So-Yeon Kang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
So-Yeon Kang has a total of 21 co-authors. Here are the top co-authors by number of shared publications:
- Gerard Anderson
Department of Health Policy and Management, Johns Hopkins Bloomberg School of Public Health , Baltimore, MD 21205
3 shared publications - Yunan Ji
McDonough School of Business, Georgetown University, Washington, DC, United States
3 shared publications - Abhishek Bazaz
School of Medicine, Johns Hopkins University, Baltimore, MD, United States
2 shared publications - Mariana P. Socal
Department of Health Policy and Management, Johns Hopkins Bloomberg School of Public Health , Baltimore, MD 21205
2 shared publications - Ravi Gupta
Division of General Internal Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland
2 shared publications Allison S. Kim
Department of Health Management and Policy, Georgetown University School of Health, Washington, DC
1 shared publication
Latest Publications
- Open Access
- 1 May 2026
- Published by American Medical Association (AMA) in JAMA Network Open
- Vol. 9 (5) , e2610053
International Availability and Price of Ultraexpensive Drugs in Medicare Part D
This cohort study characterizes ultraexpensive drugs in Medicare Part D and examines their availability and relative prices across industrialized countries.
- 1 May 2026
- Published by American Medical Association (AMA) in JAMA Network Open
- Vol. 9 (5) , e2610053
Open AccessInternational Availability and Price of Ultraexpensive Drugs in Medicare Part D
This cohort study characterizes ultraexpensive drugs in Medicare Part D and examines their availability and relative prices across industrialized countries.
- 27 April 2026
- Published by Wolters Kluwer Health in Medical Care
- Vol. 64 (7) , 413-418
Success Rates of Venture Capital Investment in Biopharmaceutical Development
Background: Large biopharmaceutical companies increasingly outsource early research and development (R&D) to startups funded by venture capital (VC). Yet little is known about how effective VC funding is at advancing new drugs to market. Objectives: To examine the outcomes of clinical trials conducted by VC-backed startups and identify factors associated with their progression. Research Design: Retrospective cohort study of VC-backed biopharmaceutical clinical trials. Subjects: Index phase 1 trials initiated between January 2006 and December 2015, tracked through April 2024. Measures: Progression to phase 2, phase 3, and US Food and Drug Administration (FDA) approval. Associations between trial characteristics and progression were estimated using multinomial logistic regression. Results: Among 1357 VC-backed phase 1 trials, 10.9% failed to complete phase 1, 13.3% completed phase 1 and stopped, 42.5% progressed only to phase 2, 19.2% only to phase 3, and 14.1% ultimately received FDA approval. Cancer-related trials achieved FDA approval in 8.7% of cases. Compared with small-molecule trials, biological trials were less likely to progress to phase 3 [−11.3 percentage points (pp), P P P <0.001). Conclusions: Clinical trial success rates among VC-backed startups were within the range of approval rates previously reported for large manufacturers. Progression varied substantially by drug modality, therapeutic focus, and initial investment size, underscoring how capital allocation and development strategy shape R&D outcomes.
- 27 April 2026
- Published by Wolters Kluwer Health in Medical Care
- Vol. 64 (7) , 413-418
Success Rates of Venture Capital Investment in Biopharmaceutical Development
Background: Large biopharmaceutical companies increasingly outsource early research and development (R&D) to startups funded by venture capital (VC). Yet little is known about how effective VC funding is at advancing new drugs to market. Objectives: To examine the outcomes of clinical trials conducted by VC-backed startups and identify factors associated with their progression. Research Design: Retrospective cohort study of VC-backed biopharmaceutical clinical trials. Subjects: Index phase 1 trials initiated between January 2006 and December 2015, tracked through April 2024. Measures: Progression to phase 2, phase 3, and US Food and Drug Administration (FDA) approval. Associations between trial characteristics and progression were estimated using multinomial logistic regression. Results: Among 1357 VC-backed phase 1 trials, 10.9% failed to complete phase 1, 13.3% completed phase 1 and stopped, 42.5% progressed only to phase 2, 19.2% only to phase 3, and 14.1% ultimately received FDA approval. Cancer-related trials achieved FDA approval in 8.7% of cases. Compared with small-molecule trials, biological trials were less likely to progress to phase 3 [−11.3 percentage points (pp), P P P <0.001). Conclusions: Clinical trial success rates among VC-backed startups were within the range of approval rates previously reported for large manufacturers. Progression varied substantially by drug modality, therapeutic focus, and initial investment size, underscoring how capital allocation and development strategy shape R&D outcomes.
- Open Access
- 25 February 2026
- Published by JMIR Publications Inc. in Journal of Medical Internet Research
- Vol. 28, e84968
Trends in Venture Capital Investment in AI-Driven Biopharmaceutical Startups
This study analyzes 2010-2024 venture capital trends in international artificial intelligence–driven biopharmaceutical startups, revealing rapid growth in discovery tool investments and concentrated US funding in California and Massachusetts.
- 25 February 2026
- Published by JMIR Publications Inc. in Journal of Medical Internet Research
- Vol. 28, e84968
Open AccessTrends in Venture Capital Investment in AI-Driven Biopharmaceutical Startups
This study analyzes 2010-2024 venture capital trends in international artificial intelligence–driven biopharmaceutical startups, revealing rapid growth in discovery tool investments and concentrated US funding in California and Massachusetts.
- Open Access
- 30 October 2025
- Published by Oxford University Press (OUP) in Health Affairs Scholar
- Vol. 3 (11)
Global venture capital flows and US health care innovation
- 30 October 2025
- Published by Oxford University Press (OUP) in Health Affairs Scholar
- Vol. 3 (11)
Open AccessGlobal venture capital flows and US health care innovation
- Open Access
- 28 September 2025
- Published by JMIR Publications Inc. in JMIR Preprints
Trends in Venture Capital Investment in AI-Driven Biopharmaceutical Startups (Preprint)
UNSTRUCTURED This study analyzes 2010-2024 venture capital trends in international artificial intelligence–driven biopharmaceutical startups, revealing rapid growth in discovery tool investments and concentrated US funding in California and Massachusetts.
- 28 September 2025
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessTrends in Venture Capital Investment in AI-Driven Biopharmaceutical Startups (Preprint)
UNSTRUCTURED This study analyzes 2010-2024 venture capital trends in international artificial intelligence–driven biopharmaceutical startups, revealing rapid growth in discovery tool investments and concentrated US funding in California and Massachusetts.
- Open Access
- 4 August 2025
- Published by Oxford University Press (OUP) in Health Affairs Scholar
- Vol. 3 (8) , qxaf161
Research and development financing models for new biopharmaceuticals in the United States
- 4 August 2025
- Published by Oxford University Press (OUP) in Health Affairs Scholar
- Vol. 3 (8) , qxaf161
Open AccessResearch and development financing models for new biopharmaceuticals in the United States
- Open Access
- 1 July 2025
- Published by Oxford University Press (OUP) in Health Affairs Scholar
- Vol. 3 (7) , qxaf139
Research and development investments for biologics independently developed by US biotechnology startups, 2017–2023
Introduction: Despite policy interest in pharmaceutical innovation, little is known about the investment needed for venture capital–backed startups to develop innovative biologics, a growing segment of the pharmaceutical marketplace. Methods: In a cross-sectional analysis of Food and Drug Administration (FDA) drug approval data, investment deal records, and clinical trials data, we estimated the investment needed for a biotechnology startup company to independently develop a biologic approved by the US FDA with priority review, an indicator of innovation. To isolate a homogenous set of cases, the sample focused on independent drug development of 13 drugs by 9 biotechnology startup companies that retained ownership of the drug from the start of development to FDA approval without being acquired or licensing the product. Results: We found that the median investment per FDA-approved biologic was $304.1 million (IQR: $289.9–$790.3 million) in uncapitalized costs, accounting for the cost of failures. Conclusion: This estimate represents the direct drug R&D investment needed for independent development of innovative biologic products by biotechnology startup companies.
- 1 July 2025
- Published by Oxford University Press (OUP) in Health Affairs Scholar
- Vol. 3 (7) , qxaf139
Open AccessResearch and development investments for biologics independently developed by US biotechnology startups, 2017–2023
Introduction: Despite policy interest in pharmaceutical innovation, little is known about the investment needed for venture capital–backed startups to develop innovative biologics, a growing segment of the pharmaceutical marketplace. Methods: In a cross-sectional analysis of Food and Drug Administration (FDA) drug approval data, investment deal records, and clinical trials data, we estimated the investment needed for a biotechnology startup company to independently develop a biologic approved by the US FDA with priority review, an indicator of innovation. To isolate a homogenous set of cases, the sample focused on independent drug development of 13 drugs by 9 biotechnology startup companies that retained ownership of the drug from the start of development to FDA approval without being acquired or licensing the product. Results: We found that the median investment per FDA-approved biologic was $304.1 million (IQR: $289.9–$790.3 million) in uncapitalized costs, accounting for the cost of failures. Conclusion: This estimate represents the direct drug R&D investment needed for independent development of innovative biologic products by biotechnology startup companies.
- Open Access
- 1 April 2025
- Published by American Medical Association (AMA) in JAMA Network Open
- Vol. 8 (4) , e255785
Claim Denials for Cancer-Related Next-Generation Sequencing in Medicare
Importance: The Centers for Medicare & Medicaid Services implemented a National Coverage Determination (NCD) for genetic testing using next-generation sequencing (NGS) in March 2018 and amended it in January 2020. Little is known about how often NGS insurance claims are denied and what factors are associated with these denials.Objective: To examine the prevalence of and factors associated with claim denials for NGS testing among Medicare enrollees.Design, Setting, and Participants: This cohort study used a 20% random sample of Medicare claims data from January 1, 2016, to December 31, 2021. Statistical analysis took place from June to October 2024.Main Outcomes and Measures: Claim denials were measured based on payment and denial data. Multivariable logistic regression models were estimated with patient and claim characteristics, including testing type, testing site, and time period indicators reflecting the presence and scope of the NCD: no NCD (period 1), first NCD (period 2), and amended NCD (period 3).Results: The study sample included 29 919 cancer-related NGS claims among 24 443 unique Medicare beneficiaries (51.8% males; 44.7% aged >75 years). The number of cancer-related NGS testing claims increased from 1912 in 2016 to 9177 in 2021. Claim denial occurred among 23.3% of the NGS claims in the sample. The claim denial rate was 16.8% before the NCD, 20.3% after the implementation of the NCD in 2018, and 27.4% after the amended NCD in 2020. Claims for NGS testing were more likely to be denied if they were performed in independent laboratories (odds ratio [OR], 2.76 [95% CI, 2.58-2.95]; P < .001) or other nonhospital sites (OR, 2.55 [95% CI, 2.12-3.07]; P < .001) vs a hospital or if they were for 50 or more genes vs 50 or fewer genes for solid tumors (OR, 1.32 [95% CI, 1.23-1.43]; P < .001). The likelihood of a claim denial was higher in both time periods after the original 2018 NCD compared with before the NCD (period 2: OR, 1.23 [95% CI, 1.10-1.36]; P < .001; period 3: OR, 1.64 [95% CI, 1.49-1.80]; P < .001). The claim denial likelihood for period 3 was significantly higher than period 2 (OR, 1.34 [95% CI, 1.24-1.44]; P < .001).Conclusions and Relevance: In this cohort study of cancer-related NGS testing claims among Medicare enrollees, denial rates varied by testing type and testing site and increased over time. The findings suggest the continued existence of uncertainty regarding coverage for NGS, despite the NCD. Policy approaches to further reduce uncertainty regarding NGS coverage and raise awareness of potential financial liability warrant consideration.
- 1 April 2025
- Published by American Medical Association (AMA) in JAMA Network Open
- Vol. 8 (4) , e255785
Open AccessClaim Denials for Cancer-Related Next-Generation Sequencing in Medicare
Importance: The Centers for Medicare & Medicaid Services implemented a National Coverage Determination (NCD) for genetic testing using next-generation sequencing (NGS) in March 2018 and amended it in January 2020. Little is known about how often NGS insurance claims are denied and what factors are associated with these denials.Objective: To examine the prevalence of and factors associated with claim denials for NGS testing among Medicare enrollees.Design, Setting, and Participants: This cohort study used a 20% random sample of Medicare claims data from January 1, 2016, to December 31, 2021. Statistical analysis took place from June to October 2024.Main Outcomes and Measures: Claim denials were measured based on payment and denial data. Multivariable logistic regression models were estimated with patient and claim characteristics, including testing type, testing site, and time period indicators reflecting the presence and scope of the NCD: no NCD (period 1), first NCD (period 2), and amended NCD (period 3).Results: The study sample included 29 919 cancer-related NGS claims among 24 443 unique Medicare beneficiaries (51.8% males; 44.7% aged >75 years). The number of cancer-related NGS testing claims increased from 1912 in 2016 to 9177 in 2021. Claim denial occurred among 23.3% of the NGS claims in the sample. The claim denial rate was 16.8% before the NCD, 20.3% after the implementation of the NCD in 2018, and 27.4% after the amended NCD in 2020. Claims for NGS testing were more likely to be denied if they were performed in independent laboratories (odds ratio [OR], 2.76 [95% CI, 2.58-2.95]; P < .001) or other nonhospital sites (OR, 2.55 [95% CI, 2.12-3.07]; P < .001) vs a hospital or if they were for 50 or more genes vs 50 or fewer genes for solid tumors (OR, 1.32 [95% CI, 1.23-1.43]; P < .001). The likelihood of a claim denial was higher in both time periods after the original 2018 NCD compared with before the NCD (period 2: OR, 1.23 [95% CI, 1.10-1.36]; P < .001; period 3: OR, 1.64 [95% CI, 1.49-1.80]; P < .001). The claim denial likelihood for period 3 was significantly higher than period 2 (OR, 1.34 [95% CI, 1.24-1.44]; P < .001).Conclusions and Relevance: In this cohort study of cancer-related NGS testing claims among Medicare enrollees, denial rates varied by testing type and testing site and increased over time. The findings suggest the continued existence of uncertainty regarding coverage for NGS, despite the NCD. Policy approaches to further reduce uncertainty regarding NGS coverage and raise awareness of potential financial liability warrant consideration.
- Open Access
- 4 October 2024
- Published by Oxford University Press (OUP) in Health Affairs Scholar
- Vol. 2 (10) , qxae124
Biopharmaceutical pipeline funded by venture capital firms, 2014 to 2024
Venture capital (VC) firms fund biopharmaceutical research and development (R&D) while incurring substantial financial risk. VC firms seek to invest in clinical areas with the greatest potential for financial return. Using a combination of data for clinical trials and VC investment deals between January 2014 and March 2024, we found that approximately 75% of VC investments were allocated to clinical trials studying small-molecule drugs compared to biologics or gene therapies, without substantial changes over the study period. Most of VC firms’ investment in biopharmaceutical R&D was concentrated in phase 1 and phase 2 clinical trials. This trend has increased in recent years, with phase 1 trials accounting for nearly half of total deals and capital investments in 2023. VC investments were concentrated in several therapeutic areas, including cancer.
- 4 October 2024
- Published by Oxford University Press (OUP) in Health Affairs Scholar
- Vol. 2 (10) , qxae124
Open AccessBiopharmaceutical pipeline funded by venture capital firms, 2014 to 2024
Venture capital (VC) firms fund biopharmaceutical research and development (R&D) while incurring substantial financial risk. VC firms seek to invest in clinical areas with the greatest potential for financial return. Using a combination of data for clinical trials and VC investment deals between January 2014 and March 2024, we found that approximately 75% of VC investments were allocated to clinical trials studying small-molecule drugs compared to biologics or gene therapies, without substantial changes over the study period. Most of VC firms’ investment in biopharmaceutical R&D was concentrated in phase 1 and phase 2 clinical trials. This trend has increased in recent years, with phase 1 trials accounting for nearly half of total deals and capital investments in 2023. VC investments were concentrated in several therapeutic areas, including cancer.
- 1 September 2024
- Published by Academy of Managed Care Pharmacy in Journal of Managed Care & Specialty Pharmacy
- Vol. 30 (9) , 903-907
Manufacturer-sponsored drug coupon use and drug-switching behavior among patients with type 2 diabetes
BACKGROUND: Patients often use manufacturer-sponsored coupons reduce their out-of-pocket spending. However, little is known whether coupon use is associated with medication-switching behaviors. OBJECTIVE: To examine if using a manufacturer-sponsored coupon to initiate a medication is associated with patterns of medication-switching behaviors among patients with type 2 diabetes. METHODS: Using IQVIA's retail pharmacy claims data from October 2017 to September 2019, we analyzed commercially insured patients with type 2 diabetes who had newly started taking the following noninsulin diabetes drugs: generic metformin (nearly no coupon use), Sodium-glucose Cotransporter-2 (SGLT2) Inhibitors (SGLT2, high coupon use), and dipeptidyl peptidase IV inhibitors (DPP-IV inhibitors, moderate coupon use). We assessed if drug-switching behaviors, defined as no switching, switching to a same-class drug, or switching a drug in a different class, differed among patients who did and did not use coupons to initiate treatments. We performed multinomial logistic regression to estimate the probability of each switching type associated with patients' initial coupon use. RESULTS: Among 9,781 patients in our sample, 83.7% of them initiated treatments with metformin, 8.2% with SGLT2, and 8.1% with DPP-IV inhibitors. The overall switching rate was the lowest for generic metformin (40%) than brand-name drugs (56%-57%). Among the brand-name drug users, patients who used a coupon to initiate these drugs were less likely to switch to any drug compared with patients without coupon use (SGLT2 = -18% [95% CI = -24% to -13%]; DPP-IV inhibitors = -9% [-16% to -2%]). These patients were also less likely to switch to drugs in other competing classes (SGLT2 = -16% [95% CI = -22% to -10%]; DPP-IV inhibitors = -9% [-16% to -2%]). CONCLUSIONS: Patients who started their treatment with generic metformin had the lowest rate of drug switching. Using coupons to initiate brand-name drugs in classes with prevalent coupons was associated with reduced medication switching to other class drugs.
- 1 September 2024
- Published by Academy of Managed Care Pharmacy in Journal of Managed Care & Specialty Pharmacy
- Vol. 30 (9) , 903-907
Manufacturer-sponsored drug coupon use and drug-switching behavior among patients with type 2 diabetes
BACKGROUND: Patients often use manufacturer-sponsored coupons reduce their out-of-pocket spending. However, little is known whether coupon use is associated with medication-switching behaviors. OBJECTIVE: To examine if using a manufacturer-sponsored coupon to initiate a medication is associated with patterns of medication-switching behaviors among patients with type 2 diabetes. METHODS: Using IQVIA's retail pharmacy claims data from October 2017 to September 2019, we analyzed commercially insured patients with type 2 diabetes who had newly started taking the following noninsulin diabetes drugs: generic metformin (nearly no coupon use), Sodium-glucose Cotransporter-2 (SGLT2) Inhibitors (SGLT2, high coupon use), and dipeptidyl peptidase IV inhibitors (DPP-IV inhibitors, moderate coupon use). We assessed if drug-switching behaviors, defined as no switching, switching to a same-class drug, or switching a drug in a different class, differed among patients who did and did not use coupons to initiate treatments. We performed multinomial logistic regression to estimate the probability of each switching type associated with patients' initial coupon use. RESULTS: Among 9,781 patients in our sample, 83.7% of them initiated treatments with metformin, 8.2% with SGLT2, and 8.1% with DPP-IV inhibitors. The overall switching rate was the lowest for generic metformin (40%) than brand-name drugs (56%-57%). Among the brand-name drug users, patients who used a coupon to initiate these drugs were less likely to switch to any drug compared with patients without coupon use (SGLT2 = -18% [95% CI = -24% to -13%]; DPP-IV inhibitors = -9% [-16% to -2%]). These patients were also less likely to switch to drugs in other competing classes (SGLT2 = -16% [95% CI = -22% to -10%]; DPP-IV inhibitors = -9% [-16% to -2%]). CONCLUSIONS: Patients who started their treatment with generic metformin had the lowest rate of drug switching. Using coupons to initiate brand-name drugs in classes with prevalent coupons was associated with reduced medication switching to other class drugs.