Fernando Alarid-Escudero, PhD, is an assistant professor of health policy at the Department of Health Policy in the Stanford University School of Medicine. He obtained his PhD in Health Decision Sciences from the University of Minnesota School of Public Health, and was an assistant professor at the Center for Research and Teaching in Economics (CIDE) Región Centro, Aguascalientes, Mexico, from 2018 to 2022, prior to coming to Stanford. His research focuses on developing statistical and decision-analytic models to identify optimal prevention, control, and treatment policies to address a wide range of public health problems and develops novel methods to quantify the value of future research. 

In this Faculty Focus, Alarid-Escudero Tells Us About His Research

My current research focuses on developing and applying decision-analytic models to health policy and public health problems at a local, regional, national, and international level. I am a three-time awardee of the Lee B. Lusted Prize for Outstanding Student Research from the Society for Medical Decision Making (SMDM) in 2014, 2015, and 2016, and was elected Trustee of SMDM in 2018. In 2020, I was named a COVID-19 Decision Modeling Initiative (CDMI) Leader by SMDM.

I’m a member of the Cancer Intervention and Surveillance Modeling Network (CISNET) consortium, a team of investigators sponsored by the National Cancer Institute in the U.S. that uses modeling to evaluate the impact of cancer control interventions on trends in incidence and mortality. As part of my research, I develop simulation models to evaluate screening policies for three cancers (colorectal [CRC], bladder, and gastric) in the U.S. and other low- and middle-income countries (LMIC) at a population level. I also implement novel Bayesian methodologies using high-performance computing (HPC) systems to quantify the impact of uncertainty of models of the natural history of cancer on health policy analyses.

I co-lead a cost-effectiveness and feasibility analysis of a city-wide CRC screening program in Mexico City using a modified version of a US-specific CRC CISNET model. I am co-developing a microsimulation model of CRC patients, incorporating genomic biomarkers to explore the clinical utility and evaluate the cost-effectiveness of genomic testing. I am leading the comparative modeling of the three CISNET CRC models by integrating large-scale probabilistic inference algorithms with detailed microsimulation models to develop a more transparent, comprehensive understanding of the impact of uncertainty on model outcomes, ultimately leaving the models better suited to inform policy decisions.

I also lead the evaluation of the potential benefits of a country-wide FIT-based CRC screening program in Chile using our CRC CISNET microsimulation model adapted to the Chilean population. The results of this project will inform the design of screening programs in Chile and potentially guide screening policies in other countries in Latin America. 

My work on gastric cancer focuses on disparities and prevention in the U.S. and globally. I am developing a policy decision tool to evaluate different screening-and-treating policies for the Helicobacter pylori (H. pylori) bacteria on gastric cancer accounting for antibiotic resistance in the U.S. and various LMIC in collaboration with the International Agency for Research on Cancer (IARC). This tool is an extension of my doctoral dissertation, where I developed a dynamic epidemiologic model of H. pylori and a natural history model of gastric cancer to evaluate the cost-effectiveness of different screen-and-treat strategies for H. pylori on the prevalence of H. pylori infection and antibiotic resistance and gastric cancer incidence in Mexico. With this work, I showed that mass-treatment strategies in the presence of high antibiotic resistance could provide small benefits, at best, at the expense of significantly increasing resistance levels.

As part of my work in bladder cancer, I am co-developing a novel natural history individual-level simulation model of bladder cancer that will be used in population modeling for early detection and control in the U.S.

Collaborating with the American Cancer Society (ACS), I’m developing a policy decision tool to evaluate state-level screening policies for cervical cancer and human papillomavirus (HPV) vaccination strategies in the U.S. using a comprehensive model that combines both a dynamic transmission model of HPV and a progression model of cervical cancer.

I also have co-founded international collaborative research networks, such as the Decision Analysis in R for Technologies in Health (DARTH) workgroup, an international, multi-institutional collaborative effort comprised of researchers with a passion for transparent and open-source solutions to decision analysis in health. Our ultimate aim is to expand knowledge in simulation-based health policy analysis using R statistical software. As part of my work in DARTH, I co-authored several tutorials, developed new or more efficient methodologies, and taught workshops in different research institutions across the globe. I also co-founded the Collaborative Network for Value of Information (ConVoI), an international and multi-institutional collaborative research group that develops transparent and open-source solutions to quantify the value of potential future research efficiently. As part of my work in ConVoI, I have co-authored several research articles and taught workshops at international conferences. At the Department of Health Policy, I plan to organize workshops on cost-effectiveness, decision modeling, and VOI for participants worldwide.

Most recently, I co-founded the Stanford-CIDE Coronavirus Simulation Model (SC-COSMO) modeling consortium, where I co-developed simulation models of the COVID-19 virus that provide policymakers in the U.S. and Mexico with insights into the impact of different mitigation strategies and policy decisions on health outcomes. These collaborations speak to my commitment to creating bridges among researchers and decision-makers from other parts of the world to address timely relevant public health issues and reduce knowledge barriers of state-of-the-art decision modeling methods. At the Department of Health Policy, I plan to design and organize a workshop on infectious disease modeling applied to cost-effectiveness and health policy analysis for participants worldwide.

I am particularly interested in developing new or enhancing current methods to improve decision-analytic modeling for health policy analyses. Some illustrations of such interest are as follows. I co-developed a novel, efficient method to perform value of information analysis (VOI). This method can be applied to estimate the value of different types of potential future research and prioritize feasible study designs in cancer prevention and early detection and across the cancer continuum. I have also described potential biases when modeling the effectiveness of health technologies as a reduction in overall causes of death beyond the time horizon of a trial instead of reducing the disease-specific mortality and proposed a new approach to overcome this bias. I use machine learning methods to improve the accessibility and transparency of models and model-based analyses. 

For example, I co-developed a machine learning metamodel-based Bayesian calibration approach to estimate the parameters of computationally expensive simulation models. In more statistical work, I proposed and developed a catalytic epidemic model to estimate the force of infection (FOI) of H. pylori in Mexico using Bayesian methods on a nationally representative seroepidemiological survey in the pre-treatment era. Currently, I use this method to estimate the FOI of H. pylori for different racial and ethnic groups in the U.S. and other countries. I also developed open-source software for analyzing and visualizing the health economic outputs of decision-analytic models, to quantify the value of further research and determine optimal research study designs, estimate bias-corrected effectiveness parameters of health technologies when used in decision-analytic models, and evaluate the cost-effectiveness of testing cancer patients for biomarker expressions followed by adjuvant chemotherapy.