Miguel Ángel Luque-Fernandez
Sections
  • Research
Academic Background, Positions, and Collaborations
Academic Background
PhD in Preventive Medicine and Public Health (Summa Cum Laude) from University of Granada and University Free of Brussels (UFB).
Holds a BSc in Mathematics and Statistics (Open University, UK), MSc in Biostatistics (Newcastle University, Australia), MSc in Epidemiology (UFB), and a Master in Public Health (University of Granada, Spain).
Completed postdoctoral fellowships at University of Cape Town and Harvard, specializing in epidemiological methods and casual inference (2012-2015).
Trained as a Field Epidemiologist (Epidemic Intelligence Officer), he has worked with Médecins Sans Frontières and WHO-GOARN responding to health emergencies in Haiti and sub-Saharan Africa.
His research focuses on epidemiological methods, causal inference, and socioeconomic inequalities in cancer outcomes.
Publication list and metrics at UGR repository:

produccioncientifica.ugr.es

MIGUEL ÁNGEL LUQUE FERNÁNDEZ - Universidad de Granada

I received my Ph.D. in Preventive Medicine (Epidemiology) and Public Health, awarded Summa Cum Laude, from the University of Granada (UGR, Spain) and the ULB (Université Libre de Bruxelles, Belgium). Also, I hold a Bachelor's in Maths and Stats from the Open University, UK, an MSc in Biostatistics from the University of Newcastle, Australia, an MSc in Epidemiology from ULB, and a Master's in Public Health from the UGR. After the completion of my Ph.D. in 2010, I moved to the Center for Infectiou

Academic Positions
Associated Professor in Statistics at the University of Granada, Faculty of Sciences and Biostatistics Unit at the Faculty of Medicine.

Universidad de Granada

Miguel Ángel Luque Fernández | Universidad de Granada

Fundada en 1531, institución de carácter docente e investigador. Con más de 54000 estudiantes, es primer destino Erasmus y está en posiciones destacadas en rankings como Shanghai

Honorary Associated Professor of Epidemiology and Biostatistics at the London School of Hygiene and Tropical Medicine and member of the ICON group.

LSHTM

Miguel-Angel Luque-Fernandez | LSHTM

I received my Ph.D. in Preventive Medicine (Epidemiology) and Public Health, awarded with Summa Cum Laude, from the University of Granada (UGR, Spain) and the ULB (Universite Libre de Bruxelles, Belgium). Also, I hold a BSc in Mathematics and Statistics from the Open University, an MSc in Biostatistics from the University of Newcastle, Australia, an MSc in Epidemiology from the ULB, and an MPH from the UGR. After finishing my Ph.D. in 2010, I moved to the Center for Infectious Disease Epidemiolo

Inequalities in Cancer Outcomes Network

Inequalities in Cancer Outcomes Network

Why are cancer outcomes poorer in some population groups in countries with universal access to healthcare? Why does access to optimal cancer care vary between different cancer patients? What are the mechanisms leading to such disparities? These are the type of questions our research aims to answer.

Collaborations
He was a visiting Associated Professor at the Division of Biostatistics of the Berkeley School of Public Health and since then he is promoting the Targeted Maximum Likelihood (TMLE) framework for causal inference—a methodology that has become standard in comparative effectiveness research. His work on TMLE has been cited extensively in leading journals and has influenced best practices in observational study design across major academic institutions.

vanderlaan-lab.org

Lab Members and Alumni · The Research Group of Mark van der Laan

Students, Post-Docs, and Alumni of the van der Laan Group

He also collaborates with the Harvard Medical School helping to identify plasma glycated CD59 as a powerful early biomarker for Gestational Diabetes Mellitus and predicting large-for-gestational-age newborns, showing high accuracy even in the first trimester, potentially earlier and better than traditional screening methods like HbA1c or the glucose challenge test alone, especially in high-risk women.

American Diabetes Association

Plasma Glycated CD59, a Novel Biomarker for Detection of Pregnancy-Induced Glucose Intolerance

OBJECTIVE. Plasma glycated CD59 (pGCD59) is an emerging biomarker in diabetes. We assessed whether pGCD59 could predict the following: the results of the g

Research
Causal Inference
Advanced statistical methods for causal effect estimation including Targeted Maximum Likelihood Estimation (TMLE), ensemble learning, and cross-validation techniques. Development of data-adaptive methods for model selection and evaluation in epidemiological research.
Key achievements and applications:
  • Development of TMLE software and cross-validated TMLE for observational studies in STATA.
  • Publications in high-impact journals such as Statistics in Medicine and American Journal of Epidemiology.
  • Application in comparative effectiveness studies to evaluate health interventions.
Cancer Epidemiology
Socioeconomic inequalities in cancer incidence, mortality, and survival. Multimorbidity patterns in cancer patients. Health disparities and comparative effectiveness research. Population-based cohort studies using administrative health data.
Specific research areas:
  • Socioeconomic inequalities in breast, colorectal, and prostate cancer.
  • Population-based cohort studies using health data in Spain and the United Kingdom.
  • Analysis of comorbidity and multimorbidity patterns in cancer patients.
Teaching
1
Interactive Applications

watzile.shinyapps.io

Some flexible parametric distributions for modelling time-to-event data

Time-to-event data appears in many research areas such as medicine, engineering, biology, and finance, to name but a few. For instance, modelling and understanding the survival times of cancer patients has become an important aim of many countries. In statistical terms, times-to-event correspond to a sample of positive observations, possibly censored due to lost of follow-up or administrative censoring. These observations represent either the time at which the event of interest happens (e.g. death of a patient or failure of an electric device) or the last time of follow-up. There exist several approaches to model time to event data, such as nonparametric, semiparametric, and parametric methods. Parametric distributions represent a parsimonious approach to modelling time-to-event data since they are typically easier to implement and to interpret. There is a large catalogue of flexible parametric distributions that can be used to model this kind of data. This shiny app illustrates different features (Hazard, Cumulative Hazard, and Survival functions) of interest of some of the most popular parametric distributions used to model time-to-event data. By changing the values of the parameters of the corresponding distributions, using the bars on the left hand side, the user can visualize the effect on the shapes of the hazard, cumulative hazard, and survival functions.

watzile.shinyapps.io

Collider

Based on a motivating example in non-communicable disease epidemiology, we generated a dataset with 1,000 observations to contextualize the effect of conditioning on a collider. Nearly 1 in 3 Americans suffer from high blood pressure and more than half do not have it under control [1]. Increased levels of systolic blood pressure over time are associated with increased cardio-vascular

watzile.shinyapps.io

Pattern of Comorbidities and Multimorbidity among Colorectal Cancer Patients in Spain

We found a consistent pattern of factors associated with a higher prevalence of comorbidities and multimorbidity at diagnosis among colorectal cancer patients at diagnosis in Spain. This pattern may add valuable insights for further etiological and preventive research and may help to identify patients at higher risk for poorer cancer outcomes and treatment.

2
Survival Analysis Tutorials

migariane.github.io

Tutorial: Introduction to Survival Analyses

migariane.github.io

TUTORIAL: AGE-STANDARDISED NET SURVIVAL BY DEPRIVATION COHORT APPROACH

Note: the data used in the tutorial has been modified from the original source for a teaching purpose and represent breast cancer incident cases between 1971 and 2001 in England

migariane.github.io

Modeling Relative Survival/Excess Mortality using stpm2 (Part 2)

Flexible parametric modelling using restricted cubic splines

3
Case-Control Study Methods

GitHub

GitHub - migariane/cmatch: Stata module for tabulation of matched pairs in 1:1 case control study by exposure status

Stata module for tabulation of matched pairs in 1:1 case control study by exposure status - migariane/cmatch

GitHub

GitHub - migariane/cmatch: Stata module for tabulation of matched pairs in 1:1 case control study by exposure status

Stata module for tabulation of matched pairs in 1:1 case control study by exposure status - migariane/cmatch

4
Cross-Validation & Model Evaluation

migariane.github.io

Cross-Validation in Practice by Miguel Angel Luque Fernandez

Cross-Validation is a data partitioning method that can be used to: i) Assess the stability of parameter estimates, ii) Assess the accuracy of classification algorithms, and iii) Assess the adequacy of a fitted model

GitHub

GitHub - migariane/cvAUROC: Cross-validated Area Under the Curve for ROC Analysis after Predictive Modelling for Binary Outcomes

Cross-validated Area Under the Curve for ROC Analysis after Predictive Modelling for Binary Outcomes - migariane/cvAUROC

5
Causal Inference Methods

migariane.github.io

CIM.html

Copyright (c) 2017 Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this per

migariane.github.io

Targeted Maximum Likelihood Estimation for a Binary Outcome: Tutorial and Guided Implementation

During the last 30 years, modern epidemiology has been able to identify significant limitations of classic epidemiologic methods when the focus is to explain the main effect of a risk factor on a disease or outcome.

GitHub

GitHub - migariane/eltmle: Ensemble Learning Targeted Maximum Likelihood for Stata users

Ensemble Learning Targeted Maximum Likelihood for Stata users - GitHub - migariane/eltmle: Ensemble Learning Targeted Maximum Likelihood for Stata users

migariane.github.io

OneSimulation.html

Note: Data reproduce a Cancer Epidemiological example. The interpretation of the results is not applicable to real-world

6
Statistical Methods

migariane.github.io

Delta Method in Epidemiology: An Applied and Reproducible Tutorial.

Approximate statistical inference via determination of the asymptotic distribution of the statistics is a technique that epidemiologists and applied statisticians use routinely, we name it the classical delta-method, but there is a gap as it is not routinely taught and many applied researchers do not understand it neither its uses. The delta-method is a theorem which states that a smooth function of an asymptotically normal estimator is also asymptotically normally distributed. It can also be vi

7
Statistical Methods

migariane.github.io

The Comprehensive Guide to Conformal Statistical Inference

This interactive guide provides a visual and intuitive introduction to Conformal Inference, a modern framework for quantifying uncertainty in Machine Learning. Unlike traditional point predictions, which offer a single "best guess," Conformal Inference generates prediction intervals that come with a rigorous mathematical guarantee: the true value will fall within the predicted range at a user-defined confidence level (e.g., 95%). The core strength of this method lies in being model-agnostic and distribution-free. It doesn’t matter if you are using a simple linear regression or a complex deep neural network, and it doesn't assume that data follows a specific pattern (like a Normal distribution). The process typically involves: Splitting the data: Using a "calibration set" separate from the training data. Calculating Non-conformity Scores: Measuring how much a model’s prediction deviates from the actual values in the calibration set. Computing a Quantile: Finding an error threshold (q^​) that covers the desired percentage of cases. The resource excels at breaking down these technical steps into digestible, interactive visualizations, making it accessible for practitioners who want to move beyond "black box" models and provide reliable, safety-critical predictions.

8
Medical Statistics:
Access a rich collection of learning materials covering fundamental concepts to advanced techniques in medical statistics. Ideal for practical application in health research.

migariane.github.io

Análisis Estadístico con Ordenador de Datos Médicos

Este libro de texto está dirigido a estudiantes de medicina, residentes y otros profesionales de la salud que deseen aprender a utilizar el lenguaje de programación R para realizar análisis estadísticos en medicina. También es adecuado para cualquier persona interesada en aprender a usar R para el análisis estadístico en general, con un enfoque en aplicaciones médicas.

Medical Statistics Tutorials
Explore essential tutorials on advanced statistical and epidemiological methods, designed to enhance understanding and application in research. These resources, often co-authored by Dr. Miguel Angel Luque Fernandez, cover pivotal concepts for robust study design and analysis.
1
Statistics in Medicine Journal
These tutorials delve into modern causal inference techniques, providing practical guidance for complex epidemiological questions.

Wiley Online Library

Introduction to computational causal inference using reproducible Stata, R, and Python code: A tutorial

The main purpose of many medical studies is to estimate the effects of a treatment or exposure on an outcome. However, it is not always possible to randomize the study participants to a particular tr...

Wiley Online Library

Targeted maximum likelihood estimation for a binary treatment: A tutorial

When estimating the average effect of a binary treatment (or exposure) on an outcome, methods that incorporate propensity scores, the G-formula, or targeted maximum likelihood estimation (TMLE) are p...

Wiley Online Library

Using longitudinal targeted maximum likelihood estimation in complex settings with dynamic interventions

Longitudinal targeted maximum likelihood estimation (LTMLE) has very rarely been used to estimate dynamic treatment effects in the context of time-dependent confounding affected by prior treatment wh...

2
International Journal of Epidemiology
Understand key methodological challenges such as collider stratification bias in epidemiological research through this accessible tutorial.

watzile.shinyapps.io

Collider

Based on a motivating example in non-communicable disease epidemiology, we generated a dataset with 1,000 observations to contextualize the effect of conditioning on a collider. Nearly 1 in 3 Americans suffer from high blood pressure and more than half do not have it under control [1]. Increased levels of systolic blood pressure over time are associated with increased cardio-vascular

Software
ELTMLE
Ensemble Learning Targeted Maximum Likelihood Estimation for Stata. Implements advanced causal inference methods for epidemiological research. Combines ensemble learning with TMLE for robust causal effect estimation.
cvAUROC
Cross-validated Area Under the Curve for ROC Analysis. Provides cross-validation techniques for evaluating predictive models in binary outcomes. Essential tool for model selection and validation.
All packages are open-source and available on GitHub. GitHub: @migariane
Publications
Over 100 peer-reviewed articles published in leading journals including The Lancet, BMJ, JAMA, PLoS Medicine, and Cancer. Research spans perinatal epidemiology, social epidemiology, non-communicable disease epidemiology, biostatistics, statistical methods, and cancer epidemiology.
40
h-index
4500+
Citations
76
i10-index
For a complete list of publications, visit:
Books
Explore essential academic texts and resources, meticulously crafted or co-authored by Dr. Miguel Ángel Luque-Fernández. These publications offer deep insights into biostatistics, epidemiological methods, and causal inference, providing invaluable tools for researchers and students alike.
Medical Statistics:
Access a rich collection of learning materials covering fundamental concepts to advanced techniques in medical statistics. Ideal for practical application in health research.

migariane.github.io

Análisis Estadístico con Ordenador de Datos Médicos

Este libro de texto está dirigido a estudiantes de medicina, residentes y otros profesionales de la salud que deseen aprender a utilizar el lenguaje de programación R para realizar análisis estadísticos en medicina. También es adecuado para cualquier persona interesada en aprender a usar R para el análisis estadístico en general, con un enfoque en aplicaciones médicas.

Causal Inference in Practice
Understand and apply cutting-edge causal inference methodologies, including Targeted Maximum Likelihood Estimation (TMLE). This resource bridges theory with practical application for real-world scenarios.
Guides on data-adaptive methods and effect estimation.
In progress …
Cancer Epidemiology & Health Disparities
Examine critical research on cancer incidence, mortality, and survival, with a focus on socioeconomic inequalities and multimorbidity. Explore population-based studies and health disparities. Insights into prevention and public health strategies.
Links to the books:
  1. Social disparities in survival from lung cancer in Europe Social Environment and Cancer in Europe: Towards an Evidence-Based Public Health Policy (Springer International Publishing), pp. 121-140
  1. The role of comorbidities in the social gradient in cancer survival in Europe Social Environment and Cancer in Europe: Towards an Evidence-Based Public Health Policy (Springer International Publishing), pp. 261-286
Contact
Connect & Follow
Academic Affiliations
University of Granada
Department of Statistics and Operations Research, Granada, Spain
Division of Biostatistics, Faculty of Medicine
London School of Hygiene and Tropical Medicine
London, UK
Made with