Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Min Jung Kim has a total of 48 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Thirty‐Day All‐Cause Readmissions After Pregnancy‐Associated Hospitalizations With Heart Failure in the United States: Analysis of the Nationwide Readmissions Database

    by Emma Barry, Min Jung Kim, Sarah Goldstein, Anna Denoble, Christine Hsueh, Patricia Chavez, Sara Tabtabai

    Background: Cardiovascular disease is the leading cause of pregnancy‐related morbidity and mortality in the United States. There is a paucity of data regarding pregnancy‐associated heart failure readmissions. This study aims to characterize the 30‐day heart failure readmission patterns in pregnancy‐associated admissions. Methods: The National Readmission Database from 2016 to 2021 was used to identify women aged 13 to 49 with an index hospitalization with heart failure as the primary or secondary diagnosis during a pregnancy‐related admission, identified by diagnosis‐related group codes and International Classification of Diseases, Tenth Revision ( ICD‐10 ) codes. The primary outcome was 30‐day all‐cause readmission. Survey‐weighted descriptive analyses and hierarchical multivariable logistic regression were used to characterize readmission patterns and identify independently associated factors. Results: Among 19 452 eligible index hospitalizations, 2659 were followed by a 30‐day readmission; the survey‐weighted readmission rate was 13.5%. In the fully adjusted model, Medicare and Medicaid coverage, residence in a small metropolitan area, greater All Patient Refined Diagnosis‐Related Group severity, discharge with home health care, and discharge against medical advice were associated with higher odds of readmission. Deficiency anemia, chronic pulmonary disease, diabetes without chronic complications, renal failure, solid tumor without metastasis, and cardiovascular or respiratory disorders complicating pregnancy were also independently associated with increased readmission. Conclusions: Approximately 1 in 7 pregnancy‐associated hospitalizations complicated by heart failure was followed by a 30‐day all‐cause readmission. Readmission risk reflected a combination of illness severity, discharge disposition, payer status, geographic location, and selected medical and pregnancy‐related comorbidities. These findings support the development and evaluation of individualized, multidisciplinary transitional‐care strategies for patients at highest risk.

  • A machine learning approach to identifying students at risk for USMLE Step 1 examination delay using academic performance data

    by Min-Jung Kim, Zita Lazzarini, Thomas M Manger, Christine Thatcher

    Purpose: Delaying the United States Medical Licensing Examination Step 1 examination is often linked to poorer performance on Step 1, clerkship, and Step 2 Clinical Knowledge examinations. This study developed and evaluated explainable machine learning models to predict Step 1 examination delay using student performance data and identify key predictors contributing to delay risk. Method: Data from 610 medical students at University of Connecticut School of Medicine for matriculation years 2016, 2017, 2019, 2020, 2021, and 2022 were analyzed, with 167 students (27.4%) delaying their Step 1 examinations. Explainable machine learning models, including RF (random forest), XGBoost (Extreme Gradient Boosting), CatBoost (Categorical Boosting), and regularized logistic regression, were evaluated using 5-fold cross-validation and testing datasets. Misclassification rate, area under the receiver operating characteristic curve, and area under the precision-recall curve were used to compare model performance. Shapley Additive Explanations and permutation feature importance were applied to interpret key predictors. Results: XGBoost outperformed all other models, achieving a median (interquartile range [IQR]) misclassification rate of 0.14 (0.11-0.17), area under the receiver operating characteristic curve of 0.80 (0.76-0.84), and area under the precision-recall curve of 0.72 (0.60-0.78). Shapley Additive Explanations and permutation feature importance analyses revealed that incorrect responses on the first Comprehensive Basic Science Self-Assessment (CBSSA), along with preclerkship foundational basic science and laboratory course scores, were the most influential predictors of Step 1 examination delay. Students with delays had a median (IQR) of 99 (82-112) incorrect responses on the first CBSSA vs 75 (56-92) for nondelayers (P < .001). Students with delays had significantly lower scores across the 5 preclerkship course blocks. Conclusions: Ensemble machine learning models, particularly XGBoost, provide strong predictive performance for identifying students at risk for delaying the Step 1 examination. Early academic indicators, such as CBSSA and course performance, can inform timely interventions, supporting students’ academic progression and success.

  • Open Access

    Predicting 30-Day Heart Failure Readmissions Using Machine Learning: Insights From the Kansas Health Information Network (KHIN)

    by Minjung Kim, Jun Yan, Jason H. Wasfy, Robert H. Aseltine

    Background: Heart failure (HF) is a major contributor to inpatient hospital utilization, with persistently high 30-day readmission rates. Existing prediction tools are frequently restricted to primary-diagnosis HF admissions, potentially excluding clinically relevant HF-related hospitalizations. Objectives: To develop and validate risk prediction models using machine learning (ML)-based risk prediction models to predict 30-day readmissions among patients with HF using the Kansas Health Information Network, a statewide health information exchange. Methods: This retrospective cohort study analyzed HF hospitalizations using predictors including demographics, comorbidities, laboratory results, medications, clinical quality metrics for diabetes and kidney disease management, and prior healthcare utilization. Five ML models, including regularized logistic regression, random forest, extreme gradient boosting, categorical boosting, and deep neural network, were trained using stratified 5-fold cross-validation. Model performance was evaluated on an independent test set using the area under the receiver operating characteristic curve (AUROC), area under the precision-recall curve (AUPRC), misclassification rate (MCR), and Brier score. Results: Among 2,734 HF patients, the 30-day readmission rate was 27%. The XGBoost model achieved the best discrimination (AUROC=0.75; AUPRC=0.58; MCR=0.21). Patients in the highest-risk decile had a positive predictive value of 76%, accounted for approximately one-third of all 30-day readmissions, and had a 3.3-fold enrichment compared with baseline risk. The key predictors included prior hospital utilization, diabetes and kidney disease management indicators, and comorbidity burden. Conclusions: Risk stratification using routinely collected clinical data identified a subgroup at elevated risk for 30-day readmission. These findings support the potential role of data-driven risk prediction to inform targeted transitional care. CLINICAL PERSPECTIVES: What is New?: This study is among the first to utilize health information exchange data to develop machine learning models for predicting 30-day heart failure readmissions, including patients with HF as either a primary or secondary diagnosis. Unlike prior models restricted to primary HF admissions, this approach includes hospitalizations in any diagnosis position, reflecting real-world acute-care utilization. Notably, our machine learning model correctly identified approximately three in four patients classified as high-risk in the top decile of predicted scores, highlighting its ability to identify individuals at greatest likelihood of early readmission What are the Clinical Implications?: Accurately identifying a group of high-risk patients has meaningful implications for care delivery. By correctly capturing actual readmissions within a small, prioritized segment of the population, ML-driven risk stratification can guide clinicians and care management teams in targeting transitional care resources where they are most needed. Such focused intervention strategies may allow for efficient targeting of limited health care resources, improve post-discharge outcomes, enhance operational efficiency, and potentially reduce preventable HF readmissions.

  • Open Access

    Pregnancy-related Heart Failure Readmissions in the US: Analysis of the Nationwide Readmissions Database

    by Emma Barry, Min Jung Kim, Sarah A Goldstein, Anna E. Denoble, Patricia Chavez, Christine Hsueh, Sara Tabtabai

    Background: Cardiovascular disease (CVD) is the leading cause of pregnancy-related morbidity and mortality in the United States. Several studies have evaluated readmission rates in the general HF population, but in patients with pregnancy-related HF, readmissions have been understudied. This study aims to characterize the 30-day HF readmission patterns in pregnancy-related admissions to identify vulnerable patient populations. Methods: The National Readmission Database from 2016 to 2021 was used to identify women aged 13-49 with an index hospitalization in which HF was coded as either the primary or secondary diagnosis during a pregnancy-related antepartum, delivery, or postpartum admission, identified by diagnosis-related group (DRG) codes and ICD-10 codes. The primary outcome was 30-day all-cause readmission. We performed descriptive and comparative analyses to describe the differences in patient characteristics and readmission patterns between groups. Results: The overall 30-day all-cause readmission rate was 13% when readmissions for delivery were excluded. The readmission rate increased with age, peaking at 15.1% in the 38-49yr age group. Higher readmission rates were also associated with combined (systolic and diastolic) HF (16.1%), systolic HF (14.8%), lower socioeconomic status (15.3%), substance use disorder (17.2%), and alcohol use (18.6%). Patients whose index hospitalization was for delivery had the highest absolute risk of 30-day readmission at 19.3%. Readmissions peaked between days 6 and 8 post discharge, with more than 50% of all readmissions occurring within the first two weeks post-discharge Conclusions: In our study, the highest risk of readmission occurred after an index hospitalization for delivery, and most readmissions occurred in the first 2 weeks post-discharge. Our findings suggest that a post-discharge follow up within 7 days of admission complicated by HF should be extended to patients with pregnancy-related HF and effective readmission reduction strategies must include a better understanding of heart failure phenotypes, and a proactive approach to addressing social risk factors.

  • Outcomes and Predictors of In-hospital Mortality in Nonagenarians with NSTEMI: A Comparison of PCI and Medical Management

    by Minjung Kim, Priyesh Thakurathi, Jai Nagpal, Vaibhav Satija, Juyong Lee, Kyutae Park, Agnes S. Kim

  • Open Access

    Prognostic Value of Coronary CT in Outpatients with Chest Pain: A 5-Year MACE Analysis Stratified by ASCVD Risk

    by Dongwon Kim, Jinha Lee, Hyunhee Choi, Taesun Kim, Minjung Kim, Agnes S. Kim, Michael Azrin, Juyong Lee, Kyungsoon Hong, Kyutae Park

    Background Coronary computed tomography angiography (CCTA) is recommended for the diagnosis of initial coronary artery disease (CAD) in patients with stable chest pain. We sought to understand the prevalence and severity of coronary stenosis observed via CCTA and to determine how integrating these anatomical findings with conventional 10-year atherosclerotic cardiovascular disease (ASCVD) scores could enhance risk stratification and guide clinical decisions. Methods This was an open-label, prospective, single-center observational study including 1,492 outpatients with stable chest pain who underwent CCTA. We collected data on ASCVD risk factors and followed up patients for 5 years to monitor for major adverse cardiovascular events (MACE). We analyzed the prevalence of obstructive CAD (OCAD, ≥50% stenosis) across different ASCVD risk categories and its relationship with MACE. Results Among 1,492 patients, CCTA revealed OCAD in 16.0%. Over a 5-year follow-up, 7.2% of patients experienced MACE. The presence of OCAD significantly improved MACE prediction beyond ASCVD scores alone. Notably, patients with ASCVD < 7.5% and OCAD had a significantly higher MACE risk (adjusted hazards ratio: 3.634; p = 0.023) compared with those without OCAD. The highest risk was found in the ASCVD ≥ 7.5% with OCAD group (adjusted hazards ratio: 5.101; pConclusions CCTA provides significant incremental value for risk stratification in outpatients with stable chest pain. It helps uncover a high-risk group that might be underestimated by conventional ASCVD scores, supporting CCTA integration into clinical workups for earlier intervention and improved patient outcomes.

  • Detection of Subclinical Cardiac Dysfunction in Patients With Sickle Cell Disease Using Speckle-Tracking Echocardiography

    by Min-Jung Kim, Grace Lee, Gian Lima, Osama Mukarram, Samuel Crooks, Kristin Marshall, Agnes S. Kim


  • Open Access

    Cardiac and noncardiac biomarkers in patients undergoing anthracycline chemotherapy – a prospective analysis

    by Matthew Dean, Min Jung Kim, Sharon Dimauro, Susan Tannenbaum, Garth Graham, Bruce T. Liang, Agnes S. Kim

    Background: Biomarkers represent a potential tool to identify individuals at risk for anthracycline-induced cardiotoxicity (AICT) prior to symptom onset or left ventricular dysfunction. Methods: This study examined the levels of cardiac and noncardiac biomarkers before, after the last dose of, and 3–6 months after completion of doxorubicin chemotherapy. Cardiac biomarkers included 5th generation high-sensitivity cardiac troponin T (cTnT), N-terminal pro-brain natriuretic peptide, growth/differentiation factor-15 (GDF-15), and soluble suppression of tumorigenesis-2 (sST2). Noncardiac biomarkers included activated caspase-1 (CASP-1), activated caspase-3, C-reactive protein, tumor necrosis factor-α, myeloperoxidase (MPO), galectin-3, and 8-hydroxy-2’-deoxyguanosine. Echocardiographic data (LVEF and LVGLS) were obtained at pre- and post-chemotherapy. Subanalysis examined interval changes in biomarkers among high (cumulative doxorubicin dose ≥ 250 mg/m2) and low exposure groups. Results: The cardiac biomarkers cTnT, GDF-15, and sST2 and the noncardiac biomarkers CASP-1 and MPO demonstrated significant changes over time. cTnT and GDF-15 levels increased after anthracycline exposure, while CASP-1 and MPO decreased significantly. Subanalysis by cumulative dose did not demonstrate a larger increase in any biomarker in the high-dose group. Conclusions: The results identify biomarkers with significant interval changes in response to anthracycline therapy. Further research is needed to understand the clinical utility of these novel biomarkers.


  • Open Access

    SARS-CoV-2 Antibody Dynamics in Healthcare Workers after mRNA Vaccination

    by Kevin D. Dieckhaus, Min-Jung Kim, Jian-Bing Shen, Tina S. Liang, Michael J. Kleinberg, Kristen M. Siedlarz, David B. Banach, Mark L. Metersky, Rob P. Fuller, Eric M. Mortensen, et al.

    Since the emergence of SARS-CoV-2, maintaining healthcare worker (HCW) health and safety has been fundamental to responding to the global pandemic. Vaccination with mRNA-base vaccines targeting SARS-CoV-2 spike protein has emerged as a key strategy in reducing HCW susceptibility to SARS-CoV-2, however, neutralizing antibody responses subside with time and may be influenced by many variables. We sought to understand the dynamics between vaccine products, prior clinical illness from SARS-CoV-2, and incidence of vaccine-associated adverse reactions on antibody decay over time in HCWs at a university medical center. A cohort of 296 HCWs received standard two-dose vaccination with either bnt162b2 (Pfizer/BioNTech) or mRNA-1273 (Moderna) and were evaluated after two, six, and nine months. Subjects were grouped by antibody decay curve into steep antibody decliners gentle decliners. Vaccination with mRNA-1273 led to more sustained antibody responses compared to bnt162b2. Subjects experiencing vaccine-associated symptoms were more likely to experience a more prolonged neutralizing antibody response. Subjects with clinical SARS-CoV-2 infection prior to vaccination were more likely to experience vaccination-associated symptoms after first vaccination and were more likely to have a more blunted antibody decay. Understanding factors associated with vaccine efficacy may assist clinicians in determining appropriate vaccine strategies in HCWs.