Tina Yi-Jin Hsieh

Tina Yi-Jin Hsieh

Auto-generated profile, based on publicly available information

Tina Yi-Jin Hsieh

Beth Israel Deaconess Medical Center

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 September 2024 in Annals of Internal Medicine
Verified
Journal Article
Closed access
YYYee Hui Yeo
Yee Hui Yeo
ARAli Rezaie
Ali Rezaie
THTina Yi-Jin Hsieh
Tina Yi-Jin Hsieh
XHXiaoqin Hu
Xiaoqin Hu
SGSrinivas Gaddam
Srinivas Gaddam
RWRabindra Watson
Rabindra Watson


16 November 2022 in American Journal of Perinatology
Verified
Journal Article
Closed access
PDF file
JSJocelyn Alayne Stairs
Jocelyn Alayne Stairs
THTina Yi Jin Hsieh
Tina Yi Jin Hsieh
DRDaniel Rolnik
Daniel Rolnik
Objective Our objective was to estimate the association between in vitro fertilization (IVF) pregnancy and adverse pregnancy outcomes during delivery hospital admission in a contemporary, nation-wide cohort of births in the United States. Study Design This retrospective, population-based cohort study used the National Inpatient Sample database to identify patients discharged from the hospital following delivery from 2014 to 2019. IVF pregnancies were identified using the International Classification of Disease-Revision 9/10 codes. Crude and adjusted odds ratios of preterm birth and other clinically significant adverse pregnancy outcomes were evaluated using multivariable logistic regression models. Trends in preterm birth and multiple pregnancy were estimated over the study period. The contribution of multiple pregnancy to preterm birth in IVF pregnancy was estimated in a mediation analysis. Results Among 4,451,667 delivery-related discharges, IVF pregnancies were associated with 3.25 times the odds of preterm birth (95% confidence interval [CI]: 3.05–3.46, p < 0.001) compared with non-IVF pregnancy deliveries. Odds of preterm birth in IVF pregnancy delivery discharges decreased over the study period (p-value for linear trend = 0.009). The proportion of multiple pregnancies decreased in IVF pregnancy delivery discharges but remained stable in non-IVF pregnancy deliveries. The proportion of the adjusted effect of IVF pregnancy on preterm birth mediated through multiple pregnancy was 67.6% (95% CI: 62.6–72.7%). Conclusion While the odds of adverse pregnancy outcomes are increased compared with non-IVF pregnancies, the odds of preterm birth and multiple gestation have decreased among IVF pregnancies in the United States. Key Points

18 December 2025 in medRxiv
Verified
Preprint
Open access
PDF file
DTDiego Trujillo
Diego Trujillo
DWDulin Wang
Dulin Wang
NBNathan Bahr
Nathan Bahr
THTina Yi-Jin Hsieh
Tina Yi-Jin Hsieh
BCByeongyeon Cho
Byeongyeon Cho
JGJeanne-Marie Guise
Jeanne-Marie Guise
Large language models (LLMs) have shown incredible promise in medicine. While LLMs may be particularly useful in areas requiring extensive review of clinical records, their use remains limited due to their tendency to hallucinate and fabricate information. Hallucination issues, as well as their consequences, are exacerbated in low–probability, high–stakes scenarios such as rare adverse safety events or medical errors. We present SAFE–AI (Structured and Automated Framework for Explainable AI), a novel method for clinical decision making that combines the strengths of clinical expert knowledge with LLMs in an ontology–driven model that minimizes hallucinations using strict rules. We test this method to identify medication errors in medical charts. We collected a sample of 18,402 lines of clinical information from 300 EMS clinical charts that were independently dually reviewed by two expert physicians for epinephrine adverse safety events (ASEs), with 96% inter-rater agreement. We tested SAFE–AI against these labels, achieving human–like performance in detecting epinephrine overdoses with 97.9% accuracy, and 91.6% accuracy in identifying delays in epinephrine administration, greatly outperforming baseline LLMs models. Notably, some disagreements between clinicians and the model were found to be justifiable differences in judgment rather than errors. SAFE-AI presents a novel approach for clinical AI applications that addresses two key limitations of current machine learning methods: 1) over-reliance on probabilistic pattern recognition instead of established medical knowledge, and 2) perpetuation of biases present in training data. This framework is easily adaptable to a range of clinical applications, paving the way for provable and trustworthy AI in medicine. Author Summary: LLMs have shown promise in analyzing clinical records but their use is limited due to their tendency to hallucinate and fabricate information. Misinformation could threaten patient safety and jeopardize trust. We developed SAFE–AI (Structured and Automated Framework for Explainable AI), which combines knowledge from clinical experts with LLM inference to detect adverse safety events (ASEs) with minimal errors. We tested our method in identifying medication errors in medical charts and compared results to reviews by expert physicians. Our method detected epinephrine delays and overdoses with a high level of accuracy. SAFE-AI presents a novel approach for clinical AI applications that overcomes reliance on pattern recognition instead of medical knowledge biases present in training data.

9 December 2025 in JCO Oncology Practice
Verified
Journal Article
Closed access
RYRen-Hua Ye
Ren-Hua Ye
HJHong-Jie Jhou
Hong-Jie Jhou
HCHsin-Yu Chen
Hsin-Yu Chen
Li-Ting KaoLi-Ting Kao
Li-Ting Kao
THTina Yi-Jin Hsieh
Tina Yi-Jin Hsieh
CHO-HAO LEECHO-HAO LEE
CHO-HAO LEE
PURPOSE: Although clinical trials support osimertinib's efficacy over standard epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) in metastatic EGFR-mutant non–small cell lung cancer (NSCLC), large real-world comparisons against first-generation TKIs are limited. We compared the real-world effectiveness and safety of first-line osimertinib versus first-generation TKIs (erlotinib, gefitinib) using a target trial emulation approach. METHODS: This retrospective cohort study used the TriNetX Research Network global database. We identified adults with newly diagnosed metastatic EGFR-mutant NSCLC initiating first-line EGFR-TKI therapy (January 2014-March 2025). After 1:1 propensity score matching (PSM), overall survival (OS; primary outcome) and safety were compared between treatment arms. RESULTS: Following PSM, a total of 3,168 patients (1,584 per arm) were analyzed. The median OS was significantly longer with osimertinib compared with first-generation TKIs (1,505 v 901 days; hazard ratio [HR], 0.612 [95% CI, 0.549 to 0.683]; P < .001). Survival benefits were consistent across subgroups, particularly pronounced in Asian patients (HR, 0.439). First-generation TKIs had higher rates of dermatologic toxicities, severe infections, and hospitalizations. Osimertinib showed higher rates of abnormal ECG findings and thrombocytopenia. Sensitivity analyses confirmed the robustness of the primary findings. CONCLUSION: In this large real-world comparative effectiveness study using target trial emulation, first-line osimertinib was associated with substantially prolonged OS and a favorable safety profile, including lower rates of severe infections and hospitalizations, compared with first-generation EGFR-TKIs in patients with metastatic EGFR-mutant NSCLC.

2 September 2026 in Acta Paediatrica
Verified
Journal Article
Closed access
PDF file
ENEleanor Nicholson
Eleanor Nicholson
RBRitobhas Bhowmik
Ritobhas Bhowmik
ESEmily Stonestreet
Emily Stonestreet
GNGanad Neka
Ganad Neka
AMAudrey Mutoni
Audrey Mutoni
ACAi‐Ris Y. Collier
Ai‐Ris Y. Collier