Michael Baiocchi
· Associate Professor of Epidemiology & Population Health and, by courtesy, of Statistics and of MedicineVerifiedStanford University · Epidemiology and Population Health
Active 2002–2026
About
Michael Baiocchi is an associate professor of Epidemiology and Population Health at Stanford University, with courtesy appointments in Statistics and Medicine. He holds a PhD in Statistics from the University of Pennsylvania and a BA in Mathematics from Williams College. His research focuses on behavioral interventions and their rigorous evaluation, emphasizing the development of statistically rigorous methods for causal inference that are transparent and critique-friendly. Baiocchi has designed and served as the principal investigator for large randomized studies, including interventions to prevent sexual assault in Nairobi, Kenya. His work unites quantitative approaches with real-world questions to improve people's lives. He is actively involved in academic teaching, advising doctoral and master's students, and contributing to interdisciplinary research through memberships in Bio-X and the Wu Tsai Human Performance Alliance.
Research topics
- Medicine
- Political Science
- Psychiatry
- Psychology
- Clinical psychology
- Environmental health
- Sociology
- Public relations
- Internal medicine
- Artificial Intelligence
- Computer Science
- Machine Learning
- Social psychology
- Medical education
- Biology
- Physical therapy
- Engineering
- Surgery
- Radiology
- Obstetrics
- Applied psychology
Selected publications
Figshare · 2026-02-25
articleOpen accessSenior authorSupplementary Material 1.
BMC Surgery · 2026-02-25
articleOpen accessSenior authorBACKGROUND: Minimally invasive surgical approaches offers patient benefit, such as expedited recovery, and could reduce hospital cost. This study examines how sociodemographic factors influences surgical approach for aortic valve surgery. METHODS: We used data from The Society of Thoracic Surgeons' database to model selection into minimally invasive surgery vs. traditional sternotomy. Pair-matches were created between two identity types: male-vs-female sex and White-vs-Black race. Patients were matched on facility and other covariates. These pair-matches were summarized using generalized linear mixed models (logit-link), regressing surgery type on the relevant identity, with random effects for facility and matched pair. Additionally, a regional analysis summarizing variation in the mortality-risk profiles of patients was conducted. RESULTS: From 2015 to 2020, of the patients that met inclusion criteria, 68,956 patients underwent traditional sternotomy and 23,811 underwent minimally invasive surgery. For matched pairs of the examined covariate, the null hypothesis was that each patient would have the same odds of receiving each surgical approach. Our models estimate the odds ratio for receiving the minimally invasive surgery are 1.13 and 1.56 times higher for female and White patients respectively (both p-values <= 0.005). We also identified regional variation across levels of mortality-risk score and race. CONCLUSIONS: Our study demonstrates a pattern of variation in sorting in minimally invasive surgical aortic valve replacement vs traditional sternotomy via patient sex and race. These findings infer non-medical features guide patient candidacy for surgical approach, even within the same facility.
Journal of Adolescent Health · 2026-02-13
articleZenodo (CERN European Organization for Nuclear Research) · 2026-04-03
articleOpen accessThis protocol describes a cross-sectional observational study using NHANES 2011–2014 to estimate the effects of lifetime stroke and traumatic brain injury on healthcare access and utilization among U.S. adults. Planned analyses use propensity score matching and survey-weighted models, with primary outcomes of usual source of care and health insurance coverage.
Figshare · 2026-02-25
articleOpen accessSenior authorSupplementary Material 1.
Zenodo (CERN European Organization for Nuclear Research) · 2026-04-03
articleOpen accessThis protocol describes a cross-sectional observational study using NHANES 2011–2014 to estimate the effects of lifetime stroke and traumatic brain injury on healthcare access and utilization among U.S. adults. Planned analyses use propensity score matching and survey-weighted models, with primary outcomes of usual source of care and health insurance coverage.
The Prevalence and Characteristics of Labor Trafficking in Brazilian Agriculture
Journal of Human Trafficking · 2025-10-02
articleJournal of Educational and Behavioral Statistics · 2025-07-18
preprintOpen accessSenior authorRandomized experiments are considered the gold standard for estimating causal effects. However, out of the set of possible randomized assignments, some may be more likely to produce poor effect estimates and misleading conclusions. Restricted randomization is an experimental design strategy that filters out undesirable treatment assignments, but its application has primarily been limited to ensuring covariate balance in two-arm studies where the target estimand is the average treatment effect. Other experimental settings with different design desiderata and target effect estimands could also stand to benefit from a restricted randomization approach. We introduce inspection-guided randomization (IGR), a transparent and flexible framework for restricted randomization that filters out undesirable treatment assignments by inspecting assignments against analyst-specified, domain-informed design desiderata. In IGR, the acceptable treatment assignments are locked in ex ante and preregistered in the trial protocol, thus safeguarding against p-hacking and promoting reproducibility. Through illustrative simulation studies motivated by behavioral health and education interventions, we demonstrate how IGR can improve effect estimates compared to benchmark designs in experiments involving interference and group formation.
2025-06-03
supplementary-materialsOpen access<p>Supplemental Table S5 presents results from subgroup analysis by age with 40, 60, and 70 years as cut-offs.</p>
2025-06-03
supplementary-materialsOpen access<p>Supplemental Table S7 presents additional analyses that restricted the survivors and comparators in the study population to those lived longer than three or five years after the index date.</p>
Recent grants
NIH · $106k · 2014
The development of statistical tools for comparative effectiveness research
NIH · $747k · 2013–2017
Frequent coauthors
- 26 shared
Jonathan Chen
Stanford University
- 22 shared
Dylan S. Small
- 21 shared
Rachael C. Aikens
Stanford University
- 18 shared
Mary Amuyunzu‐Nyamongo
African Institute for Development Policy
- 18 shared
Scott A. Lorch
- 16 shared
Clea Sarnquist
Stanford University
- 14 shared
Y. Joseph Woo
Stanford University
- 12 shared
Judith J. Prochaska
Stanford University
Labs
Michael Baiocchi's ProfilePI
Education
- 2011
Ph.D., Statistics
University of Pennsylvania
- 2003
B.A., Mathematics
Williams College
- Resume-aware match score
- Save to shortlist
- AI-drafted outreach
See your match with Michael Baiocchi
PhdFit ranks faculty by your research interests, methods, and publications — grounded in their actual work, not templates.
- Free to start
- No credit card
- 30-second signup