitis cases. The Bacterial Meningitis Score showed a sensitivity of 80.0% and a specificity of 57.6%. The recently published UK-ChiMES-pre- and post-lumbar puncture scores demonstrated sensitivities of 84.6% and 76.9% and specificities of 86.3% and 92.7%, respectively. DiscussionBacterial meningitis was rare in this contemporary cohort, while viral and etiologically unresolved infections predominated despite routine multiplex PCR diagnostics. Clinical prediction scores supported risk stratification, with the UK-ChiMES-pre-lumbar puncture score showing the most favorable balance between sensitivity and specificity and potential to guide diagnostic decisions and antiinfective therapy.

Machine learning adoption in clinical decision support systems remains limited by concerns about transparency and robustness. Causal structure learning (CSL) combined with expert knowledge may address these concerns by identifying potentially causal predictors, enabling more interpretable and clinically aligned models. In this study, we show that by integrating clinician expertise with CSL algorithms we can identify plausible causal drivers of acquired acute brain dysfunction (ABD) in the pediatric intensive care unit (PICU), which enables the development of parsimonious predictive models without substantial loss in performance. To do so, we analyzed 18,568 PICU encounters from the University of Pittsburgh Medical Center Childrens Hospital (2010-2022) and elicited knowledge from experienced clinicians. Encounters with acquired ABD were defined using the validated ABD computable phenotype. Expert knowledge was elicited from four clinicians through iterative interviews to construct a consensus directed acyclic graph (DAG). Clinician consensus achieved acceptable inter-rater reliability (Fleiss Kappa = 0.62) after two rounds of interviews and identified 16 biomarkers as potential causes of acquired ABD. Two CSL algorithms, GOLEM and PC-MB, were applied to enrich the clinicians consensus DAG. The PC-MB algorithm showed 78% concordance with expert consensus, while GOLEM showed 46%. Together, the CSL algorithms identified seven biomarkers as potential causes that were not included in the clinicians DAG: blood urea nitrogen, creatinine, dobutamine, glucose, potassium, PTT, SpO2. Using multiple variations of the enriched DAGs, XGBoost models were trained using biomarkers identified as potential causes of acquired ABD; these were evaluated primarily by area under the precision-recall curve (AUPRC). Models trained on the intersection of clinician consensus and PC-MB DAGs achieved an AUPRC of 0.79 (95% CI: 0.75-0.82) using only 14 biomarkers, compared to 0.81 (95% CI: 0.78-0.84) for the control model using all 45 biomarkers. When restricted to vitals and laboratory results alone, the best-performing model achieved an AUPRC of 0.77. Combining clinical expertise with causal structure learning enables the identification of causal hypotheses consistent with the clinical understanding of the participating clinicians and the development of parsimonious predictive models for acquired ABD in the PICU.

Pediatric asthma exacerbations are a frequent cause of emergency department (ED) visits and hospitalizations, yet accurate risk prediction remains limited and no consensus risk scores exist. Using UF Health electronic health records (EHRs) from 2011-2023, we evaluated two computable phenotypes (i.e., CAPriCORN and COMPAC) to predict exacerbations over 6-, 12-, and 24-month horizons. Exacerbations were defined using a validated composite of diagnosis codes from ED, inpatient, or outpatient encounters combined with systemic corticosteroids prescriptions. Several commonly used machine learning (ML) models were trained with stratified five-fold cross-validation, Bayesian hyperparameter optimization, and Youdens J thresholding. XGBoost achieved the best performance, with SHapley Additive exPlanations (SHAP) highlighting note-derived symptom terms and rescue-medication use as dominant predictors. Future work will focus on external validation and assessment of generalizability. This interpretable, text-integrated framework may support child-specific risk stratification and inform EHR-based decision support for timely pediatric asthma management.

RationaleObstructive sleep apnea (OSA) is linked to cardiovascular, metabolic, and cognitive morbidity. Although COVID-19 has been associated with long-term respiratory and neurological sequelae, its role in precipitating new-onset OSA remains unclear. ObjectivesTo evaluate whether SARS-CoV-2 infection increases risk of developing OSA up to 4.5 years post-infection and how risk varies by hospitalization status, demographics, comorbidities, and vaccination status. MethodsThis retrospective cohort study used electronic health records from the Montefiore Health System in the Bronx. Adults tested for SARS-CoV-2 between March 1, 2020, and August 17, 2024, were classified as hospitalized COVID+, non-hospitalized COVID+, or COVID-. Patients with prior OSA or inadequate follow-up were excluded. Inverse probability weighting adjusted for demographic, clinical, socioeconomic, and vaccination covariates. New-onset OSA was assessed using weighted Cox proportional hazards models. Secondary outcomes including hypertension, myocardial infarction, heart failure, stroke, arrhythmia, pulmonary hypertension, type 2 diabetes, and obesity were evaluated with Poisson regression. Sensitivity analysis used a pre-pandemic control cohort. ResultsAmong 910,393 eligible patients, hospitalized [HR 1.41 (95% CI 1.14-1.73)] and non-hospitalized [HR 1.33 (95% CI 1.22-1.46)] COVID+ patients had higher adjusted risk of new-onset OSA versus COVID- controls. Similar findings were observed using historical controls (n=621046). After OSA onset, hospitalized COVID+ patients had higher risks of heart failure and pulmonary hypertension, while non-hospitalized COVID+ patients had higher risk of obesity vs COVID- patients. ConclusionsSARS-CoV-2 infection is independently associated with increased risk of new-onset OSA. These findings support targeted screening in post-COVID populations.

IntroductionRecreational and medical cannabis use (CU) information is often available within the electronic health record (EHR) in a format that is impractical for health care provider use. Transformation of free-text EHR documentation in notes to discrete elements is possible using natural language processing (NLP) and has the potential to characterize CU efficiently. The objective of this study was to develop an NLP algorithm to identify documentation of CU within EHR unstructured clinical notes. MethodsWe identified EHR notes with cannabis-related terminologies through a keyword search among all Geisinger patients with at least one encounter between 1/1/2013 and 6/30/2022. We trained four NLP models to classify notes into six categories based on time, context, and reliability of CU documentation identified through manual annotation. We compared the demographic characteristics of patients with positive classification for CU using the best-performing model to those of the overall population. ResultsOf the over 1.7 million eligible patients, 150,726 (8.6%) were flagged as cannabis users. The Bio-ClinicalBERT, a transformer-based NLP model, achieved close to human performance in classifying CU (weighted Precision=91.4, Recall=93.3, F-score=92.4). Cannabis users had higher BMI and were at least nine-fold more likely to use tobacco, alcohol, and illicit substances. ConclusionOur study evaluated the prevalence of CU documentation across the entire corpus of EHR notes data without population segmentation. The NLP methodologies used achieved performance close to that of human annotation and laid the foundation for identifying and classifying CU within unstructured data sources, with future applications in research and patient care. Plain Language SummaryMarijuana, also known as cannabis, may impact the health of patients, yet it is not routinely captured in medical records, and when documented, it is often found in unstructured formats (e.g., progress notes) rather than in discrete fields. Incomplete and unstructured capture limits many functional capabilities within the EHR that enhance patient care (e.g., drug interactions, notifications) and limit researchers from identifying patients routinely exposed to marijuana use. The transformation of free-text documentation of cannabis use (CU) into discrete elements can be performed using natural language processing (NLP). The objective of this study was to develop an NLP model to identify CU in unstructured clinical notes in the EHR. We examined the EHRs of Geisinger patients in Pennsylvania over a 10-year period. Among 1.7 million patients, 9% were identified as CU. One of the NLP models tested, Bio-ClinicalBERT, achieved the highest performance. Cannabis users had a higher BMI and were ten-fold more likely to be tobacco users, ten-fold more likely to use alcohol, and nine-fold more likely to use illicit substances. NLP can be used to better understand the risks and benefits of CU at a population level and may improve patient identification to assist clinical decision-making. Future CU epidemiological research should continue to explore other avenues to automate and improve CU documentation by leveraging rapidly evolving technologies, such as artificial intelligence-driven tools.

Introduction"Encompass" is a participatory group-based intervention originating from low- and middle-income countries, co-developed with parents and professionals to enhance the wellbeing, health literacy and empowerment of caregivers of young children with complex neurodisability. We aimed to assess feasibility and acceptability of a) intervention delivery in two socially deprived United Kingdom (UK) urban areas and b) evaluation methods including data collection on programme outcomes and costs. MethodsWe conducted a mixed-methods pilot and feasibility study with caregivers of children under five years with complex neurodisability. Feasibility and acceptability of intervention delivery were assessed based on recruitment rates, group attendance, fidelity checklists and qualitative interviews with caregivers and facilitators. Feasibility and acceptability of evaluation methods were explored through follow-up rates, questionnaire completeness, and caregiver feedback on outcome measures. Data relating to implementation at organisational and system levels were explored through interviews with facilitators and key partners. Results were compared to predefined traffic light criteria (green, amber, red) to determine whether a larger scale evaluation was warranted. ResultsEight caregivers participated in the programme. Fidelity of delivery and follow-up questionnaire completion met green criteria, while recruitment and attendance met amber criteria, indicating that minor adaptations are required before scaling up. Qualitative findings demonstrated high acceptability of the programme among caregivers and facilitators, particularly valuing the co-facilitation model, participatory approach, and peer support. Flexible delivery, including online participation and communication support, enhanced accessibility for families with diverse needs. Capturing programme delivery costs was feasible and provided preliminary estimates to inform future economic evaluation. ConclusionsOur findings provide proof of principle that "Encompass" can feasibly and acceptably be delivered and evaluated with caregivers of children with complex neurodisability in an ethnically diverse UK community health setting. The findings support progression to a larger-scale evaluation, with refinements to recruitment strategies and delivery logistics. Patient or Public ContributionCaregivers with lived experience were central to developing the "Encompass" programme and this study. Four local mothers of children with complex neurodisability contributed to planning, recruitment, and sense-checking the findings.

BackgroundHelmet use is a proven safety measure that reduces the risk of head injury among cyclists and e-scooter riders. Despite legal requirements for pedal bikes and e-bikes in Australia, compliance varies, particularly among users of electric vehicles. The growing popularity of e-bikes and e-scooters in urban areas presents new public health challenges, yet observational data on helmet use, behavioural determinants, and the effectiveness of safety interventions remain limited. AimPhase 4 of the Helmet Use in Canberra study aims to identify demographic and behavioural predictors of unsafe riding and to explore perceived barriers and facilitators to helmet use, including compliance with existing regulations. MethodsA cross-sectional survey will be administered to Canberra residents aged 18 years or older, both online and in-person. The survey will assess attitudes toward helmet use, perceptions of head injury risk, and the deterrent effect of fines. Data will capture demographic characteristics, vehicle type, riding behaviours under varying conditions, and opinions regarding mandatory helmet laws and signage interventions. Survey responses will be de-identified, securely stored, and analysed using descriptive statistics and ordinal logistic regression to evaluate factors influencing compliance. Survey findings will be triangulated with observational and hospital data from earlier study phases. Expected ResultsThe survey is anticipated to provide insights into public attitudes toward helmet use, the perceived effectiveness of fines as behavioural deterrents, and the acceptability of policy interventions. These findings will inform evidence-based strategies to improve helmet compliance and reduce head injuries among urban riders. Trial RegistrationAustralian and New Zealand Clinical Trials Registry (ANZCTR) [ACTRN12626000245392].

ImportancePrenatal cannabis exposure is increasing in prevalence, yet its associations with early brain development--particularly how the timing and frequency of exposure across gestation relate to neonatal brain structure--remain insufficiently understood. Clarifying these associations is essential for informing early risk identification and guiding perinatal care. ObjectiveTo examine associations between patterns of maternal prenatal cannabis exposure, including exposure presence, gestational timing, and frequency of exposure, and neonatal brain structure and microstructure during the first month of life. Design, Setting, and ParticipantsThis cohort study included 1,782 mother-infant dyads (221 with PCE) from the HEALthy Brain and Child Development Study. Mother-reported prenatal cannabis exposure was assessed using the validated Timeline Follow-back method. Infants underwent natural-sleep magnetic resonance imaging, including T2-weighted structural imaging and diffusion imaging, within the first month of life. Main Outcomes and MeasuresAssociations between prenatal cannabis exposure and regional T2-weighted volumes and diffusion white matter microstructure metrics examined (1) exposure presence, (2) gestational timing of exposure, and (3) frequency of exposure within exposed infants. ResultsAny prenatal cannabis exposure was associated with brain volume differences in cerebellar and subcortical limbic regions, including smaller amygdala, thalamic, and cerebellar vermis volumes and larger caudate, hippocampal, and cerebellar cortex volumes. Timing-specific analyses revealed divergent patterns: first trimester exposure was associated with smaller volumes in select regions, whereas exposure that continued into the third trimester was associated with larger volumes in overlapping structures, with additional subcortical volumetric differences observed. White matter microstructure alterations were observed only among infants with exposure that continued into the third trimester. Within the exposed subgroup, higher frequency of cannabis exposure was associated with larger cerebral white matter volumes and white matter microstructural differences in white matter regions. Conclusions and RelevanceIn infants with maternal prenatal cannabis exposure, we observed timing- and frequency-dependent differences in brain development within the first month of life. These findings underscore the importance of considering not only the presence of exposure, but also when and how much cannabis is used during pregnancy to support targeted prenatal counseling and early developmental monitoring for exposed infants. Key PointsO_ST_ABSQuestionC_ST_ABSIs prenatal cannabis exposure associated with brain development in the first month of life? FindingsIn a cohort[ABS] of 1,782 mother-infant dyads, prenatal cannabis exposure was associated with region-specific differences in neonatal brain volumes. Brain volume and diffusion white matter microstructure associations differed between exposure limited to the first trimester versus exposure that continued into the third trimester. Greater frequency of exposure across gestation was also associated with volumetric and microstructural differences. MeaningThe timing and frequency of prenatal cannabis exposure is associated with alterations in neonatal brain development, underscoring the importance of addressing cannabis use in pregnancy.

BACKGROUND AND OBJECTIVESSeasonal influenza vaccination has been shown to reduce the risk of influenza and severe complications among children 6 months and older. Since 2010, reported numbers of influenza-associated pediatric deaths among children aged <18 years have ranged from 37 during the 2011-2012 season to 289 during 2024-2025. We estimated influenza vaccine effectiveness (VE) against pediatric death from 2016-2017 through 2024-2025. METHODSWe conducted a case-cohort analysis comparing current season influenza vaccination status among reported influenza-associated pediatric deaths with survey estimates of influenza vaccination coverage in pediatric age groups. Underlying medical conditions and current seasonal influenza vaccination were obtained from surveillance case reports. We estimated vaccination odds ratios (OR) and 95% confidence intervals (CI) from logistic regression comparing influenza vaccination among children who died with vaccination coverage in comparison cohorts. VE was calculated as (1 - OR) x 100. RESULTSFrom August 2016 through July 2025, 1234 laboratory-confirmed influenza-associated pediatric deaths were reported among children aged 6 months--17 years. Of 1086 reported deaths including influenza vaccination information, 124 (23%) of 530 children with underlying medical conditions and 70 (13%) of 556 children without known conditions were fully vaccinated against influenza. Average influenza vaccination coverage in survey cohorts was 49%. VE was 80% (95% CI, 75% to 84%) overall, 77% (95% CI, 71% to 82%) among children with underlying medical conditions and 87% (95% CI, 84% to 89%) among children without known conditions. CONCLUSIONSInfluenza vaccination reduced risk of fatal influenza among children with or without known underlying medical conditions.