Sze-Chuan Suen
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Sze-Chuan Suen has a total of 90 co-authors. Here are the top co-authors by number of shared publications:
- Suyanpeng Zhang
University of Southern California Viterbi School of Engineering
4 shared publications Jeremy D. Goldhaber-Fiebert
Stanford Health Policy, Centers for Health Policy and Primary Care and Outcomes Research, Stanford University, Stanford, CA, USA
3 shared publications- Brian K. Chen
Department of Health Services Policy and Management, Arnold School of Public Health, University of South Carolina, Columbia, SC
2 shared publications - Citina Liang
Daniel J. Epstein Department of Industrial and Systems Engineering, University of Southern California Viterbi School of Engineering, Los Angeles, USA
2 shared publications Fernando Ordonez
2 shared publications- Han Yu
Center for Artificial Intelligence in Society Department of Business and Public Administration Center for Artificial Intelligence in Society University of Southern California University of Cyprus University of Southern California
2 shared publications
Latest Publications
- Open Access
- 13 July 2026
- Published by Springer Nature in Health Care Management Science
- Vol. 29 (3) , 30
Optimizing vaccine site locations while considering travel inconvenience and public health outcomes
During the COVID-19 pandemic, selecting vaccination sites and allocating limited doses required balancing accessibility, disease control, and fairness. We formulate a multi-objective mixed-integer linear programming model that jointly determines the locations of mega-sites and allocates vaccine doses while explicitly incorporating travel inconvenience, disease dynamics, and equitable distribution. The model incorporates commuting patterns from both residential and workplace origins to more accurately capture population mobility, and employs a tractable objective formulation that proxies key public health goals, enabling efficient and equitable mass vaccination planning. Compared with the solution empirically used in Los Angeles County in 2020, we recommend more dispersed mega-site locations that result in a 26% reduction in travel inconvenience and avert an additional 200 infections.
- 13 July 2026
- Published by Springer Nature in Health Care Management Science
- Vol. 29 (3) , 30
Open AccessOptimizing vaccine site locations while considering travel inconvenience and public health outcomes
During the COVID-19 pandemic, selecting vaccination sites and allocating limited doses required balancing accessibility, disease control, and fairness. We formulate a multi-objective mixed-integer linear programming model that jointly determines the locations of mega-sites and allocates vaccine doses while explicitly incorporating travel inconvenience, disease dynamics, and equitable distribution. The model incorporates commuting patterns from both residential and workplace origins to more accurately capture population mobility, and employs a tractable objective formulation that proxies key public health goals, enabling efficient and equitable mass vaccination planning. Compared with the solution empirically used in Los Angeles County in 2020, we recommend more dispersed mega-site locations that result in a 26% reduction in travel inconvenience and avert an additional 200 infections.
- 22 October 2025
- Published by Elsevier in Computers & Industrial Engineering
- Vol. 211
Understanding the traffic pattern impacts of COVID-19 lockdown orders
- 22 October 2025
- Published by Elsevier in Computers & Industrial Engineering
- Vol. 211
Understanding the traffic pattern impacts of COVID-19 lockdown orders
- Open Access
- 30 September 2025
- Published by Oxford University Press (OUP) in PNAS Nexus
- Vol. 4 (10)
American Life in Realtime: Benchmark, publicly available person-generated health data for equity in precision health
Person-generated health data (PGHD) from smartphones/wearables are invaluable for precision health, a field promoting health equity through tailored disease prevention, detection, and intervention strategies. However, pervasive convenience sampling in extant PGHD research introduces selection biases that systematically underrepresent disadvantaged groups, limit model generalizability, and risk exacerbating health disparities. Benchmark PGHD (representative, validated, longitudinal, and frequently repeated) are urgently needed to support model equity. To address this fieldwide limitation, we established American Life in Realtime (ALiR), a longitudinal population health study involving PGHD collected from a probability-based, nationally representative cohort using study-provided Fitbits and (as needed) 4G tablets. As a result, ALiR's 1,038 participants are broadly representative across comprehensive sociodemographic, behavioral, and health-related US population norms, overcoming disparities in established convenience samples (e.g. NIH's All of Us; AoU). Only two sources of differential enrollment remained: older age (odds ratio [OR]: 1.27, 99% CI: 1.12–1.45) during consent, lower education (OR: 0.86, 99% CI: 0.79–0.94) during enrollment, though oversampling individuals without bachelor's degrees sufficiently counterbalanced the latter. An illustrative coronavirus disease 2019 classification model—chosen for global significance, known disparities in experience and outcomes, and methodological relevance—trained using ALiR performed equivalently when tested in sample (area under the curve [AUC] = 0.84, 95% CI: 0.79–0.89) and out of sample on AoU (AUC = 0.83, 95% CI: 0.78–0.89) overall, and in historically underserved subgroups (AUC = 0.82–1.0). Conversely, an identically trained classification model using AoU underperformed by 35% out of sample on ALiR (overall AUC = 0.68, 95% CI: 0.61–0.75 vs. AUC = 0.93, 95% CI: 0.91–0.96 in sample), with worse performance in older female and non-White subgroups (by 22–40%). Our results suggest that probability sampling and hardware provisioning enabled cohort inclusivity and generalizable model performance, supporting ALiR's benchmarking potential for equitable recruitment, PGHD collection, and precision health application.
- 30 September 2025
- Published by Oxford University Press (OUP) in PNAS Nexus
- Vol. 4 (10)
Open AccessAmerican Life in Realtime: Benchmark, publicly available person-generated health data for equity in precision health
Person-generated health data (PGHD) from smartphones/wearables are invaluable for precision health, a field promoting health equity through tailored disease prevention, detection, and intervention strategies. However, pervasive convenience sampling in extant PGHD research introduces selection biases that systematically underrepresent disadvantaged groups, limit model generalizability, and risk exacerbating health disparities. Benchmark PGHD (representative, validated, longitudinal, and frequently repeated) are urgently needed to support model equity. To address this fieldwide limitation, we established American Life in Realtime (ALiR), a longitudinal population health study involving PGHD collected from a probability-based, nationally representative cohort using study-provided Fitbits and (as needed) 4G tablets. As a result, ALiR's 1,038 participants are broadly representative across comprehensive sociodemographic, behavioral, and health-related US population norms, overcoming disparities in established convenience samples (e.g. NIH's All of Us; AoU). Only two sources of differential enrollment remained: older age (odds ratio [OR]: 1.27, 99% CI: 1.12–1.45) during consent, lower education (OR: 0.86, 99% CI: 0.79–0.94) during enrollment, though oversampling individuals without bachelor's degrees sufficiently counterbalanced the latter. An illustrative coronavirus disease 2019 classification model—chosen for global significance, known disparities in experience and outcomes, and methodological relevance—trained using ALiR performed equivalently when tested in sample (area under the curve [AUC] = 0.84, 95% CI: 0.79–0.89) and out of sample on AoU (AUC = 0.83, 95% CI: 0.78–0.89) overall, and in historically underserved subgroups (AUC = 0.82–1.0). Conversely, an identically trained classification model using AoU underperformed by 35% out of sample on ALiR (overall AUC = 0.68, 95% CI: 0.61–0.75 vs. AUC = 0.93, 95% CI: 0.91–0.96 in sample), with worse performance in older female and non-White subgroups (by 22–40%). Our results suggest that probability sampling and hardware provisioning enabled cohort inclusivity and generalizable model performance, supporting ALiR's benchmarking potential for equitable recruitment, PGHD collection, and precision health application.
- 17 February 2025
- Published by Elsevier in Resuscitation
- Vol. 209, 110552
Cost-effectiveness of drone-delivered automated external defibrillators for cardiac arrest
- 17 February 2025
- Published by Elsevier in Resuscitation
- Vol. 209, 110552
Cost-effectiveness of drone-delivered automated external defibrillators for cardiac arrest
- Open Access
- 17 July 2024
- Published by Oxford University Press (OUP) in Open Forum Infectious Diseases
- Vol. 11 (Supplement) , S137-S145
A Microsimulation Model of Mpox in Los Angeles County: Implications for Future Disease Prevention and Control Strategies among Men Who Have Sex with Men
Background: The 2022 monkeypox (mpox) outbreak in Los Angeles County (LAC) emphasized the need to prepare for emergent infectious disease outbreaks. Vaccination and promotion of sexual risk reduction practices appeared successful in LAC, as mpox cases declined starting in August. Nonetheless, questions persisted regarding the effectiveness of targeting vaccinations and the role of sexual risk reduction in reducing mpox cases. Methods: We collaborated with the LAC Department of Public Health to develop a microsimulation for men who have sex with men (MSM). This model tracked mpox dynamics by age, race/ethnicity, and HIV status and was calibrated and validated against surveillance data. We simulated counterfactual scenarios to understand the effects of variation in vaccination rates, timing of vaccination rollout, vaccine allocation, and sexual contact rates. Results: In the simulation, doubling the vaccination rate reduced cumulative cases over a 40-week time horizon by 13% but would necessitate 88 995 additional doses. Initiating vaccination 2 weeks earlier decreased cases by 11%, while an 8-week delay yielded a 20% increase in cases. A 3-week earlier decrease in sexual contact rates reduced cumulative cases by 60%, while a 3-week delay resulted in a 95% increase. Prioritizing people with HIV (PWH) for vaccination reduced cumulative cases, while allocating vaccines to a single racial/ethnic group was not effective. Conclusions: Our study highlights the significance of policies to support timely vaccination and sexual partnership reduction to address mpox outbreaks among MSM. These findings also underscore the need to target vulnerable risk groups, such as PWH.
- 17 July 2024
- Published by Oxford University Press (OUP) in Open Forum Infectious Diseases
- Vol. 11 (Supplement) , S137-S145
Open AccessA Microsimulation Model of Mpox in Los Angeles County: Implications for Future Disease Prevention and Control Strategies among Men Who Have Sex with Men
Background: The 2022 monkeypox (mpox) outbreak in Los Angeles County (LAC) emphasized the need to prepare for emergent infectious disease outbreaks. Vaccination and promotion of sexual risk reduction practices appeared successful in LAC, as mpox cases declined starting in August. Nonetheless, questions persisted regarding the effectiveness of targeting vaccinations and the role of sexual risk reduction in reducing mpox cases. Methods: We collaborated with the LAC Department of Public Health to develop a microsimulation for men who have sex with men (MSM). This model tracked mpox dynamics by age, race/ethnicity, and HIV status and was calibrated and validated against surveillance data. We simulated counterfactual scenarios to understand the effects of variation in vaccination rates, timing of vaccination rollout, vaccine allocation, and sexual contact rates. Results: In the simulation, doubling the vaccination rate reduced cumulative cases over a 40-week time horizon by 13% but would necessitate 88 995 additional doses. Initiating vaccination 2 weeks earlier decreased cases by 11%, while an 8-week delay yielded a 20% increase in cases. A 3-week earlier decrease in sexual contact rates reduced cumulative cases by 60%, while a 3-week delay resulted in a 95% increase. Prioritizing people with HIV (PWH) for vaccination reduced cumulative cases, while allocating vaccines to a single racial/ethnic group was not effective. Conclusions: Our study highlights the significance of policies to support timely vaccination and sexual partnership reduction to address mpox outbreaks among MSM. These findings also underscore the need to target vulnerable risk groups, such as PWH.
- Open Access
- 5 June 2024
- Published by Springer Nature in Health Care Management Science
- Vol. 27 (3) , 391-414
Examining chronic kidney disease screening frequency among diabetics: a POMDP approach
Forty percent of diabetics will develop chronic kidney disease (CKD) in their lifetimes. However, as many as 50% of these CKD cases may go undiagnosed. We developed screening recommendations stratified by age and previous test history for individuals with diagnosed diabetes and unknown proteinuria status by race and gender groups. To do this, we used a Partially Observed Markov Decision Process (POMDP) to identify whether a patient should be screened at every three-month interval from ages 30-85. Model inputs were drawn from nationally-representative datasets, the medical literature, and a microsimulation that integrates this information into group-specific disease progression rates. We implement the POMDP solution policy in the microsimulation to understand how this policy may impact health outcomes and generate an easily-implementable, non-belief-based approximate policy for easier clinical interpretability. We found that the status quo policy, which is to screen annually for all ages and races, is suboptimal for maximizing expected discounted future net monetary benefits (NMB). The POMDP policy suggests more frequent screening after age 40 in all race and gender groups, with screenings 2-4 times a year for ages 61-70. Black individuals are recommended for screening more frequently than their White counterparts. This policy would increase NMB from the status quo policy between $1,000 to $8,000 per diabetic patient at a willingness-to-pay of $150,000 per quality-adjusted life year (QALY).
- 5 June 2024
- Published by Springer Nature in Health Care Management Science
- Vol. 27 (3) , 391-414
Open AccessExamining chronic kidney disease screening frequency among diabetics: a POMDP approach
Forty percent of diabetics will develop chronic kidney disease (CKD) in their lifetimes. However, as many as 50% of these CKD cases may go undiagnosed. We developed screening recommendations stratified by age and previous test history for individuals with diagnosed diabetes and unknown proteinuria status by race and gender groups. To do this, we used a Partially Observed Markov Decision Process (POMDP) to identify whether a patient should be screened at every three-month interval from ages 30-85. Model inputs were drawn from nationally-representative datasets, the medical literature, and a microsimulation that integrates this information into group-specific disease progression rates. We implement the POMDP solution policy in the microsimulation to understand how this policy may impact health outcomes and generate an easily-implementable, non-belief-based approximate policy for easier clinical interpretability. We found that the status quo policy, which is to screen annually for all ages and races, is suboptimal for maximizing expected discounted future net monetary benefits (NMB). The POMDP policy suggests more frequent screening after age 40 in all race and gender groups, with screenings 2-4 times a year for ages 61-70. Black individuals are recommended for screening more frequently than their White counterparts. This policy would increase NMB from the status quo policy between $1,000 to $8,000 per diabetic patient at a willingness-to-pay of $150,000 per quality-adjusted life year (QALY).
- 10 April 2024
- Published by Taylor & Francis in IISE Transactions
- Vol. 57 (5) , 469-483
Quantifying the benefits of increasing decision-making frequency for health applications with regular decision epochs
- 10 April 2024
- Published by Taylor & Francis in IISE Transactions
- Vol. 57 (5) , 469-483
Quantifying the benefits of increasing decision-making frequency for health applications with regular decision epochs
- Open Access
Extending Dynamic Origin-Destination Estimation to Understand Traffic Patterns During COVID-19
Estimating dynamic Origin-Destination (OD) traffic flow is crucial for understanding traffic patterns and the traffic network. While dynamic origin-destination estimation (DODE) has been studied for decades as a useful tool for estimating traffic flow, few existing models have considered its potential in evaluating the influence of policy on travel activity. This paper proposes a data-driven approach to estimate OD traffic flow using sensor data on highways and local roads. We extend prior DODE models to improve accuracy and realism in order to estimate how policies affect OD traffic flow in large urban networks. We applied our approach to a case study in Los Angeles County, where we developed a traffic network, estimated OD traffic flow between health districts during COVID-19, and analyzed the relationship between OD traffic flow and demographic characteristics such as income. Our findings demonstrate that the proposed approach provides valuable insights into traffic flow patterns and their underlying demographic factors for a large-scale traffic network. Specifically, our approach allows for evaluating the impact of policy changes on travel activity. The approach has practical applications for transportation planning and traffic management, enabling a better understanding of traffic flow patterns and the impact of policy changes on travel activity.
Open AccessExtending Dynamic Origin-Destination Estimation to Understand Traffic Patterns During COVID-19
Estimating dynamic Origin-Destination (OD) traffic flow is crucial for understanding traffic patterns and the traffic network. While dynamic origin-destination estimation (DODE) has been studied for decades as a useful tool for estimating traffic flow, few existing models have considered its potential in evaluating the influence of policy on travel activity. This paper proposes a data-driven approach to estimate OD traffic flow using sensor data on highways and local roads. We extend prior DODE models to improve accuracy and realism in order to estimate how policies affect OD traffic flow in large urban networks. We applied our approach to a case study in Los Angeles County, where we developed a traffic network, estimated OD traffic flow between health districts during COVID-19, and analyzed the relationship between OD traffic flow and demographic characteristics such as income. Our findings demonstrate that the proposed approach provides valuable insights into traffic flow patterns and their underlying demographic factors for a large-scale traffic network. Specifically, our approach allows for evaluating the impact of policy changes on travel activity. The approach has practical applications for transportation planning and traffic management, enabling a better understanding of traffic flow patterns and the impact of policy changes on travel activity.
- Open Access
RIFLE: Imputation and Robust Inference from Low Order Marginals
The ubiquity of missing values in real-world datasets poses a challenge for statistical inference and can prevent similar datasets from being analyzed in the same study, precluding many existing datasets from being used for new analyses. While an extensive collection of packages and algorithms have been developed for data imputation, the overwhelming majority perform poorly if there are many missing values and low sample sizes, which are unfortunately common characteristics in empirical data. Such low-accuracy estimations adversely affect the performance of downstream statistical models. We develop a statistical inference framework for regression and classification in the presence of missing data without imputation. Our framework, RIFLE (Robust InFerence via Low-order moment Estimations), estimates low-order moments of the underlying data distribution with corresponding confidence intervals to learn a distributionally robust model. We specialize our framework to linear regression and normal discriminant analysis, and we provide convergence and performance guarantees. This framework can also be adapted to impute missing data. In numerical experiments, we compare RIFLE to several state-of-the-art approaches (including MICE, Amelia, MissForest, KNN-imputer, MIDA, and Mean Imputer) for imputation and inference in the presence of missing values. Our experiments demonstrate that RIFLE outperforms other benchmark algorithms when the percentage of missing values is high and/or when the number of data points is relatively small. RIFLE is publicly available at https://github.com/optimization-for-data-driven-science/RIFLE.
Open AccessRIFLE: Imputation and Robust Inference from Low Order Marginals
The ubiquity of missing values in real-world datasets poses a challenge for statistical inference and can prevent similar datasets from being analyzed in the same study, precluding many existing datasets from being used for new analyses. While an extensive collection of packages and algorithms have been developed for data imputation, the overwhelming majority perform poorly if there are many missing values and low sample sizes, which are unfortunately common characteristics in empirical data. Such low-accuracy estimations adversely affect the performance of downstream statistical models. We develop a statistical inference framework for regression and classification in the presence of missing data without imputation. Our framework, RIFLE (Robust InFerence via Low-order moment Estimations), estimates low-order moments of the underlying data distribution with corresponding confidence intervals to learn a distributionally robust model. We specialize our framework to linear regression and normal discriminant analysis, and we provide convergence and performance guarantees. This framework can also be adapted to impute missing data. In numerical experiments, we compare RIFLE to several state-of-the-art approaches (including MICE, Amelia, MissForest, KNN-imputer, MIDA, and Mean Imputer) for imputation and inference in the presence of missing values. Our experiments demonstrate that RIFLE outperforms other benchmark algorithms when the percentage of missing values is high and/or when the number of data points is relatively small. RIFLE is publicly available at https://github.com/optimization-for-data-driven-science/RIFLE.
- Open Access
- 30 June 2023
- Published by Wiley in World Journal of Surgery
- Vol. 47 (10) , 2340-2346
Assessment of Machine Learning Methods to Predict Massive Blood Transfusion in Trauma
Background: Accurately predicting which patients are most likely to benefit from massive transfusion protocol (MTP) activation may help patients while saving blood products and limiting cost. The purpose of this study is to explore the use of modern machine learning (ML) methods to develop and validate a model that can accurately predict the need for massive blood transfusion (MBT).Methods: The institutional trauma registry was used to identify all trauma team activation cases between June 2015 and August 2019. We used an ML framework to explore multiple ML methods including logistic regression with forward and backward selection, logistic regression with lasso and ridge regularization, support vector machines (SVM), decision tree, random forest, naive Bayes, XGBoost, AdaBoost, and neural networks. Each model was then assessed using sensitivity, specificity, positive predictive value, and negative predictive value. Model performance was compared to that of existing scores including the Assessment of Blood Consumption (ABC) and the Revised Assessment of Bleeding and Transfusion (RABT).Results: A total of 2438 patients were included in the study, with 4.9% receiving MBT. All models besides decision tree and SVM attained an area under the curve (AUC) of above 0.75 (range: 0.75–0.83). Most of the ML models have higher sensitivity (0.55–0.83) than the ABC and RABT score (0.36 and 0.55, respectively) while maintaining comparable specificity (0.75–0.81; ABC 0.80 and RABT 0.83).Conclusions: Our ML models performed better than existing scores. Implementing an ML model in mobile computing devices or electronic health record has the potential to improve the usability.
- 30 June 2023
- Published by Wiley in World Journal of Surgery
- Vol. 47 (10) , 2340-2346
Open AccessAssessment of Machine Learning Methods to Predict Massive Blood Transfusion in Trauma
Background: Accurately predicting which patients are most likely to benefit from massive transfusion protocol (MTP) activation may help patients while saving blood products and limiting cost. The purpose of this study is to explore the use of modern machine learning (ML) methods to develop and validate a model that can accurately predict the need for massive blood transfusion (MBT).Methods: The institutional trauma registry was used to identify all trauma team activation cases between June 2015 and August 2019. We used an ML framework to explore multiple ML methods including logistic regression with forward and backward selection, logistic regression with lasso and ridge regularization, support vector machines (SVM), decision tree, random forest, naive Bayes, XGBoost, AdaBoost, and neural networks. Each model was then assessed using sensitivity, specificity, positive predictive value, and negative predictive value. Model performance was compared to that of existing scores including the Assessment of Blood Consumption (ABC) and the Revised Assessment of Bleeding and Transfusion (RABT).Results: A total of 2438 patients were included in the study, with 4.9% receiving MBT. All models besides decision tree and SVM attained an area under the curve (AUC) of above 0.75 (range: 0.75–0.83). Most of the ML models have higher sensitivity (0.55–0.83) than the ABC and RABT score (0.36 and 0.55, respectively) while maintaining comparable specificity (0.75–0.81; ABC 0.80 and RABT 0.83).Conclusions: Our ML models performed better than existing scores. Implementing an ML model in mobile computing devices or electronic health record has the potential to improve the usability.