A pediatrician checking in with a young patient.
About
Login

Explore PEDSpace Collections

Recent Submissions

  • Systematic Data Fitness Assessment Improves Validity and Replicability of Research Using Real-World Data [Poster]
    Published:2026-10-14Journal:PEDSnet
    Secondary use of electronic health record (EHR) data gives investigators the ability to conduct research and derive insights at a large scale for a lesser cost. These data are complex, so multi-institutional research networks often implement broad data quality (DQ) programs to ensure the data is generally usable and meet network governance standards. However, study-specific needs are typically out of scope for these programs, leaving researchers with the responsibility of conducting ad hoc analyses to ensure the data is fit-for-use in the context of their study. Our team has developed a reusable, modularized tool to assess study-specific data quality (SSDQ), which is intended to reduce time spent developing ad hoc analyses and improve transparency and replicability in research. Using this tool as a foundation, we demonstrate the utility of undertaking rigorous DQ assessment by replicating a study examining the impact of hydroxyurea treatment in a cohort of pediatric patients with the SS/Sβ0 sickle-cell genotype. Poster designed and presented by Kaleigh Wieand on October 14, 2026 at the Science, Population Health, Analytics, Research, Collaboration, Community (SPARCC) data summit.
  • Systematic Data Fitness Assessment Improves Validity and Replicability of Research Using Real-World Data [Preprint]
    Published:2026-08-10Journal:medRxiv
    Research replication in real-world data is essential to build trust in evidence from clinical studies. However, methods for conducting and reporting these efforts are lacking, particularly related to data fitness and limitations. We demonstrate the importance of incorporating systematic fitness testing by replicating a single-center observational study of hydroxyurea for children with severe sickle cell disease (SS/Sβ0) in a multi-institutional learning network using EHR data (PEDSnet). An AS-IS arm applied the original study’s criteria with no major data quality adjustments, while a Data Fitness Enhanced (DFE) arm used systematic data fitness assessment to inform adjustments to cohort eligibility criteria and variable definitions; both arms then replicated the original study’s primary analyses. Data quality checks in the DFE arm refined cohort accuracy and improved hydroxyurea capture, drug era computation, and hematology specialist mapping. The DFE cohort produced average treatment effects with higher face validity and greater concordance with the original study (e.g., change in ED visits: -0.44 (CI -0.60, -0.26) versus -0.36 (CI -0.57, -0.16) in the original study) than the AS-IS cohort (-0.08 (CI -0.26, 0.09)), which yielded several implausible results. These findings show that superficially plausible cohort characteristics do not guarantee valid results without transparent, systematic data fitness assessment.
  • Declining Child Health in the United States — Addressing the Developmental Ecosystem
    Published:2026-09-24Journal:New England Journal of Medicine
    ### Abstract To address drivers of poor child health, society should move from responding to health crises one symptom and one child at a time to reimagining ecosystems in ways that facilitate healthy development.
  • Augmenting Structured Diagnoses Through Effective Use of Pre-Trained Large Language Models on Clinical Notes
    Published:2026-09-09Journal:JAMIA Open
    **Objective:** Clinical narrative provides a unique window into provider reasoning and attribution for automated diagnosis assignment, but large language models (LLMs) have traditionally not performed well at medical coding. We evaluate a reproducible method for automated diagnosis assignment using LLMs in clinical notes and compare with structured diagnoses. **Materials and Methods:** We used GPT-OSS for prompt engineering and task segmentation to create a model that extracts ICD-10-CM diagnoses, with estimates of severity, currency, and importance, from progress notes. We assessed performance across multiple cohorts of patients aged 0-21 years. For each, 100 outpatient provider notes were selected across levels of severity, along with coded diagnoses from that visit (electronic health record [EHR]); a subset of 130 notes were subjected to clinical expert review. **Results:** Comparison showed 18.7% exact code and 33.3% ICD-10-CM category match between EHR and LLM, but semantic similarity of 0.93 at the category level. Compared to expert review, LLM precision was 0.84 and recall 0.49 for exact matches, and 0.92 and 0.62, respectively, for category-level matching. In contrast, coded diagnoses showed slightly higher precision (0.94 for both cases) and substantially lower recall (0.27 and 0.43) versus expert review. Codes not identified by the LLM were more often rated by the reviewer as lower importance or certainty. **Discussion:** We demonstrate a reusable approach to optimize LLMs for use in diagnosis extraction from clinical notes that can augment structured diagnoses and provide contextualizing metadata. **Conclusion:** LLMs represent a viable and flexible approach to diagnosis code extraction from unstructured clinical notes.
  • Treatment Burden, Healthcare Experiences, and Health-Related Quality of Life Among Youth with Chronic Kidney Disease Treated for Hypertension
    Published:2026-09-24Journal:Qualitative Life Research
    Background: Controlling blood pressure (BP) is essential for slowing chronic kidney disease (CKD) progression. This study examined the associations between treatment burden, assessed by antihypertensive medication use, and health-related quality of life (HRQoL) in youth with CKD and hypertension. Methods: Youth (aged 8 to 21 years) were identified through electronic health records from 15 institutions. Youth and caregivers completed an online questionnaire assessing healthcare experiences, medication use, physical symptoms, and BP monitoring. HRQoL and related well-being domains (fatigue, anxiety, sleep disturbance, global health, and life satisfaction) were assessed using PROMIS® measures. Analyses focused on youth-reported outcomes. Linear regression estimated associations between antihypertensive medication use and HRQoL outcomes, and scores were compared to the general population. Results: The study included 395 youth (Mage = 16.8 years, 54% female) and 238 caregivers. Over 95% of caregivers reported generally positive healthcare experiences. Most youth (92%) reported at least one of the physical symptoms asssessed (median [IQR] = 5 [3, 8] symptoms). Twenty-one percent took no antihypertensive medications, 56% took one, and 23% took multiple. Use of multiple medications was associated with slightly lower global health (B = - 2.6; 95% CI: - 4.9, - 0.3) after adjustment for CKD stage. No associations were observed for other HRQoL domains. Overall, youth reported lower QoL than the general population, with the greatest impairments observed in global health and fatigue. Discussion: Youth with CKD and hypertension reported generally positive healthcare experiences, and antihypertensive medication use, one component of treatment burden, showed minimal association with HRQoL. Despite modest effects, continued attention to HRQoL remains an important aspect of care for youth with CKD and hypertension.