A5083306075- Opioid Use Disorder Treatment
- Machine Learning in Healthcare
- Substance Abuse Treatment and Outcomes
- Topic Modeling
- Emergency and Acute Care Studies
- Biomedical Text Mining and Ontologies
- COVID-19 Clinical Research Studies
- Pharmaceutical Practices and Patient Outcomes
- Data Quality and Management
- COVID-19 and Mental Health
- Artificial Intelligence in Healthcare and Education
- Electronic Health Records Systems
- Mental Health Treatment and Access
- Prenatal Substance Exposure Effects
- Long-Term Effects of COVID-19
- Alcohol Consumption and Health Effects
- Acute Ischemic Stroke Management
- Blockchain Technology Applications and Security
- COVID-19 and healthcare impacts
- Pain Management and Opioid Use
- Pharmacovigilance and Adverse Drug Reactions
- Alcoholism and Thiamine Deficiency
- Clinical Reasoning and Diagnostic Skills
- Digital Mental Health Interventions
- Homelessness and Social Issues
University of Washington
2024
University of Wisconsin–Madison
2022-2023
Rush University Medical Center
2020-2023
Rush University
2021
Loyola University Chicago
2018-2020
Loyola University Medical Center
2019
Abstract Objectives To assess fairness and bias of a previously validated machine learning opioid misuse classifier. Materials & Methods Two experiments were conducted with the classifier’s original (n = 1000) external validation 53 974) datasets from 2 health systems. Bias was assessed via testing for differences in type II error rates across racial/ethnic subgroups (Black, Hispanic/Latinx, White, Other) using bootstrapped 95% confidence intervals. A local surrogate model estimated to...
Background The clinical narrative in electronic health records (EHRs) carries valuable information for predictive analytics; however, its free-text form is difficult to mine and analyze decision support (CDS). Large-scale natural language processing (NLP) pipelines have focused on data warehouse applications retrospective research efforts. There remains a paucity of evidence implementing NLP at the bedside care delivery. Objective We aimed detail hospital-wide, operational pipeline implement...
Approaches are needed to better delineate the continuum of opioid misuse that occurs in hospitalized patients. A prognostic enrichment strategy with latent class analysis (LCA) may facilitate treatment strategies subtypes misuse. We aim identify patients and examine distinctions between by examining patient characteristics, topic models from clinical notes, outcomes.This was an observational study inpatient hospitalizations at a tertiary care center 2007 2017. Patients were identified using...
Abstract Background Automated de-identification methods for removing protected health information (PHI) from the source notes of electronic record (EHR) rely on building systems to recognize mentions PHI in text, but they remain inadequate at ensuring perfect removal. As an alternative relying systems, we propose following solutions: (1) Mapping corpus documents standardized medical vocabulary (concept unique identifier [CUI] codes mapped Unified Medical Language System) thus eliminating as...
Abstract Objective Natural language processing (NLP) engines such as the clinical Text Analysis and Knowledge Extraction System are a solution for notes research, but optimizing their performance data warehouse remains challenge. We aim to develop high throughput NLP architecture using present predictive model use case. Materials Methods The CDW was comprised of 1 103 038 patients across 10 years. constructed Hadoop repository source 3 large-scale symmetric servers NLP. Each named entity...
Background Addiction medicine consultation services (ACS) may improve outcomes of hospitalized patients with substance use disorders (SUD). Our aim was to examine the difference in length stay and hazard ratio for a routine hospital discharge between SUD receiving not ACS. Methods Structured EHR data from 2018 1,900 adult SUD-related diagnostic code at an urban academic health center were examined among 35,541 total encounters. Cox proportional hazards regression models fit using...
Abstract Background Opioid misuse screening in hospitals is resource-intensive and rarely done. Many hospitalized patients are never offered opioid treatment. An automated approach leveraging routinely captured electronic health record (EHR) data may be easier for to institute. We previously derived internally validated an classifier a separate hospital setting. The aim externally validate our published open-source machine-learning at different identifying cases of misuse. Methods...
Background Unhealthy alcohol use (UAU) is known to disrupt pulmonary immune mechanisms and increase the risk of acute respiratory distress syndrome in patients with pneumonia; however, little about effects UAU on outcomes COVID-19 pneumonia. To our knowledge, this first observational cross-sectional study that aims understand effect severity COVID-19. Objective We aim determine if associated more severe clinical presentation worse health related socioeconomic status, smoking, age, BMI,...
Generative artificial intelligence (AI) is a promising direction for augmenting clinical diagnostic decision support and reducing errors, leading contributor to medical errors. To further the development of AI systems, Diagnostic Reasoning Benchmark (DR.BENCH) was introduced as comprehensive generative framework, comprised six tasks representing key components in reasoning. We present comparative analysis in-domain versus out-of-domain language models well multi-task single task training...
Abstract Background and Aims Accurate case discovery is critical for disease surveillance, resource allocation research. International Classification of Disease (ICD) diagnosis codes are commonly used this purpose. We aimed to determine the sensitivity, specificity positive predictive value (PPV) ICD‐10 opioid misuse in emergency department (ED) setting. Design Setting Retrospective cohort study ED encounters from January 2018 December 2020 at an urban academic hospital United States. A...
Protecting the integrity of medical records is critical to patients, professionals, governments, insurance companies, and hospital systems. Integrity protecting data against accidental or fraudulent changes. Owing onset electronic in physician offices health-care facilities, increasingly mandated incentivized by federal Acts USA, ensuring an environment has become far more important than days paper-based records. In this research, we propose innovative Merkle tree-based approach describe its...
Unhealthy alcohol use (UAU) is one of the leading causes global morbidity. A machine learning approach to screening could accelerate best practices when integrated into electronic health record (EHR) systems. This study aimed validate externally a natural language processing (NLP) classifier developed at an independent medical center.Retrospective cohort study.The site for validation was midwestern United States tertiary-care, urban center that has inpatient structured universal model...
Background Automated and data-driven methods for screening using natural language processing (NLP) machine learning may replace resource-intensive manual approaches in the usual care of patients hospitalized with conditions related to unhealthy substance use. The rigorous evaluation tools that use artificial intelligence (AI) is necessary demonstrate effectiveness before system-wide implementation. An NLP tool routinely collected data electronic health record was previously validated...
The emergency department (ED) is a critical setting for the treatment of patients with opioid misuse. Detecting relevant clinical profiles allows tailored approaches. We sought to identify and characterize subphenotypes ED opioid-related encounters. A latent class analysis was conducted using 14,057,302 encounters from 2016 through 2017 National Emergency Department Sample (NEDS), largest all-payer database in United States. optimal model determined by face validity information...
The COVID-19 pandemic has exacerbated health inequities in the United States. People with unhealthy opioid use (UOU) may face disproportionate challenges precautions, and disrupted access to opioids UOU treatments. impairs immunological, cardiovascular, pulmonary, renal, neurological systems increase severity of outcomes for COVID-19.
Background: Substance misuse is a heterogeneous and complex set of behavioral conditions that are highly prevalent in hospital settings frequently co-occur. Few solutions exist to comprehensively reliably identify these hospital-wide prioritize care guide treatment. The aim apply natural language processing (NLP) admission notes the electronic health record (EHR) accurately screen for substance misuse.Methods: reference dataset was derived from program used structured diagnostic interviews...
Generative artificial intelligence (AI) is a promising direction for augmenting clinical diagnostic decision support and reducing errors, leading contributor to medical errors. To further the development of AI systems, Diagnostic Reasoning Benchmark (DR.BENCH) was introduced as comprehensive generative framework, comprised six tasks representing key components in reasoning. We present comparative analysis in-domain versus out-of-domain language models well multi-task single task training...