Early Identification of Candidates for Epilepsy Surgery: A Multicenter, Machine Learning, Prospective Validation Study.

Background And Objectives: Epilepsy surgery is often delayed. We previously developed machine learning (ML) models to identify candidates for resective epilepsy surgery earlier in the disease course. In this study, we report the prospective validation.

Machine Learning-Based Prediction of Escitalopram and Sertraline Side Effects With Pharmacokinetic Data in Children and Adolescents.

Selective serotonin reuptake inhibitors (SSRI) are the first-line pharmacologic treatment for anxiety and depressive disorders in children and adolescents. Many patients experience side effects that are difficult to predict, are associated with significant morbidity, and can lead to treatment discontinuation. Variation in SSRI pharmacokinetics could explain differences in treatment outcomes, but this is often overlooked as a contributing factor to SSRI tolerability.

Development and User Testing of a Dynamic Tool for Rheumatic Heart Disease Management.

Objective: Most rheumatic heart disease (RHD) registries are static and centralized, collecting epidemiological and clinical data without providing tools to improve care. We developed a dynamic cloud-based RHD case management application with the goal of improving care for patients with RHD in Uganda.

Methods: The Active Community Case Management Tool (ACT) was designed to improve community-based case management for chronic disease, with RHD as the first test case.

Automated, machine learning-based alerts increase epilepsy surgery referrals: A randomized controlled trial.

Objective: To determine whether automated, electronic alerts increased referrals for epilepsy surgery.

Methods: We conducted a prospective, randomized controlled trial of a natural language processing-based clinical decision support system embedded in the electronic health record (EHR) at 14 pediatric neurology outpatient clinic sites. Children with epilepsy and at least two prior neurology visits were screened by the system prior to their scheduled visit.

Built environment factors predictive of early rapid lung function decline in cystic fibrosis.

Background: The extent to which environmental exposures and community characteristics of the built environment collectively predict rapid lung function decline, during adolescence and early adulthood in cystic fibrosis (CF), has not been examined.

Objective: To identify built environment characteristics predictive of rapid CF lung function decline.

Methods: We performed a retrospective, single-center, longitudinal cohort study (n = 173 individuals with CF aged 6-20 years, 2012-2017).

Implementation of Machine Learning Pipelines for Clinical Practice: Development and Validation Study.

Background: Artificial intelligence (AI) technologies, such as machine learning and natural language processing, have the potential to provide new insights into complex health data. Although powerful, these algorithms rarely move from experimental studies to direct clinical care implementation.

Objective: We aimed to describe the key components for successful development and integration of two AI technology-based research pipelines for clinical practice.

Standardizing electronic health record ventilation data in the pediatric long-term mechanical ventilator-dependent population.

Background: Sharing data across institutions is critical to improving care for children who are using long-term mechanical ventilation (LTMV). Mechanical ventilation data are complex and poorly standardized. This lack of data standardization is a major barrier to data sharing.

The Risk of Coding Racism into Pediatric Sepsis Care: The Necessity of Antiracism in Machine Learning.

Machine learning holds the possibility of improving racial health inequalities by compensating for human bias and structural racism. However, unanticipated racial biases may enter during model design, training, or implementation and perpetuate or worsen racial inequalities if ignored. Pre-existing racial health inequalities could be codified into medical care by machine learning without clinicians being aware.

Aligning Provider Prescribing With Guidelines for Soft Tissue Infections.

Objective: Despite evidence-based guidelines, antibiotics prescribed for uncomplicated skin and soft tissue infections can involve inappropriate microbial coverage. Our aim was to evaluate the appropriateness of antibiotic prescribing practices for mild nonpurulent cellulitis in a pediatric tertiary academic medical center over a 1-year period.

Methods: Eligible patients treated in the emergency department or urgent care settings for mild nonpurulent cellulitis from January 2017 to December 2017 were identified by an International Classification of Diseases, Tenth Revision, code for cellulitis.

An ontology-based review of transgender literature: Revealing a history of medicalization and pathologization.

Objectives: To evaluate the linguistic changes of transgender-related resources prior to 1999 to create a comprehensive dataset of resources using an ontology-derived search system, laying a framework for ontology-based reviews to be used in informatics.

Methods: We analyzed 77 bibliographies and 11 databases for transgender resources published prior to 31 December 1999. We used 858 variants of the term "transgender" to identify resources.

Early identification of epilepsy surgery candidates: A multicenter, machine learning study.

Objectives: Epilepsy surgery is underutilized. Automating the identification of potential surgical candidates may facilitate earlier intervention. Our objective was to develop site-specific machine learning (ML) algorithms to identify candidates before they undergo surgery.

Development of the Gender, Sex, and Sexual Orientation ontology: Evaluation and workflow.

Objective: The study sought to create an integrated vocabulary system that addresses the lack of standardized health terminology in gender and sexual orientation.

Materials And Methods: We evaluated computational efficiency, coverage, query-based term tagging, randomly selected term tagging, and mappings to existing terminology systems (including ICD (International Classification of Diseases), DSM (Diagnostic and Statistical Manual of Mental Disorders ), SNOMED (Systematized Nomenclature of Medicine), MeSH (Medical Subject Headings), and National Cancer Institute Thesaurus).

Results: We published version 2 of the Gender, Sex, and Sexual Orientation (GSSO) ontology with over 10 000 entries with definitions, a readable hierarchy system, and over 14 000 database mappings.

An interactive online dashboard for tracking COVID-19 in U.S. counties, cities, and states in real time.

Objective: The study sought to create an online resource that informs the public of coronavirus disease 2019 (COVID-19) outbreaks in their area.

Materials And Methods: This R Shiny application aggregates data from multiple resources that track COVID-19 and visualizes them through an interactive, online dashboard.

Results: The Web resource, called the COVID-19 Watcher, can be accessed online (https://covid19watcher.

Information Technology-Assisted Screening for Gonorrhea and Chlamydia in a Pediatric Emergency Department.

Purpose: The aim of the study was to design and implement a novel, universally offered, computerized clinical decision support (CDS) gonorrhea and chlamydia (GC/CT) screening tool embedded in the emergency department (ED) clinical workflow and triggered by patient-entered data.

Methods: The study consisted of the design and implementation of a tablet-based screening tool based on qualitative data of adolescent and parent/guardian acceptability of GC/CT screening in the ED and an advisory committee of ED leaders and end users. The tablet was offered to adolescents aged 14-21 years and informed patients of Centers for Disease Control and Prevention GC/CT screening recommendations, described the testing process, and assessed whether patients agreed to testing.

Automated Patient Linking for Electronic Health Record and Child Welfare Databases.

There are 427,000 children in protective custody in the United States. A lack of integration between the child welfare data system and electronic health record systems complicates the communication of critical health history details to caregivers. We created and evaluated automated ten custom algorithms linking these data.

Investigation of bias in an epilepsy machine learning algorithm trained on physician notes.

Racial disparities in the utilization of epilepsy surgery are well documented, but it is unknown whether a natural language processing (NLP) algorithm trained on physician notes would produce biased recommendations for epilepsy presurgical evaluations. To assess this, an NLP algorithm was trained to identify potential surgical candidates using 1097 notes from 175 epilepsy patients with a history of resective epilepsy surgery and 268 patients who achieved seizure freedom without surgery (total N = 443 patients). The model was tested on 8340 notes from 3776 patients with epilepsy whose surgical candidacy status was unknown (2029 male, 1747 female, median age = 9 years; age range = 0-60 years).

Sharing personal health record data elements in protective custody: youth and stakeholder perspectives.

Objective: The study sought to develop the necessary elements for a personalized health record (PHR) for youth emancipating from child protective custody (eg, foster care) by collecting thoughts and ideas from current and former foster youth and community stakeholders who have a significant amount of experience working with emancipating young people.

Materials And Methods: We employed a mixed methods, participatory research design using concept mapping to identify key features for PHR across stakeholders.

Results: In the clusters, common themes for necessary elements for a PHR included health education, health tips, medication instructions, diagnoses including severity, and website resources that could be trusted to provide reliable information, and addressed data privacy issues such as the primary user being able to choose what diagnoses to share with their trusted adult and the ability to assign a trusted adult to view a part of the record.

Improving Information Sharing for Youth in Foster Care.

There are ∼443 000 children in child protective custody (ie, foster care) in the United States. Children in protective custody have more medical, behavioral, and developmental problems that require health care services than the general population. These health problems are compounded by poor information exchange impeding care coordination.

Mathematical Model for Computer-Assisted Modification of Medication Dosing Rules.

Objective: Medication dosing in pediatrics is complex and prone to errors that may lead to patient harm. To improve computer-assisted dosing, a mathematical model and algorithm were developed to optimize clinical decision support dosing rules and reduce spurious alerts. The objective was to evaluate the feasibility of using this algorithm to adjust dosing rules.