About

David D. Kim, PhD, is the principal investigator of the Kim Research Group at the University of Chicago and serves as an Assistant Professor of Medicine and Public Health Sciences. His interdisciplinary research integrates health economics, policy evaluation, and decision modeling to advance value-based healthcare decisions. Dr. Kim has made significant contributions in evaluating long-term health outcomes, economic impacts, and equity implications of health interventions to guide evidence-informed policy. He also develops rigorous methodological approaches in economic evaluation and research prioritization to optimize healthcare resource allocation, and identifies geographic and racial disparities in the use of low-value healthcare services to inform targeted policy interventions and resource reallocation strategies. Dr. Kim has an extensive publication record with over 60 peer-reviewed articles in prominent journals such as JAMA, Annals of Internal Medicine, Health Affairs, and Medical Decision Making. He has developed innovative analytical tools and disease simulation models, including the Criteria for Health Economic Quality Evaluation (CHEQUE) tool and the Diabetes, Obesity, Cardiovascular Disease Microsimulation (DOC-M) model. His research has been featured in leading media outlets like The New York Times and The Washington Post, reflecting its broader public impact. Dr. Kim provides strategic guidance to national and international advisory groups, including the World Health Organization’s Guidelines Development Group, the Institute for Clinical and Economic Review (ICER), and currently serves as a Methods Consultant for the Annals of Internal Medicine. Prior to his current role, he was an Assistant Professor and Program Director of the Cost-Effectiveness Analysis Registry at the Center for the Evaluation of Value and Risk in Health at Tufts Medical Center. He earned his PhD in Health Economics from the University of Washington and holds a Master's degree in Biostatistics from the University of Michigan.

Research topics

• Medicine
• Intensive care medicine
• Computer Science
• Economics
• Endocrinology
• Medical physics
• Business
• Nursing
• Accounting
• Environmental health
• Demography
• Internal medicine
• Gerontology
• Operations management
• Actuarial science

Selected publications

• Impact of Self-Reported Patient–Provider Communication on the Use of High- and Low-Value Care Among U.S. Adults

  American Journal of Preventive Medicine · 2026-01-29

  articleSenior author
  PublisherDOI

• Cost-effectiveness thresholds used in the United States vs most favored nations

  Health Affairs Scholar · 2026-04-01

  articleOpen access

  Objectives: Cost-effectiveness thresholds inform whether health interventions represent good value for money, yet their use varies across countries. This study compares thresholds cited in published cost-effectiveness analyses (CEAs) in the United States with those in countries designated as the Most Favored Nations (MFNs) under the 2025 President's Executive Order on prescription drug pricing. Methods: We analyzed 6876 cost-per-QALY studies published between 1979 and 2023 from the Tufts CEA Registry. We standardized thresholds as multiples of each country's GDP per capita, and used logistic regression to estimate the probability of citing a threshold >1 × GDP per capita, adjusting for region, intervention type, disease area, and study period. Results: Over time, MFN studies shifted toward citing lower thresholds, whereas US thresholds consistently cited thresholds >1 × GDP per capita. After adjusting for other factors, MFN studies were less likely to cite higher thresholds than US studies. Cancer-related CEAs and CEAs of pharmaceutical interventions were more likely to cite higher thresholds. Conclusions: CEAs in the United States and peer high-income nations cite remarkably different thresholds, with MFNs citing lower value benchmarks over time. Policymakers should be cautious about adopting pricing policies that would implicitly subject US pharmaceutical spending to benchmarks developed in different institutional and fiscal contexts.

  PublisherDOI

• Detecting Where Effects Occur by Testing Hypotheses in Order

  Open MIND · 2026-02-24

  preprint

  Experimental evaluations of public policies often randomize a new intervention within many sites or blocks. After a report of an overall result -- statistically significant or not -- the natural question from a policy maker is: \emph{where} did any effects occur? Standard adjustments for multiple testing provide little power to answer this question. In simulations modeled after a 44-block education trial, the Hommel adjustment -- among the most powerful procedures controlling the family-wise error rate (FWER) -- detects effects in only 11\% of truly non-null blocks. We develop a procedure that tests hypotheses top-down through a tree: test the overall null at the root, then groups of blocks, then individual blocks, stopping any branch where the null is not rejected. In the same 44-block design, this approach detects effects in 44\% of non-null blocks -- roughly four times the detection rate. A stopping rule and valid tests at each node suffice for weak FWER control. We show that the strong-sense FWER depends on how rejection probabilities accumulate along paths through the tree. This yields a diagnostic: when power decays fast enough relative to branching, no adjustment is needed; otherwise, an adaptive $α$-adjustment restores control. We apply the method to 25 MDRC education trials and provide an R package, \texttt{manytestsr}.

  DOI

• Detecting Where Effects Occur by Testing Hypotheses in Order

  arXiv (Cornell University) · 2026-02-24

  articleOpen access

  Experimental evaluations of public policies often randomize a new intervention within many sites or blocks. After a report of an overall result -- statistically significant or not -- the natural question from a policy maker is: \emph{where} did any effects occur? Standard adjustments for multiple testing provide little power to answer this question. In simulations modeled after a 44-block education trial, the Hommel adjustment -- among the most powerful procedures controlling the family-wise error rate (FWER) -- detects effects in only 11\% of truly non-null blocks. We develop a procedure that tests hypotheses top-down through a tree: test the overall null at the root, then groups of blocks, then individual blocks, stopping any branch where the null is not rejected. In the same 44-block design, this approach detects effects in 44\% of non-null blocks -- roughly four times the detection rate. A stopping rule and valid tests at each node suffice for weak FWER control. We show that the strong-sense FWER depends on how rejection probabilities accumulate along paths through the tree. This yields a diagnostic: when power decays fast enough relative to branching, no adjustment is needed; otherwise, an adaptive $α$-adjustment restores control. We apply the method to 25 MDRC education trials and provide an R package, \texttt{manytestsr}.

  PublisherOA PDF

• Impact of Trainee Doctors’ Walkout on Mortality, Health Care Utilization, and Spending in South Korea

  SSRN Electronic Journal · 2026-01-01

  preprintOpen access1st authorCorresponding
  PublisherDOI

• Health and Economic Impact and Cost-Effectiveness of Produce Prescriptions for Diabetes in 50 US States: A Microsimulation Study

  2025-08-06

  preprintOpen access

  <p dir="ltr">Objective: Produce prescription (PRx) programs have shown improved dietary quality, diabetes control, and cardiometabolic outcomes. However, their potential health and economic impacts across 50 U.S. states, where Medicaid and commercial plans determine coverage, are unclear.</p><p dir="ltr">Research Design and Methods: Using a validated microsimulation model, we projected the health and economic outcomes of implementing PRx for adults aged 40–79 years with both diabetes and food insecurity in each U.S. state, over 5- and 10- years horizons. The model incorporated state-specific data, pooled intervention effects and costs from 20 PRx programs, diet-disease associations, and payer-specific healthcare expenditures. Outcomes and costs were discounted at 3% annually. Probabilistic sensitivity analyses accounted for uncertainty.</p><p dir="ltr">Results: An estimated 693,000 adults in California to 7,000 in Wyoming were eligible for PRx. Over 10 years, PRx was projected to avert between 9,240 CVD events (95% UI: 4,710–14,500) in Texas and 94 (48–147) in Alaska and gain 9,990 (4,810–15,500) to 92.2 (-30.8–159) QALYs across states. From a healthcare perspective, PRx was projected to be net cost-saving in 43 out of 50 states and was cost-effective (ICER<$150,000/QALY) in all, with New York having the largest net saving ($345M) and California the largest net costs ($155M). By insurance type, PRx was most likely to be cost-saving in Medicare beneficiaries, followed by Medicaid and private payers. Similar patterns were observed over 5 years. </p><p dir="ltr">Conclusions: PRx for patients with diabetes and food insecurity could substantially improve health and be cost saving or cost-effective in all U.S. states.</p>

  PublisherDOI

• Reply to Editorial Comment on “Characterization of Newborn Circumcision Revisions and Complications Among Privately Insured Boys in the United States Between 2010 and 2022”

  Urology · 2025-04-02

  letter1st authorCorresponding
  PublisherDOI

• Health and Economic Impact and Cost-effectiveness of Produce Prescriptions for Diabetes in 50 U.S. States: A Microsimulation Study

  Diabetes Care · 2025-08-06 · 1 citations

  articleOpen access

  OBJECTIVE: Produce prescription (PRx) programs have been shown to result in improved dietary quality, diabetes control, and cardiometabolic outcomes. However, their potential health and economic impacts across 50 U.S. states, where Medicaid and commercial plans determine coverage, are unclear. RESEARCH DESIGN AND METHODS: Using a validated microsimulation model, we projected the health and economic outcomes of implementing PRx for adults aged 40-79 years with both diabetes and food insecurity in each U.S. state, over 5- and 10-year horizons. The model incorporated state-specific data, pooled intervention effects and costs from 20 PRx programs, diet-disease associations, and payer-specific health care expenditures. Outcomes and costs were discounted at 3% annually. Probabilistic sensitivity analyses accounted for uncertainty. RESULTS: An estimated 693,000 adults in California to 7,000 in Wyoming were eligible for PRx. Over 10 years, PRx was projected to avert between 9,240 cardiovascular disease (CVD) events (95% uncertainty interval [UI] 4,710-14,500) in Texas and 94 (95% UI 48-147) in Alaska and gain 9,990 (95% UI 4,810-15,500) to 92.2 (95% UI -30.8 to 159) quality-adjusted life-years (QALYs) across states. From a health care perspective, PRx was projected to be net cost-saving in 43 of 50 states and was cost-effective (incremental cost-effectiveness ratio <$150,000/QALY) in all, with New York having the largest net saving ($345 million) and California the largest net costs ($155 million). By insurance type, PRx was most likely to be cost-saving for Medicare beneficiaries, followed by Medicaid and private payers. Similar patterns were observed over 5 years. CONCLUSIONS: PRx for patients with diabetes and food insecurity could substantially improve health and be cost saving or cost-effective in all U.S. states.

  PublisherOA PDFDOI

• Sexual dysfunction and other prolactin-related side effects of antipsychotic drugs in schizophrenia: Protocol for a systematic review with single-arm, pairwise, and network meta-analyses of randomized controlled trials and non-randomized studies

  F1000Research · 2025-06-19

  reviewOpen accessCorresponding

  <ns3:p>Introduction Sexual dysfunctions are common yet underreported side effects of antipsychotics for schizophrenia, affecting 30-80% of treated individuals. These side effects can severely impact social interactions and treatment adherence for individuals with schizophrenia, but comprehensive comparative evidence assessing the risk profiles of different antipsychotics is lacking. This study aims to address this gap using network meta-analysis that integrates data from both randomized-controlled trials (RCTs) and non-randomized studies (NRS). Protocol This systematic review will include both RCTs and NRS focusing on participants with schizophrenia or schizophrenia-like psychoses, without restrictions on symptoms, gender, ethnicity, age, or setting. For interventions, all second-generation antipsychotics will be included. The primary outcome will be the occurrence of at least one sexual adverse event of any kind. Secondary outcomes will be the occurrence of any sexual adverse event evaluated in men and women separately, and any adverse event related to the three phases of sexual response cycle separately: desire (e.g. libido, sexual thoughts), arousal (e.g. erection, lubrication) and orgasm (e.g. ejaculation, anorgasmia), and any adverse effect related to breast dysfunction and menstruation irregularities. Study selection and data extraction will be performed independently by two reviewers. The Cochrane Risk of Bias tool 1 and ROBINS-I will be employed to evaluate the risk of bias for RCTs and NRS, respectively. Single-arm meta-analysis of proportions will synthesize the average frequency of sexual adverse events in treated participants. Pairwise and network meta-analysis of RCTs and NRS will be used to evaluate comparative tolerability. Subgroup and sensitivity analyses will explore possible heterogeneity in results and validate the findings’ robustness. The quality of the evidence will be evaluated using GRADE. Discussion This study will provide vital insights into the sexual side effects of antipsychotics by combining evidence from clinical trials and real-world practice, facilitating better decision-making in choosing the optimal antipsychotic for individuals.</ns3:p>

  PublisherOA PDFDOI

• Health and Economic Impact and Cost-Effectiveness of Produce Prescriptions for Diabetes in 50 US States: A Microsimulation Study

  2025-08-06

  preprintOpen access

  &lt;p dir="ltr"&gt;Objective: Produce prescription (PRx) programs have shown improved dietary quality, diabetes control, and cardiometabolic outcomes. However, their potential health and economic impacts across 50 U.S. states, where Medicaid and commercial plans determine coverage, are unclear.&lt;/p&gt;&lt;p dir="ltr"&gt;Research Design and Methods: Using a validated microsimulation model, we projected the health and economic outcomes of implementing PRx for adults aged 40–79 years with both diabetes and food insecurity in each U.S. state, over 5- and 10- years horizons. The model incorporated state-specific data, pooled intervention effects and costs from 20 PRx programs, diet-disease associations, and payer-specific healthcare expenditures. Outcomes and costs were discounted at 3% annually. Probabilistic sensitivity analyses accounted for uncertainty.&lt;/p&gt;&lt;p dir="ltr"&gt;Results: An estimated 693,000 adults in California to 7,000 in Wyoming were eligible for PRx. Over 10 years, PRx was projected to avert between 9,240 CVD events (95% UI: 4,710–14,500) in Texas and 94 (48–147) in Alaska and gain 9,990 (4,810–15,500) to 92.2 (-30.8–159) QALYs across states. From a healthcare perspective, PRx was projected to be net cost-saving in 43 out of 50 states and was cost-effective (ICER&lt;$150,000/QALY) in all, with New York having the largest net saving ($345M) and California the largest net costs ($155M). By insurance type, PRx was most likely to be cost-saving in Medicare beneficiaries, followed by Medicaid and private payers. Similar patterns were observed over 5 years. &lt;/p&gt;&lt;p dir="ltr"&gt;Conclusions: PRx for patients with diabetes and food insecurity could substantially improve health and be cost saving or cost-effective in all U.S. states.&lt;/p&gt;

  PublisherOA PDFDOI

Frequent coauthors

• Peter J. Neumann

  Tufts Medical Center

  56 shared

• Daniel A. Ollendorf

  Institute for Clinical and Economic Review

  44 shared

• John B. Wong

  Tufts Medical Center

  32 shared

• Kalipso Chalkidou

  Imperial College London

  29 shared

• William Reuben

  Tanzania Industrial Research and Development Organization

  25 shared

• Patrice Prognon

  Assistance Publique – Hôpitaux de Paris

  25 shared

• Schiffon Wong

  25 shared

• Gina Nicholson

  25 shared

Labs

• Kim Research GroupPI

Education

• Ph.D., Public Health Sciences

  University of Chicago

• M.D.

  University of Chicago

• B.A.

  University of Chicago