Miguel A. Hernán

ORCID: 0000-0003-1619-8456
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Advanced Causal Inference Techniques
  • Health Systems, Economic Evaluations, Quality of Life
  • Statistical Methods in Clinical Trials
  • Statistical Methods and Inference
  • Statistical Methods and Bayesian Inference
  • HIV/AIDS Research and Interventions
  • COVID-19 Clinical Research Studies
  • SARS-CoV-2 and COVID-19 Research
  • HIV/AIDS drug development and treatment
  • HIV Research and Treatment
  • Colorectal Cancer Screening and Detection
  • Vaccine Coverage and Hesitancy
  • Multiple Sclerosis Research Studies
  • HIV-related health complications and treatments
  • COVID-19 epidemiological studies
  • Meta-analysis and systematic reviews
  • Parkinson's Disease Mechanisms and Treatments
  • Nutritional Studies and Diet
  • COVID-19 and healthcare impacts
  • Obesity, Physical Activity, Diet
  • Long-Term Effects of COVID-19
  • Bayesian Modeling and Causal Inference
  • Global Cancer Incidence and Screening
  • Genetic Associations and Epidemiology
  • Lipoproteins and Cardiovascular Health

Harvard University
2016-2025

Boston University
2004-2025

Karolinska Institutet
2021-2025

Harvard University Press
2009-2024

Brigham and Women's Hospital
2000-2024

National Health Service
2024

Harvard–MIT Division of Health Sciences and Technology
2014-2023

Instituto de Salud Carlos III
2016-2023

Centre for Biomedical Network Research on Rare Diseases
2023

Centro de Investigación Biomédica en Red de Epidemiología y Salud Pública
2023

Non-randomised studies of the effects interventions are critical to many areas healthcare evaluation, but their results may be biased. It is therefore important understand and appraise strengths weaknesses. We developed ROBINS-I ("Risk Of Bias In Studies - Interventions"), a new tool for evaluating risk bias in estimates comparative effectiveness (harm or benefit) from that did not use randomisation allocate units (individuals clusters individuals) comparison groups. The will particularly...

10.1136/bmj.i4919 article EN cc-by-nc BMJ 2016-10-12

In observational studies with exposures or treatments that vary over time, standard approaches for adjustment of confounding are biased when there exist time-dependent confounders also affected by previous treatment. This paper introduces marginal structural models, a new class causal models allow improved in those situations. The parameters model can be consistently estimated using estimators, the inverse-probability-of-treatment weighted estimators.

10.1097/00001648-200009000-00011 article EN Epidemiology 2000-09-01

Observational cohort studies and a secondary prevention trial have shown inverse associations between adherence to the Mediterranean diet cardiovascular risk.In multicenter in Spain, we assigned 7447 participants (55 80 years of age, 57% women) who were at high risk, but with no disease enrollment, one three diets: supplemented extra-virgin olive oil, mixed nuts, or control (advice reduce dietary fat). Participants received quarterly educational sessions and, depending on group assignment,...

10.1056/nejmoa1800389 article EN New England Journal of Medicine 2018-06-13

BackgroundAs mass vaccination campaigns against coronavirus disease 2019 (Covid-19) commence worldwide, vaccine effectiveness needs to be assessed for a range of outcomes across diverse populations in noncontrolled setting. In this study, data from Israel's largest health care organization were used evaluate the BNT162b2 mRNA vaccine.MethodsAll persons who newly vaccinated during period December 20, 2020, February 1, 2021, matched unvaccinated controls 1:1 ratio according demographic and...

10.1056/nejmoa2101765 article EN New England Journal of Medicine 2021-02-24

The term "selection bias" encompasses various biases in epidemiology. We describe examples of selection bias case-control studies (eg, inappropriate controls) and cohort informative censoring). argue that the causal structure underlying each example is essentially same: conditioning on a common effect 2 variables, one which either exposure or cause other outcome outcome. This shared by adjustment for variables affected prior exposure). A structural classification distinguishes between...

10.1097/01.ede.0000135174.63482.43 article EN Epidemiology 2004-08-12

The method of inverse probability weighting (henceforth, weighting) can be used to adjust for measured confounding and selection bias under the four assumptions consistency, exchangeability, positivity, no misspecification model estimate weights. In recent years, several published estimates effect time-varying exposures have been based on weighted estimation parameters marginal structural models because, unlike standard statistical methods, appropriately confounders affected by prior...

10.1093/aje/kwn164 article EN American Journal of Epidemiology 2008-07-15

Ideally, questions about comparative effectiveness or safety would be answered using an appropriately designed and conducted randomized experiment. When we cannot conduct a experiment, analyze observational data. Causal inference from large databases (big data) can viewed as attempt to emulate experiment—the target experiment trial—that answer the question of interest. goal is guide decisions among several strategies, causal analyses data need evaluated with respect how well they particular...

10.1093/aje/kwv254 article EN American Journal of Epidemiology 2016-03-18
Marina Pollán Beatriz Pérez‐Gómez Roberto Pastor‐Barriuso Jesús Oteo Miguel A. Hernán and 95 more Mayte Pérez‐Olmeda Jose L Sanmartín Aurora Fernández-García Israel Cruz Nerea Fernández de Larrea Marta Molina Francisco Rodríguez-Cabrera Mariano Martín Paloma Merino Amador José León Paniagua Juan F Muñoz-Montalvo Faustino Blanco Raquel Yotti Faustino Blanco Rodrigo Gutiérrez Fernández Mariano Martín Saturnino Mezcua Navarro Marta Molina Juan F Muñoz-Montalvo Matías Salinero Hernández Jose L Sanmartín Manuel Cuenca‐Estrella Raquel Yotti José León Paniagua Nerea Fernández de Larrea José María Navarro‐Marí Roberto Pastor‐Barriuso Beatriz Pérez‐Gómez Marina Pollán Ana Avellón Giovanni Fedele Aurora Fernández-García Jesús Oteo Mayte Pérez‐Olmeda Israel Cruz Elena Fernández‐Martínez Francisco Rodríguez-Cabrera Miguel A. Hernán Susana Padrones Fernández José Manuel Rumbao Aguirre José María Navarro‐Marí Begoña Palop Borrás Ana Belén Jiménez Manuel Rodríguez‐Iglesias Ana María Calvo Gascón María Luz Lou Alcaine Ignacio Donate Suárez O. Alvarez Mercedes Rodríguez Pérez Margarita Cases Sanchís Carlos Javier Villafáfila-Gomila Lluís Carbó Saladrigas Adoración Hurtado Fernández Antonio Oliver Elías Castro Feliciano María Noemí González Quintana José María Barrasa Fernández María Araceli Hernández Betancor Melisa Hernández Febles Leopoldo Martín Martín Luis-Mariano López López Teresa Ugarte Miota Inés De Benito Población María Sagrario Celada Pérez María Natalia Vallés Fernández Tomás Maté M. Arranz Marta Domínguez-Gil González Isabel Fernández-Natal Gregoria Megías Lobón Juan Luis Muñoz Bellido Pilar Ciruela Ariadna Mas i Casals María Doladé María Ángeles Marcos Dúnia Pérez del Campo Antonio Félix de Castro Ramón Limón Ramírez Maria Francisca Elías Retamosa Manuela Rubio González María Sinda Blanco Lobeiras Alberto Fuentes Losada Antonio Aguilera Germán Bou Yolanda Caro Noemí Marauri Luis Miguel Soria Blanco Isabel del Cura-González Montserrat Hernández Pascual Roberto Alonso Paloma Merino Amador Natalia Cabrera Castro Aurora Tomás Lizcano Cristóbal Ramírez Almagro M. Segovia Hernández

10.1016/s0140-6736(20)31483-5 article EN other-oa The Lancet 2020-07-06

Standard methods for survival analysis, such as the time-dependent Cox model, may produce biased effect estimates when there exist confounders that are themselves affected by previous treatment or exposure. Marginal structural models a new class of causal parameters which estimated through inverse-probability-of-treatment weighting; these allow appropriate adjustment confounding. We describe marginal proportional hazards model and use it to estimate zidovudine on human immunodeficiency...

10.1097/00001648-200009000-00012 article EN Epidemiology 2000-09-01

Common strategies to decide whether a variable is confounder that should be adjusted for in the analysis rely mostly on statistical criteria. The authors present findings from Slone Epidemiology Unit Birth Defects Study, 1992–1997, case-control study folic acid supplementation and risk of neural tube defects. When confounding evaluation are used, odds ratio 0.80 (95% confidence interval: 0.62, 1.21). However, consideration priori causal knowledge suggests crude 0.65 0.46, 0.94) used because...

10.1093/aje/155.2.176 article EN American Journal of Epidemiology 2002-01-15

<b><i>Background:</i></b> A protective effect of vitamin D on risk multiple sclerosis (MS) has been proposed, but no prospective studies have addressed this hypothesis. <b><i>Methods:</i></b> Dietary intake was examined directly in relation to MS two large cohorts women: the Nurses' Health Study (NHS; 92,253 women followed from 1980 2000) and II (NHS II; 95,310 1991 2001). Diet assessed at baseline updated every 4 years thereafter. During follow-up, 173 cases with onset symptoms after were...

10.1212/01.wnl.0000101723.79681.38 article EN Neurology 2004-01-13

The hazard ratio (HR) is the main, and often only, effect measure reported in many epidemiologic studies. For dichotomous, non–time-varying exposures, HR defined as exposed groups divided by unexposed groups. all practical purposes, hazards can be thought of incidence rates thus roughly interpreted rate ratio. commonly conveniently estimated via a Cox proportional model, which include potential confounders covariates. Unfortunately, use for causal inference not straightforward even absence...

10.1097/ede.0b013e3181c1ea43 article EN Epidemiology 2009-12-09
Shruti Gupta Salim S. Hayek Wei Wang Lili Chan Kusum S. Mathews and 95 more Michal L. Melamed Samantha K. Brenner Amanda K. Leonberg‐Yoo Edward J. Schenck Jared Radbel Jochen Reiser Anip Bansal Anand Srivastava Yan Zhou Anne Sutherland Adam Green Alexandre M. Shehata Nitender Goyal Anitha Vijayan Juan Carlos Q. Velez Shahzad Shaefi Chirag R. Parikh Justin Arunthamakun Ambarish M. Athavale Allon N. Friedman Samuel Short Zoé A. Kibbelaar Samah Abu Omar Andrew J. Admon John P. Donnelly Hayley B. Gershengorn Miguel A. Hernán Matthew W. Semler David E. Leaf Carl P. Walther Samaya J. Anumudu Kathleen F. Kopecky Gregory P. Milligan Peter A. McCullough Thuy-Duyen Nguyen Megan L. Krajewski Sidharth Shankar Ameeka Pannu Juan D. Valencia Sushrut S. Waikar Peter C. Hart Oyintayo Ajiboye Matthew Itteera Jean-Sébastien Rachoin Christa Schorr Lisa Shea Daniel Edmonston Christopher L. Mosher Aaron Karp Zaza Cohen Valerie Allusson Gabriela Bambrick‐Santoyo Noor ul aain Bhatti Bijal Mehta Aquino Williams Patricia Walters Rolando C Go Keith M. Rose Amy M. Zhou Ethan C. Kim Rebecca Lisk Steven G. Coca Deena R. Altman Aparna Saha Howard Soh Huei Hsun Wen Sonali Bose Emily Leven Jing G. Wang Gohar Mosoyan Girish N. Nadkarni John Guirguis Rajat Kapoor Christopher Meshberger Brian T. Garibaldi Celia P. Corona-Villalobos Yumeng Wen Steven Menez Rubab F. Malik Carmen Elena Cervantes Samir C. Gautam Hang Nguyen Afshin Ahoubim Leslie F. Thomas Dheeraj Reddy Sirganagari Pramod Guru Paul A. Bergl Jesús Rodríguez Jatan A. Shah Mrigank S. Gupta Princy Kumar Deepa G. Lazarous Seble Kassaye Tanya S. Johns Ryan Mocerino

<h3>Importance</h3> The US is currently an epicenter of the coronavirus disease 2019 (COVID-19) pandemic, yet few national data are available on patient characteristics, treatment, and outcomes critical illness from COVID-19. <h3>Objectives</h3> To assess factors associated with death to examine interhospital variation in treatment for patients <h3>Design, Setting, Participants</h3> This multicenter cohort study assessed 2215 adults laboratory-confirmed COVID-19 who were admitted intensive...

10.1001/jamainternmed.2020.3596 article EN JAMA Internal Medicine 2020-07-15

In ideal randomised experiments, association is causation: measures can be interpreted as effect because randomisation ensures that the exposed and unexposed are exchangeable. On other hand, in observational studies, not generally cannot However, research often only alternative for causal inference. This article reviews a condition permits estimation of effects from data, two methods -- standardisation inverse probability weighting to estimate population under condition. For simplicity, main...

10.1136/jech.2004.029496 article EN Journal of Epidemiology & Community Health 2006-06-21

Preapproval trials showed that messenger RNA (mRNA)-based vaccines against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) had a good safety profile, yet these were subject to size and patient-mix limitations. An evaluation of the BNT162b2 mRNA vaccine with respect broad range potential adverse events is needed.We used data from largest health care organization in Israel evaluate vaccine. For each event, population persons no previous diagnosis we individually matched vaccinated...

10.1056/nejmoa2110475 article EN New England Journal of Medicine 2021-08-25

The use of instrumental variable (IV) methods is attractive because, even in the presence unmeasured confounding, such may consistently estimate average causal effect an exposure on outcome. However, for this consistent estimation to be achieved, several strong conditions must hold. We review definition variable, describe required obtain estimates effects, and explore their implications context a recent application variables approach. also present (1) description connection between 4...

10.1097/01.ede.0000222409.00878.37 article EN Epidemiology 2006-05-24

Many countries are experiencing a resurgence of COVID-19, driven predominantly by the delta (B.1.617.2) variant SARS-CoV-2. In response, these considering administration third dose mRNA COVID-19 vaccine as booster to address potential waning immunity over time and reduced effectiveness against variant. We aimed use data repositories Israel's largest health-care organisation evaluate BNT162b2 for preventing severe outcomes.Using from Clalit Health Services, which provides mandatory coverage...

10.1016/s0140-6736(21)02249-2 article EN other-oa The Lancet 2021-10-29

Patients with ESRD commonly experience secondary hyperparathyroidism, a condition primarily managed activated injectable vitamin D. The biologic effects of D, however, are widespread, and it is possible that D alters survival in ESRD. This hypothesis was tested historical cohort study incident hemodialysis patients who lived throughout the United States between January 1996 December 1999. primary outcome 2-yr among those survived for at least 90 d after initiation chronic hemodialysis....

10.1681/asn.2004070573 article EN Journal of the American Society of Nephrology 2005-02-24

Background: The Women's Health Initiative randomized trial found greater coronary heart disease (CHD) risk in women assigned to estrogen/progestin therapy than those placebo. Observational studies had previously suggested reduced CHD hormone users. Methods: Using data from the observational Nurses' Study, we emulated design and intention-to-treat (ITT) analysis of trial. study was conceptualized as a sequence "trials," which eligible were classified initiators or noninitiators therapy....

10.1097/ede.0b013e3181875e61 article EN Epidemiology 2008-10-09

10.1016/j.cmpb.2003.10.004 article EN Computer Methods and Programs in Biomedicine 2003-12-19
Coming Soon ...