2024
Validation of an Electronic Health Record–Based Machine Learning Model Compared With Clinical Risk Scores for Gastrointestinal Bleeding
Shung D, Chan C, You K, Nakamura S, Saarinen T, Zheng N, Simonov M, Li D, Tsay C, Kawamura Y, Shen M, Hsiao A, Sekhon J, Laine L. Validation of an Electronic Health Record–Based Machine Learning Model Compared With Clinical Risk Scores for Gastrointestinal Bleeding. Gastroenterology 2024, 167: 1198-1212. PMID: 38971198, PMCID: PMC11493512, DOI: 10.1053/j.gastro.2024.06.030.Peer-Reviewed Original ResearchElectronic health recordsGlasgow-Blatchford scoreEmergency departmentVery-low-risk patientsRisk scoreOakland scoreMachine learning modelsStructured data fieldsClinical risk scoreGastrointestinal bleedingAll-cause mortalityHealth recordsLearning modelsManual data entrySecondary analysisRisk stratification scoresAssess proportionRed blood-cell transfusionPrimary outcomeProportion of patientsData entryOvert gastrointestinal bleedingPrimary analysisReceiver-operating-characteristic curveVery-low-riskKamino: A Scalable Architecture to Support Medical AI Research Using Large Real World Data
Lin F, Young P, He H, Huang J, Gagne R, Rice D, Price N, Byron W, Hu Y, Felker D, Button W, Meeker D, Hsiao A, Xu H, Torre C, Schulz W. Kamino: A Scalable Architecture to Support Medical AI Research Using Large Real World Data. 2024, 00: 500-504. DOI: 10.1109/ichi61247.2024.00072.Peer-Reviewed Original ResearchElectronic health recordsAI researchNatural language processing tasksElectronic health record dataLanguage processing tasksComputing resource managementLarge-scale data retrievalMedical AI researchLeveraging electronic health recordsStandard data modelKubernetes orchestratorScalable architectureProcessing tasksResource allocation systemsSecurity considerationsAccess managementData retrievalData modelArchitectural solutionsOMOP CDMReal World DataWorld DataHealth recordsOMOPData
2023
Mo1365 IDENTIFYING PATIENTS WITH ACUTE GASTROINTESTINAL BLEEDING USING NOTE TEXT IN THE ELECTRONIC HEALTH RECORD: A HYBRID NATURAL LANGUAGE PROCESSING AND DEEP LEARNING APPROACH
Dupont J, Zheng N, Laine L, Hsiao A, Thomas P, Partridge C, Shung D. Mo1365 IDENTIFYING PATIENTS WITH ACUTE GASTROINTESTINAL BLEEDING USING NOTE TEXT IN THE ELECTRONIC HEALTH RECORD: A HYBRID NATURAL LANGUAGE PROCESSING AND DEEP LEARNING APPROACH. Gastroenterology 2023, 164: s-838. DOI: 10.1016/s0016-5085(23)02935-9.Peer-Reviewed Original ResearchReal-World Observational Evaluation of Common Interventions to Reduce Emergency Department Prescribing of Opioid Medications
Sangal R, Rothenberg C, Hawk K, D'Onofrio G, Hsiao A, Solad Y, Venkatesh A. Real-World Observational Evaluation of Common Interventions to Reduce Emergency Department Prescribing of Opioid Medications. The Joint Commission Journal On Quality And Patient Safety 2023, 49: 239-246. PMID: 36914528, DOI: 10.1016/j.jcjq.2023.01.013.Peer-Reviewed Original ResearchConceptsOpioid prescribingED visitsElectronic health recordsOpioid prescriptionsEmergency department opioid prescriptionsAnalgesia prescriptionOpioid stewardshipOpioid medicationsSecondary outcomesPrimary outcomePreintervention periodInterruptive alertsCommon interventionPrescribingAlert fatigueElectronic prescribingPrevious interventionsHospital systemObservational evaluationHealth recordsVisitsStewardship policiesInterventionOutcomesPrescription
2022
Alerting Clinicians to 1-Year Mortality Risk in Patients Hospitalized With Heart Failure
Ahmad T, Desai NR, Yamamoto Y, Biswas A, Ghazi L, Martin M, Simonov M, Dhar R, Hsiao A, Kashyap N, Allen L, Velazquez EJ, Wilson FP. Alerting Clinicians to 1-Year Mortality Risk in Patients Hospitalized With Heart Failure. JAMA Cardiology 2022, 7: 905-912. PMID: 35947362, PMCID: PMC9366654, DOI: 10.1001/jamacardio.2022.2496.Peer-Reviewed Original ResearchConceptsUsual care groupElectronic health recordsHeart failureAlert groupMedian agePrimary outcomeFemale patientsTerminal pro-brain natriuretic peptide levelsYale New Haven Health SystemPro-brain natriuretic peptide levelsMedian NT-proBNP levelLeft ventricular ejection fractionLarge tertiary care centerHeart failure hospitalizationHeart failure medicationsNatriuretic peptide levelsNT-proBNP levelsHealth recordsHours of admissionTertiary care centerVentricular ejection fractionRate of hospitalizationIntensive care unitEthnicity groupsSmall community hospital
2021
REVeAL-HF Design and Rationale of a Pragmatic Randomized Controlled Trial Embedded Within Routine Clinical Practice
Ahmad T, Yamamoto Y, Biswas A, Ghazi L, Martin M, Simonov M, Hsiao A, Kashyap N, Velazquez EJ, Desai NR, Wilson FP. REVeAL-HF Design and Rationale of a Pragmatic Randomized Controlled Trial Embedded Within Routine Clinical Practice. JACC Heart Failure 2021, 9: 409-419. PMID: 33992566, DOI: 10.1016/j.jchf.2021.03.006.Peer-Reviewed Original ResearchConceptsHeart failurePatient outcomesClinical decision makingN-terminal pro-B-type natriuretic peptide levelsPro-B-type natriuretic peptide levelsPragmatic Randomized Controlled TrialIntravenous diuretic agentsNatriuretic peptide levelsRandomized Controlled TrialsRoutine clinical practiceEvidence-based interventionsElectronic health recordsControlled TrialsPatient populationAccurate prognosticationCommon causeDiuretic agentsPeptide levelsClinical practicePg/TrialsHealth recordsSignificant riskMortalityOutcomesEarly identification of patients with acute gastrointestinal bleeding using natural language processing and decision rules
Shung D, Tsay C, Laine L, Chang D, Li F, Thomas P, Partridge C, Simonov M, Hsiao A, Tay JK, Taylor A. Early identification of patients with acute gastrointestinal bleeding using natural language processing and decision rules. Journal Of Gastroenterology And Hepatology 2021, 36: 1590-1597. PMID: 33105045, DOI: 10.1111/jgh.15313.Peer-Reviewed Original ResearchConceptsNatural language processingElectronic health recordsLanguage processingNLP algorithmSystematized NomenclatureReal timeAcute gastrointestinal bleedingBidirectional Encoder RepresentationsDecision rulesEHR-based phenotyping algorithmsGastrointestinal bleedingRisk stratification scoresEncoder RepresentationsData elementsPhenotyping algorithmStratification scoresHealth recordsAlgorithmPhenotyping of patientsEmergency department patientsTime of presentationRisk stratification modelED reviewDeploymentExternal validation