Eva Tseng
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
Research Subjects
Research Interests
Eva Tseng has a total of 90 co-authors. Here are the top co-authors by number of shared publications:
Nisa M. Maruthur
Division of General Internal Medicine, Johns Hopkins University School of Medicine 1 , Baltimore, MD
16 shared publicationsJeanne M. Clark
Department of Medicine, Rutgers Robert Wood Johnson Medical School 4 , New Brunswick, NJ
8 shared publicationsJill A. Marsteller
Welch Center for Prevention, Epidemiology, and Clinical Research, Johns Hopkins University 2 , Baltimore, MD
6 shared publicationsHsin-Chieh Yeh
From the Division of General Internal Medicine, Johns Hopkins University School of Medicine, Baltimore, MD (E.T., L.J.A., H.-C.Y., S.J.P., E.R.M., N.M.M.)
4 shared publications- Nae-Yuh Wang
Departments of Medicine, Biostatistics, and Epidemiology, Johns Hopkins University, Baltimore, MD (N.-Y.W.).
4 shared publications - Hsin-Chieh Yeh
Department of Medicine, Johns Hopkins University 1 , Baltimore, MD
3 shared publications
Latest Publications
- 1 March 2026
- Published by American Diabetes Association in Diabetes Obesity and Cardiometabolic CARE
- Vol. 1  (2) , 230-238
START Diabetes Prevention: A Multilevel Strategy for Primary Care Clinics
OBJECTIVE: Prediabetes affects more than one in three adults in the U.S.; however, fewer than 5% receive treatment. Most prior interventions focused on individual-level barriers, neglecting the clinician- and system-level barriers. This study aimed to evaluate the effectiveness of a multilevel intervention in increasing prediabetes treatment uptake in primary care. RESEARCH DESIGN AND METHODS: We conducted a pragmatic study of the Screen, Test, Act, Refer, and Treat (START) Diabetes Prevention intervention, which included a structured workflow, shared decision-making resources, and clinical decision support tools within electronic health records. The intervention was implemented over 12 months in a suburban primary care clinic near Baltimore, MD, and compared with a control clinic in the same area. The primary outcome measured was the uptake of prediabetes treatment, defined as referral to the Diabetes Prevention Program and/or metformin prescription within 30 days of a primary care visit. RESULTS: Both clinics saw increased use of preventive treatment among patients with prediabetes, with no significant differences between the two. However, the intervention clinic had a higher percentage of patients reporting they had discussions about prediabetes with their primary care clinician (PCP; 60% vs. 48%; P = 0.002), felt informed, and believed that their opinion was valued. PCPs reported increased acceptability and perceived usefulness of the START Diabetes Prevention strategy for engaging patients with prediabetes about treatment by study end. CONCLUSIONS: The low-touch multilevel intervention resulted in a modest increase in preventive treatment and improved patient-centered care by increasing the number of discussions and enhancing patients’ understanding of prediabetes.
- 1 March 2026
- Published by American Diabetes Association in Diabetes Obesity and Cardiometabolic CARE
- Vol. 1  (2) , 230-238
START Diabetes Prevention: A Multilevel Strategy for Primary Care Clinics
OBJECTIVE: Prediabetes affects more than one in three adults in the U.S.; however, fewer than 5% receive treatment. Most prior interventions focused on individual-level barriers, neglecting the clinician- and system-level barriers. This study aimed to evaluate the effectiveness of a multilevel intervention in increasing prediabetes treatment uptake in primary care. RESEARCH DESIGN AND METHODS: We conducted a pragmatic study of the Screen, Test, Act, Refer, and Treat (START) Diabetes Prevention intervention, which included a structured workflow, shared decision-making resources, and clinical decision support tools within electronic health records. The intervention was implemented over 12 months in a suburban primary care clinic near Baltimore, MD, and compared with a control clinic in the same area. The primary outcome measured was the uptake of prediabetes treatment, defined as referral to the Diabetes Prevention Program and/or metformin prescription within 30 days of a primary care visit. RESULTS: Both clinics saw increased use of preventive treatment among patients with prediabetes, with no significant differences between the two. However, the intervention clinic had a higher percentage of patients reporting they had discussions about prediabetes with their primary care clinician (PCP; 60% vs. 48%; P = 0.002), felt informed, and believed that their opinion was valued. PCPs reported increased acceptability and perceived usefulness of the START Diabetes Prevention strategy for engaging patients with prediabetes about treatment by study end. CONCLUSIONS: The low-touch multilevel intervention resulted in a modest increase in preventive treatment and improved patient-centered care by increasing the number of discussions and enhancing patients’ understanding of prediabetes.
- 10 October 2025
- Published by American Diabetes Association in Clinical Diabetes
- Vol. 43  (5) , 757-771
Scaling the Diabetes Prevention Program and Diabetes Self-Management Education and Support Services Across a Health System
This article describes and evaluates a multipronged strategy to expand use of the Diabetes Prevention Program (DPP) and diabetes self-management education and support (DSMES) services across a large academic health system. Strategies included streamlining referral processes, embedding diabetes educators in primary care, leveraging consistent leadership messaging, and building partnerships with community organizations. These interventions led to substantial increases in referrals, with 10.8% of 4,397 patients enrolling in DPP and 35.9% of 10,210 enrolling in DSMES. Coordinated, multilevel interventions can effectively increase engagement in these underutilized, evidence-based diabetes care programs.
- 10 October 2025
- Published by American Diabetes Association in Clinical Diabetes
- Vol. 43  (5) , 757-771
Scaling the Diabetes Prevention Program and Diabetes Self-Management Education and Support Services Across a Health System
This article describes and evaluates a multipronged strategy to expand use of the Diabetes Prevention Program (DPP) and diabetes self-management education and support (DSMES) services across a large academic health system. Strategies included streamlining referral processes, embedding diabetes educators in primary care, leveraging consistent leadership messaging, and building partnerships with community organizations. These interventions led to substantial increases in referrals, with 10.8% of 4,397 patients enrolling in DPP and 35.9% of 10,210 enrolling in DSMES. Coordinated, multilevel interventions can effectively increase engagement in these underutilized, evidence-based diabetes care programs.
- Open Access
- 15 October 2024
- Published by Elsevier in AJPM Focus
- Vol. 3  (6) , 100287
Evaluation of a Diabetes Screening Clinical Decision Support Tool
- 15 October 2024
- Published by Elsevier in AJPM Focus
- Vol. 3  (6) , 100287
Open AccessEvaluation of a Diabetes Screening Clinical Decision Support Tool
- Open Access
START Diabetes Prevention: A Multi-Level Strategy for Primary Care Clinics
Background Prediabetes, a high-risk state for developing diabetes, affects more than 1 in 3 adults nationally. However, Objective To measure the effectiveness of a multi-level intervention on uptake of prediabetes treatment in a primary care clinic. Design Pragmatic study of the START (Screen, Test, Act, Refer and Treat) Diabetes Prevention intervention. Participants The START Diabetes Prevention intervention was implemented in a suburban primary care clinic outside of Baltimore compared to a control clinic in the same area over a 12-month period. Intervention START Diabetes Prevention intervention included a structured workflow, shared decision-making resources and electronic health record clinical decision support tools. Main Measures Uptake of prediabetes treatment, defined as Diabetes Prevention Program referral, metformin prescription and/or medical nutrition referral within 30 days of any PCC visit. Key Results We demonstrated greater uptake of preventive treatment among patients with prediabetes in the intervention clinic vs. control clinic receiving usual care (11.6% vs. 6.7%, pConclusions A low-touch multi-level intervention is effective in increasing prediabetes treatment uptake. The intervention was also acceptable and feasible for clinicians, and enhanced patient understanding and discussions of prediabetes with their clinicians.
Open AccessSTART Diabetes Prevention: A Multi-Level Strategy for Primary Care Clinics
Background Prediabetes, a high-risk state for developing diabetes, affects more than 1 in 3 adults nationally. However, Objective To measure the effectiveness of a multi-level intervention on uptake of prediabetes treatment in a primary care clinic. Design Pragmatic study of the START (Screen, Test, Act, Refer and Treat) Diabetes Prevention intervention. Participants The START Diabetes Prevention intervention was implemented in a suburban primary care clinic outside of Baltimore compared to a control clinic in the same area over a 12-month period. Intervention START Diabetes Prevention intervention included a structured workflow, shared decision-making resources and electronic health record clinical decision support tools. Main Measures Uptake of prediabetes treatment, defined as Diabetes Prevention Program referral, metformin prescription and/or medical nutrition referral within 30 days of any PCC visit. Key Results We demonstrated greater uptake of preventive treatment among patients with prediabetes in the intervention clinic vs. control clinic receiving usual care (11.6% vs. 6.7%, pConclusions A low-touch multi-level intervention is effective in increasing prediabetes treatment uptake. The intervention was also acceptable and feasible for clinicians, and enhanced patient understanding and discussions of prediabetes with their clinicians.
- Open Access
- 5 June 2024
- Published by BMJ in BMJ Open Quality
- Vol. 13  (2) , e002752
Using the Translating Research into Practice framework to develop a diabetes prevention intervention in primary care: a mixed-methods study
Background Pre-diabetes affects one-third of US adults and increases the risk of type 2 diabetes. Effective evidence-based interventions, such as the Diabetes Prevention Program, are available, but a gap remains in effectively translating and increasing uptake of these interventions into routine care. Methods We applied the Translating Research into Practice (TRiP) framework to guide three phases of intervention design and development for diabetes prevention: (1) summarise the evidence, (2) identify local barriers to implementation and (3) measure performance. In phase 1, we conducted a retrospective cohort analysis of linked electronic health record claims data to evaluate current practices in the management of pre-diabetes. In phase 2, we conducted in-depth interviews of 16 primary care physicians, 7 payor leaders and 31 patients to elicit common barriers and facilitators for diabetes prevention. In phase 3, using findings from phases 1 and 2, we developed the core elements of the intervention and performance measures to evaluate intervention uptake. Results In phase 1 (retrospective cohort analysis), we found few patients with pre-diabetes received diabetes prevention interventions. In phase 2 (stakeholder engagement), we identified common barriers to include a lack of knowledge about pre-diabetes among patients and about the Diabetes Prevention Program among clinicians. In phase 3 (intervention development), we developed the START Diabetes Prevention Clinical Pathway as a systematic change package to address barriers and facilitators identified in phases 1 and 2, performance measures and a toolkit of resources to support the intervention components. Conclusions The TRiP framework supported the identification of evidence-based care practices for pre-diabetes and the development of a well-fitted, actionable intervention and implementation plan designed to increase treatment uptake for pre-diabetes in primary care settings. Our change package can be adapted and used by other health systems or clinics to target prevention of diabetes or other related chronic conditions.
- 5 June 2024
- Published by BMJ in BMJ Open Quality
- Vol. 13  (2) , e002752
Open AccessUsing the Translating Research into Practice framework to develop a diabetes prevention intervention in primary care: a mixed-methods study
Background Pre-diabetes affects one-third of US adults and increases the risk of type 2 diabetes. Effective evidence-based interventions, such as the Diabetes Prevention Program, are available, but a gap remains in effectively translating and increasing uptake of these interventions into routine care. Methods We applied the Translating Research into Practice (TRiP) framework to guide three phases of intervention design and development for diabetes prevention: (1) summarise the evidence, (2) identify local barriers to implementation and (3) measure performance. In phase 1, we conducted a retrospective cohort analysis of linked electronic health record claims data to evaluate current practices in the management of pre-diabetes. In phase 2, we conducted in-depth interviews of 16 primary care physicians, 7 payor leaders and 31 patients to elicit common barriers and facilitators for diabetes prevention. In phase 3, using findings from phases 1 and 2, we developed the core elements of the intervention and performance measures to evaluate intervention uptake. Results In phase 1 (retrospective cohort analysis), we found few patients with pre-diabetes received diabetes prevention interventions. In phase 2 (stakeholder engagement), we identified common barriers to include a lack of knowledge about pre-diabetes among patients and about the Diabetes Prevention Program among clinicians. In phase 3 (intervention development), we developed the START Diabetes Prevention Clinical Pathway as a systematic change package to address barriers and facilitators identified in phases 1 and 2, performance measures and a toolkit of resources to support the intervention components. Conclusions The TRiP framework supported the identification of evidence-based care practices for pre-diabetes and the development of a well-fitted, actionable intervention and implementation plan designed to increase treatment uptake for pre-diabetes in primary care settings. Our change package can be adapted and used by other health systems or clinics to target prevention of diabetes or other related chronic conditions.
- 31 July 2023
- Published by Elsevier in The Joint Commission Journal on Quality and Patient Safety
- Vol. 49  (12) , 698-705
Improving Diabetes Screening in the Primary Care Clinic
- 31 July 2023
- Published by Elsevier in The Joint Commission Journal on Quality and Patient Safety
- Vol. 49  (12) , 698-705
Improving Diabetes Screening in the Primary Care Clinic
- 20 May 2023
- Published by Elsevier in American Journal of Preventive Medicine
- Vol. 65  (5) , 906-915
Lower-Intensity Interventions for Prediabetes: A Systematic Review
- 20 May 2023
- Published by Elsevier in American Journal of Preventive Medicine
- Vol. 65  (5) , 906-915
Lower-Intensity Interventions for Prediabetes: A Systematic Review
- 20 September 2022
- Published by Springer Nature in Journal of General Internal Medicine
- Vol. 38  (2) , 309-314
Engaging Payors and Primary Care Physicians Together in Improving Diabetes Prevention
Background Type 2 diabetes can be prevented through lifestyle programs like the Diabetes Prevention Programs (DPP), but few people with prediabetes participate in them, in part because their insurance does not reliably cover DPPs. Prior studies have not focused on payor-level barriers. Objective To understand barriers to DPP uptake that exist and intersect at different levels (patients, PCPs, and payors) to inform strategies to improve diabetes prevention in primary care settings through interviews with PCPs and payors. Design From May 2020 to October 2021, we conducted remote, semi-structured interviews with PCPs and payors. Participants PCPs were from primary care practices affiliated with one mid-Atlantic academic system. Payor leaders were from regional commercial, Medicare, and Medicaid plans. Approach Using a standardized interview guide focused on barriers, facilitators, and potential intervention components, interviews were audio-recorded using Zoom and professionally transcribed. Two reviewers double-coded transcripts using the framework analytic approach. Key Results We interviewed 16 PCPs from 13 primary care clinics and 7 payor leaders representing 6 insurance plans. Two themes emerged from PCP reports of patient-level barriers: (1) lack of programs and insurance coverage of resources to address nutrition and exercise and (2) inadequate resources to address social determinants of health that impact diabetes prevention. Among barriers PCPs faced, we identified two themes: (1) low PCP knowledge about DPPs and misperceptions of insurance coverage of DPPs and (2) inadequate clinical staff to address diabetes prevention. Barriers common to PCPs and payors included (1) absence of prediabetes quality measures and (2) inadequate engagement of PCPs and patients with payors. Conclusions Discussions with PCPs and payors revealed systemic barriers that suggest important priorities to improve prediabetes clinical care, including universal coverage of DPPs, clarity about coverage benefits, data reporting and outreach by payors to PCPs, and adoption of appropriate prediabetes quality measures.
- 20 September 2022
- Published by Springer Nature in Journal of General Internal Medicine
- Vol. 38  (2) , 309-314
Engaging Payors and Primary Care Physicians Together in Improving Diabetes Prevention
Background Type 2 diabetes can be prevented through lifestyle programs like the Diabetes Prevention Programs (DPP), but few people with prediabetes participate in them, in part because their insurance does not reliably cover DPPs. Prior studies have not focused on payor-level barriers. Objective To understand barriers to DPP uptake that exist and intersect at different levels (patients, PCPs, and payors) to inform strategies to improve diabetes prevention in primary care settings through interviews with PCPs and payors. Design From May 2020 to October 2021, we conducted remote, semi-structured interviews with PCPs and payors. Participants PCPs were from primary care practices affiliated with one mid-Atlantic academic system. Payor leaders were from regional commercial, Medicare, and Medicaid plans. Approach Using a standardized interview guide focused on barriers, facilitators, and potential intervention components, interviews were audio-recorded using Zoom and professionally transcribed. Two reviewers double-coded transcripts using the framework analytic approach. Key Results We interviewed 16 PCPs from 13 primary care clinics and 7 payor leaders representing 6 insurance plans. Two themes emerged from PCP reports of patient-level barriers: (1) lack of programs and insurance coverage of resources to address nutrition and exercise and (2) inadequate resources to address social determinants of health that impact diabetes prevention. Among barriers PCPs faced, we identified two themes: (1) low PCP knowledge about DPPs and misperceptions of insurance coverage of DPPs and (2) inadequate clinical staff to address diabetes prevention. Barriers common to PCPs and payors included (1) absence of prediabetes quality measures and (2) inadequate engagement of PCPs and patients with payors. Conclusions Discussions with PCPs and payors revealed systemic barriers that suggest important priorities to improve prediabetes clinical care, including universal coverage of DPPs, clarity about coverage benefits, data reporting and outreach by payors to PCPs, and adoption of appropriate prediabetes quality measures.
- 2 March 2022
- Published by Springer Nature in Journal of General Internal Medicine
- Vol. 37  (16) , 4112-4119
Clinical Care Among Individuals with Prediabetes in Primary Care: a Retrospective Cohort Study
Background The incidence of diabetes in the general US population (6.7 per 1000 adults in 2018) has not changed significantly since 2000, suggesting that individuals with prediabetes are not connecting to evidence-based interventions. Objective We sought to describe the clinical care of individuals with prediabetes, determine patient factors associated with this care, and evaluate risk for diabetes development. Design Retrospective cohort study using linked claims and electronic health record data. Participants We created a cohort of adults with prediabetes based on laboratory measures. We excluded patients with a prior history of diabetes, pregnancy in prior 6 months, or recent steroid use. Main Measures We measured ordering and completion of clinical services targeting prediabetes management and diabetes incidence within 12 months following cohort entry. We tested the strength of the association between individuals’ characteristics and outcomes of interest using bivariate and multiple logistic regression. Results Our cohort included 3888 patients with a laboratory diagnosis of prediabetes (incident or prevalent prediabetes). Within 12 months, 63.4% had repeat glycemic testing, yet only 10.4% had coded diagnoses of prediabetes, 1.0% were referred for nutrition services, and 5.4% were prescribed metformin. Most patients completed labs and nutrition visits when referred and filled metformin when prescribed. Individuals with a higher glycemic level or BMI were more likely to receive prediabetes clinical care. Six percent of individuals developed diabetes within 12 months of cohort entry and had higher glycemic levels and BMI ≥ 30 kg/m2. In the adjusted model, Black individuals had 1.4 times higher odds of developing diabetes than White individuals. Conclusions Rates of prediabetes clinical care activities are low and have not improved. Strategies are urgently needed to improve prediabetes care delivery thereby preventing or delaying incident diabetes.
- 2 March 2022
- Published by Springer Nature in Journal of General Internal Medicine
- Vol. 37  (16) , 4112-4119
Clinical Care Among Individuals with Prediabetes in Primary Care: a Retrospective Cohort Study
Background The incidence of diabetes in the general US population (6.7 per 1000 adults in 2018) has not changed significantly since 2000, suggesting that individuals with prediabetes are not connecting to evidence-based interventions. Objective We sought to describe the clinical care of individuals with prediabetes, determine patient factors associated with this care, and evaluate risk for diabetes development. Design Retrospective cohort study using linked claims and electronic health record data. Participants We created a cohort of adults with prediabetes based on laboratory measures. We excluded patients with a prior history of diabetes, pregnancy in prior 6 months, or recent steroid use. Main Measures We measured ordering and completion of clinical services targeting prediabetes management and diabetes incidence within 12 months following cohort entry. We tested the strength of the association between individuals’ characteristics and outcomes of interest using bivariate and multiple logistic regression. Results Our cohort included 3888 patients with a laboratory diagnosis of prediabetes (incident or prevalent prediabetes). Within 12 months, 63.4% had repeat glycemic testing, yet only 10.4% had coded diagnoses of prediabetes, 1.0% were referred for nutrition services, and 5.4% were prescribed metformin. Most patients completed labs and nutrition visits when referred and filled metformin when prescribed. Individuals with a higher glycemic level or BMI were more likely to receive prediabetes clinical care. Six percent of individuals developed diabetes within 12 months of cohort entry and had higher glycemic levels and BMI ≥ 30 kg/m2. In the adjusted model, Black individuals had 1.4 times higher odds of developing diabetes than White individuals. Conclusions Rates of prediabetes clinical care activities are low and have not improved. Strategies are urgently needed to improve prediabetes care delivery thereby preventing or delaying incident diabetes.
- Open Access
- 24 February 2022
- Published by JMIR Publications Inc. in JMIR Medical Informatics
- Vol. 10  (2) , e29803
Identification of Prediabetes Discussions in Unstructured Clinical Documentation: Validation of a Natural Language Processing Algorithm
Background: Prediabetes affects 1 in 3 US adults. Most are not receiving evidence-based interventions, so understanding how providers discuss prediabetes with patients will inform how to improve their care. Objective: This study aimed to develop a natural language processing (NLP) algorithm using machine learning techniques to identify discussions of prediabetes in narrative documentation. Methods: We developed and applied a keyword search strategy to identify discussions of prediabetes in clinical documentation for patients with prediabetes. We manually reviewed matching notes to determine which represented actual prediabetes discussions. We applied 7 machine learning models against our manual annotation. Results: Machine learning classifiers were able to achieve classification results that were close to human performance with up to 98% precision and recall to identify prediabetes discussions in clinical documentation. Conclusions: We demonstrated that prediabetes discussions can be accurately identified using an NLP algorithm. This approach can be used to understand and identify prediabetes management practices in primary care, thereby informing interventions to improve guideline-concordant care.
- 24 February 2022
- Published by JMIR Publications Inc. in JMIR Medical Informatics
- Vol. 10  (2) , e29803
Open AccessIdentification of Prediabetes Discussions in Unstructured Clinical Documentation: Validation of a Natural Language Processing Algorithm
Background: Prediabetes affects 1 in 3 US adults. Most are not receiving evidence-based interventions, so understanding how providers discuss prediabetes with patients will inform how to improve their care. Objective: This study aimed to develop a natural language processing (NLP) algorithm using machine learning techniques to identify discussions of prediabetes in narrative documentation. Methods: We developed and applied a keyword search strategy to identify discussions of prediabetes in clinical documentation for patients with prediabetes. We manually reviewed matching notes to determine which represented actual prediabetes discussions. We applied 7 machine learning models against our manual annotation. Results: Machine learning classifiers were able to achieve classification results that were close to human performance with up to 98% precision and recall to identify prediabetes discussions in clinical documentation. Conclusions: We demonstrated that prediabetes discussions can be accurately identified using an NLP algorithm. This approach can be used to understand and identify prediabetes management practices in primary care, thereby informing interventions to improve guideline-concordant care.