Current Research and Scholarly Interests
Fernando Alarid-Escudero uses simulation modeling and decision analysis to answer health policy questions that trials and observational studies cannot answer alone: which interventions to offer, to whom, when, and at what cost, and how much more evidence is worth collecting before deciding. His work has four connected threads.
Cancer control policy. He is a Principal Investigator of the National Cancer Institute's CISNET colorectal and gastric cancer awards and a co-investigator on the CISNET bladder cancer award. He leads SimCRC, one of the three CISNET colorectal cancer models used to inform U.S. Preventive Services Task Force screening recommendations, and DESCIPHR-GC, Stanford's new CISNET gastric cancer model, built on his lab's open-source DESCIPHR framework. He is also Co-Principal Investigator of a project supporting the Task Force's evidence reviews. Current colorectal work evaluates emerging screening options such as blood-based tests, FIT thresholds tailored by age and sex, the consequences of incomplete follow-up after a positive test, and the risks of incompletely removed polyps. In gastric cancer, his team estimates how Helicobacter pylori infection has spread across birth cohorts in the U.S. and globally, and builds these estimates into DESCIPHR-GC to evaluate prevention, screening, and surveillance strategies such as H. pylori screen-and-treat. He also collaborates on models of bladder, breast, and cervical cancer.
Global health in Latin America. He studies how middle-income countries can build cancer screening programs that fit their resources and health-system capacity. Current projects evaluate colorectal cancer screening in Mexico, recognized with Stanford's 2025 Rosenkranz Prize, and in Chile, with Stanford Impact Labs, including training for Chile's Ministry of Health in cost-effectiveness methods.
Methods for trustworthy models. He develops methods that make simulation models more credible and efficient: Bayesian calibration with machine-learning emulators (BayCANN), workflows that diagnose non-identifiability before calibration (PRE-CISE), fast sampling of event times for discrete-event simulation, value-of-information methods that show which research is worth funding, and frameworks that represent social processes and health-system constraints so that models reveal inequities instead of averaging them away.
Open-source and reproducible modeling. As co-founder of DARTH and ConVOI, he develops free R packages (dampack, darthpack, darthtools), published tutorials, and pipelines such as DESCIPHR, an open-source framework to build, calibrate, and apply cancer simulation models. He has taught these tools in courses and workshops across North America, Europe, Latin America, and Asia.
Beyond cancer, he has applied these tools to COVID-19 (the SC-COSMO model, used to project hospital demand in Mexico City, and analyses of vaccination and reopening policies in U.S. prisons), dementia care, HIV prevention, and drug policy.
He welcomes PhD students and postdoctoral fellows with quantitative training who want to build models that change policy.
