Joanna Thomson

Background: Infants with poor weight gain (PWG) are frequently hospitalized for evaluation and management. Diagnostic testing during hospitalization varies and may reflect practice variation and bias. Objective: To determine if diagnostic testing among infants hospitalized for PWG differed by community characteristics, as measured by the childhood opportunity index (COI). Methods: We conducted a retrospective cross‐sectional study of children, ages 2 weeks‐11 months, hospitalized with PWG at 43 children's hospitals in the Pediatric Health Information Systems database between 2017 and 2022. COI, measured at the zip‐code level, was categorized into nationally normed quintiles. Logistic regression models estimated adjusted associations between COI and diagnostic testing, accounting for confounders defined a priori and clustering by hospital. Results: Compared to infants from very high COI communities, those from very low COI communities were more likely to receive screening tests like CBC (64.2% vs. 54.2%, adjusted odds ratio [aOR] 1.22, 95% confidence interval [CI] 1.10, 1.35). In contrast, infants from very low COI communities were less likely to have diagnostic testing directed toward investigation of potentially specific medical etiologies, including upper gastrointestinal series (10.3% vs. 12.8%, aOR 0.67, 95% CI 0.56, 0.81). However, they were more likely to receive urine toxicology screening (1.5% vs. 0.6%, aOR 2.4, 95% CI 1.39, 4.14), suggestive of suspected neglect. Conclusions: This national study demonstrates community‐level variation in the diagnostic evaluation of infants hospitalized for PWG. While differences in psychosocial factors and healthcare access may contribute, these findings underscore the need to address potential bias in the care of infants with PWG.

We conducted a multicenter prospective cohort study to identify factors associated with discharge on higher respiratory support in children with tracheostomy hospitalized for bacterial tracheostomy‐associated infections (bTRAINs). We included children 0–21 years with tracheostomy hospitalized for bTRAIN at six freestanding children's hospitals between 2020 and 2024. The primary outcome was discharge on higher respiratory support compared to pre‐hospital baseline. Generalized linear mixed‐effects regression models accounted for patient and hospital clustering. We included 641 children representing 1180 hospitalizations, with a median age of 5 years; 45.9% were female, 35.6% Hispanic, and 20.0% Non‐Hispanic Black. 24.7% were discharged on higher respiratory support. In adjusted analyses, discharge on higher respiratory support was associated with increased oxygen requirement, escalation of chronic ventilator settings, acute ventilatory support, and longer length of stay. In our study, discharge on higher respiratory support was driven by illness severity, rather than demographic or home support factors.

Objective: Test the association between early antibiotic discontinuation and outcomes for children with a tracheostomy and positive respiratory viral PCR (RVP) hospitalized with suspected bTRAIN (bacterial TRacheostomy‐Associated INfection; pneumonia/tracheitis). Methods: We conducted a multicenter cohort study between September 1, 2021 and August 31, 2024 of children with pre‐existing tracheostomy hospitalized for suspected bTRAIN (tracheal aspirate bacterial culture sent and bTRAIN antibiotics started) at six children's hospitals with: (1) a positive RVP test; (2) no antibiotic pre‐treatment; and (3) length of stay (LOS) < 37 days (outlier). Mixed‐effects regression models identified independent associations between early antibiotic discontinuation and our primary outcome, LOS (in days). Inverse probability treatment weighting based on propensity scoring adjusted for measured confounding. We explored the association between early antibiotic discontinuation and 3 secondary outcomes: late ICU transfer, in‐hospital mortality, and 30‐day bTRAIN rehospitalization rates. Results: Of 542 hospitalizations in 365 children, the median admission age was 4 years (interquartile range [IQR]: 2–8 years). Common viruses included rhinovirus ( n = 282; 52%), SARS‐CoV‐2 ( n = 63; 12%), and respiratory syncytial virus ( n = 61; 11%). The median LOS was 7 days (IQR: 5–11 days). Only 28% ( n = 152) had early antibiotic discontinuation; full bTRAIN treatment was associated with a 23% (5%–45%) increased LOS, with estimated LOS means (95% CI) of 7.0 (5.8–8.4) days for early antibiotic discontinuation and 8.6 (7.8–9.4) days in the bTRAIN‐treated group. Secondary outcome rates (late ICU transfer = 14.6%; in‐hospital mortality = 0.9%; 30‐day bTRAIN rehospitalization rate = 8.3%) were similar rates between both groups. Conclusions: Given similar rates of serious outcomes in both groups, clinicians could consider early discontinuation of antibiotic treatment in children with suspected bTRAINs who have a positive RVP test.

Background: Radiologic imaging is routinely performed to aid in diagnosis for hospitalized children. Identifying and reducing variability in imaging practices can improve care while reducing harms and costs.Objectives: To identify common inpatient pediatric conditions with high prevalence of imaging and variation in imaging practices and imaging‐related costs across hospitals.Methods: We conducted a cross‐sectional study of children 0–18 years old admitted to 50 children's hospitals in the Pediatric Health Information Systems database between 2017 and 2019. We excluded patients with complex chronic conditions, pregnancy, newborn, or neonatal intensive care charges, and those who died or who were discharged to hospice or rehabilitation facilities. Conditions with at least 2500 discharges and in which ≥30% of patients received imaging were included. Outcomes included imaging frequency, standardized imaging‐related costs (including all modalities across all encounters), and hospital‐level variation in imaging costs using the intraclass correlation coefficient (ICC).Results: Of the 56 included conditions, imaging was most frequently conducted in patients with pectus excavatum (97.8%), scoliosis (96.2%), and intestinal obstruction (96%). Fracture ($21.4 million), trauma ($15.7 million), and appendicitis ($8.6 million) had the highest total imaging‐related costs. Conditions with the highest geometric mean standardized imaging costs were nervous system disorders ($507), osteomyelitis ($494), vesicoureteral reflux/hydronephrosis ($468). Scoliosis (ICC = 0.49) and preseptal cellulitis (ICC = 0.48) had the highest variation in imaging‐related costs.Conclusion: To reduce imaging overuse in hospitalized children, conditions with frequent imaging, high imaging‐related costs, and high hospital‐level variation in imaging practices should serve as priorities for future evidence generation, guideline development, and/or improvement initiatives.

Airway clearance therapies (ACT) are often used to optimize respiratory function for children with neurologic impairment (CNI) hospitalized with acute respiratory infections (ARI). In a five‐center retrospective cohort study of CNI aged 1–18 years hospitalized between 2013 and 2015 with ARI, we assessed the association of admission ACT with hospital outcomes (days to return to baseline respiratory support and length of stay [LOS]). Generalized estimated equation (GEE) models examined the association between ACT and outcomes, while accounting for clustering. Propensity scores estimating the likelihood of receiving ACT were included in the GEE models as inverse probability of treatment weights to account for confounding. In adjusted analyses, children receiving admission ACT required similar time to return to respiratory baseline (adjusted risk ratio [aRR] 1.24, 95%CI: 0.95,1.62) and had similar LOS (aRR 1.08, 95%CI: 0.88,1.32) compared with children not receiving ACT. Future studies should seek to identify subpopulations of CNI who benefit most from ACT.

The Pediatric Hospital Medicine (PHM) Fellowship Directors, recent fellowship graduates, and senior leaders in PHM have long identified training in scholarly activities as a key educational priority for fellowship training programs. We led a 2-day conference funded by the Agency for Healthcare Research and Quality to develop scholarship core competencies for PHM fellows. Participants included fellowship directors, national experts in PHM research, and representatives from key stakeholder organizations. Through engagement in large group presentations and small group iterative feedback and editing, participants created and refined a set of scholarship core competencies. After the conference, goals and objectives were edited and harmonized by conference leaders incorporating feedback from conference participants. Core competency development included 7 domains: (1) study design and execution, (2) data management, (3) principles of analytics, (4) critical appraisal of the medical literature, (5) ethics and responsible conduct of research, (6) peer review, dissemination, and funding, and (7) professionalism and leadership. Specific objectives for each goal were further organized into 3 levels to indicate core skills for all fellowship trainees (level 1), specialized and specific skills determined by fellow scholarly focus (level 2), and advanced skills for fellows interested in a clinical investigator career path (level 3). These newly developed scholarship core competencies provide a foundation for curricular development and implementation to ensure that the field continues to expand academically, given the 2-year training period and variable infrastructure across programs.

Background: Children and young adults with medical complexity (CMC) experience high rates of healthcare reutilization following hospital discharge. Prior studies have identified common hospital‐to‐home transition failures that may increase the risk for reutilization, including medication, technology and equipment issues, financial concerns, and confusion about which providers can help with posthospitalization needs. Few interventions have been developed and evaluated for CMC during this transition period.Objective: We will compare the effectiveness of the garnering effective telehealth 2 help optimize multidisciplinary team engagement (GET2HOME) transition bundle intervention to the standard hospital‐based care coordination discharge process by assessing healthcare reutilization and patient‐ and family‐centered outcomes.Designs, Settings, and Participants: We will conduct a pragmatic 2‐arm randomized controlled trial (RCT) comparing the GET2HOME bundle intervention to the standard hospital‐based care discharge process on CMC hospitalized and discharged from hospital medicine at two sites of our pediatric medical center between November 2022 and February 2025. CMC of any age will be identified as having complex chronic disease using the Pediatric Medical Complexity Algorithm tool. We will exclude CMC who live independently, live in skilled nursing facilities, are in custody of the county, or are hospitalized for suicidal ideation or end‐of‐life care.Intervention: We will randomize participants to the bundle intervention or standard hospital‐based care coordination discharge process. The bundle intervention includes (1) predischarge telehealth huddle with inpatient providers, outpatient providers, patients, and their families; (2) care management discharge task tracker; and (3) postdischarge telehealth huddle with similar participants within 7 days of discharge. As part of the pragmatic design, families will choose if they want to complete the postdischarge huddle. The standard hospital‐based discharge process includes a pharmacist, social worker, and care management support when consulted by the inpatient team but does not include huddles between providers and families.Main Outcome and Measures: Primary outcome will be 30‐day urgent healthcare reutilization (unplanned readmission, emergency department, and urgent care visits). Secondary outcomes include 7‐day urgent healthcare reutilization, patient‐ and family‐reported transition quality, quality of life, and time to return to baseline using electronic health record and surveys at 7, 30, 60, and 90 days following discharge. We will also evaluate heterogeneity of treatment effect for the intervention across levels of financial strain and for CMC with high‐intensity neurologic impairment. The primary analysis will follow the intention‐to‐treat principle with logistic regression used to study reutilization outcomes and generalized linear mixed modeling to study repeated measures of patient‐ and family‐reported outcomes over time.Results: This pragmatic RCT is designed to evaluate the effectiveness of enhanced discharge transition support, including telehealth huddles and a care management discharge tool, for CMC and their families. Enrollment began in November 2022 and is projected to complete in February 2025. Primary analysis completion is anticipated in July 2025 with reporting of results following.

Introduction: Children with neurologic impairment (NI) are frequently hospitalized for infectious and noninfectious illnesses. The early period of the COVID‐19 pandemic was associated with overall lower pediatric hospitalization rates, particularly for respiratory infections, but the effect on utilization for children with NI is unknown.Method: This multicenter retrospective cohort study included hospitalizations of children 1–18 years of age with NI diagnosis codes from 49 children's hospitals. We calculated the percent change in the median weekly hospitalization volumes and the hospitalization resource intensity score (H‐RISK), comparing the early‐COVID era (March 15, 2020 to December 31, 2020) with the pre‐COVID era (same timeframe of 2017–2019). Percent change was calculated over the entire study period as well as within three seasonal time periods (spring, summer, and fall/winter). Differences between infectious and noninfectious admission diagnoses were also examined.Results: Compared with the pre‐COVID era, there was a 14.4% decrease (interquartile range [IQR]: −33.8, −11.7) in the weekly median number of hospitalizations in the early‐COVID era; the weekly median H‐RISK score was 11.7% greater (IQR: 8.9, 14.9). Hospitalizations decreased for both noninfectious (−11.6%, IQR: −30.0, −8.0) and infectious (−35.5%, IQR: −51.1, −31.3) illnesses in the early‐COVID era. This decrease was the largest in spring 2020 and continued throughout 2020.Conclusions: For children with NI, there was a substantial and significant decrease in hospitalizations for infectious and noninfectious diagnoses but an increase in illness severity during the early‐COVID era compared with the pre‐COVID era. Our data suggest a need to reconsider current thresholds for hospitalization and identify opportunities to support and guide families through certain illnesses without hospitalization.