@_miguelhernan - Twitter Profile Analysis

Analysis of 200 tweets by Miguel Hernán, from 23 Sep 2020 to 19 Jan 2022.

@_miguelhernan twitter profile photo

Miguel Hernán

verified twitter profile @_MiguelHernan

Researcher, teacher. Using data to learn what works to improve health. Trying to make #causalinference less casual. Professor @HarvardEpi. Director @CAUSALab.

896 tweets
37,803 followers
328 favourites
<1 tweet per day
2015 since
183 following
665 listed
Boston location

Top Tweets by @_miguelhernan

Miguel Hernán
@_MiguelHernan
📌A New Year's resolution for all researchers: Let's make 2022 the year in which scientific journals said goodbye to statistical significance. We just won't type "statistically significant" in our papers, even if asked by clueless reviewers or editors. https://t.co/PvVD99YnMJ
31 Dec, 2021 - 02:48 PM UTC
Miguel Hernán
@_MiguelHernan
Gestionar esta pandemia es difícil, pero hay algo en lo que todos deberíamos estar de acuerdo: Los colegios son lo último que se cierra y lo primero que se abre. Si no hace falta cerrar bares mal ventilados, menos aun privar a los niños de su educación. https://t.co/sV5p1wmomB
29 Dec, 2021 - 11:30 AM UTC
Miguel Hernán
@_MiguelHernan
There're so many "multiply robust" estimators for the #gformula. No wonder it's confusing. Led by Lan Wen @CAUSALab, we describe the relation between g-formula estimators for survival outcomes, study their robustness, and construct more efficient versions https://t.co/6s7NY0X5yh
27 Dec, 2021 - 07:32 PM UTC
Miguel Hernán
@_MiguelHernan
Un titular que lo resume todo. ¿Dónde están los informes técnicos que apoyan la posición de cada presidente (central y autonómicos)? ¿En qué han basado sus decisiones desde 2020? ¿Dónde está la Agencia Española de Salud Pública? Basta de opiniones. https://t.co/uEjO5FeX0i
22 Dec, 2021 - 07:14 PM UTC
Miguel Hernán
@_MiguelHernan
The g-null paradox threatens parametric #gformula estimates. But the paradox is not about testing the causal null. The g-null paradox is a form of model misspecification that may occur even if treatment has an effect. Led by Sean McGrath, we explain 👇 https://t.co/pOGYHAJJIV
17 Dec, 2021 - 03:41 PM UTC
Miguel Hernán
@_MiguelHernan
1/ We emulated a target trial of two #COVID19 mRNA vaccines in the largest healthcare system in the US. Both vaccines were similarly effective, with Moderna slightly better than Pfizer-BioNTech. But that isn't the most important conclusion of our study. https://t.co/u8rW8sQpJE
03 Dec, 2021 - 02:40 PM UTC
Miguel Hernán
@_MiguelHernan
"A room, a bar and a classroom: how the coronavirus is spread through the air" The @AAAS Kavli Science Journalism Award goes to @Mariano_Zafra and @javisalas. The multimedia story anyone interested in indoor #SARSCov2 transmission had been waiting for. https://t.co/yFyALusOwJ
10 Nov, 2021 - 05:01 PM UTC
Miguel Hernán
@_MiguelHernan
Shame on BMJ for irresponsible reporting. The serious data problems found with Ventavia's management of Pfizer's vaccine trial: likely make the vaccine look worse than it actually is affect 3 of 153 sites You need to lead with this information, @bmj_latest. So disappointing. https://t.co/5PweQQ0LTF
03 Nov, 2021 - 05:25 PM UTC
Miguel Hernán
@_MiguelHernan
Just out: Largest peer-reviewed evaluation of #COVID19 vaccine booster effectiveness. We compared 3 vs. 2 doses of Pfizer/BioNTech. The estimated effectiveness of the third dose for COVID-19-related hospitalization was 93%. Benefit in all age groups. https://t.co/O1n2xr2VZ2
29 Oct, 2021 - 10:42 PM UTC
Miguel Hernán
@_MiguelHernan
What does a country with access to #COVID19 vaccines need to ensure that its population gets vaccinated? Two things: Trust and a good healthcare system. The case of #Spain explained in 8 minutes https://t.co/iZiI6ZfKG4
17 Oct, 2021 - 02:56 PM UTC
Miguel Hernán
@_MiguelHernan
Randomized trials and observational studies are complementary—not competing—approaches to #causalinference. Here is an example of how #observational effect estimates were used for provisional decision making until #randomized estimates became available 👇 https://t.co/7leznx1xbN
06 Oct, 2021 - 01:37 PM UTC
Miguel Hernán
@_MiguelHernan
Of 50 most populated countries, Spain first to reach 80% #COVID19 vaccination coverage. 11% of population are children yet ineligible for vaccination. Restrictions mostly lifted. Incidence low and declining. Given global vaccine shortage, unclear vaccinating children is optimal
04 Oct, 2021 - 02:20 PM UTC
Miguel Hernán
@_MiguelHernan
Statistical power calculations are often required for causal analyses of observational databases. They shouldn't be. Refuse to play the power game. If a causal question is important enough, low power can't be an excuse to do nothing. Analyze your data👇 https://t.co/sNS5izVT28
01 Sep, 2021 - 09:58 PM UTC
Miguel Hernán
@_MiguelHernan
Spain: ~70% population fully vaccinated. UK/Israel not far behind. They vaccinated the first 30% months before Spain. #COVID19 rate 2.5 times higher in UK & 4 times higher in Israel than in Spain. Waning vaccine effect? Much higher natural immunity in Spain? @OurWorldInData
28 Aug, 2021 - 08:37 PM UTC
Miguel Hernán
@_MiguelHernan
1/ Vaccine safety: We compared excess adverse events after #COVID19 vaccination (Pfizer-BioNTech) and after documented #SARSCoV2 infection. https://t.co/PNyCTqkq0t Take-home message: Low excess risk of adverse events after vaccination, higher after infection. Some thoughts👇
25 Aug, 2021 - 10:19 PM UTC

Top Retweets by @_miguelhernan

CAUSALab
@CAUSALab
Start off the New Year right. Re-watch last year's most exciting event. The CAUSALab Inaugural Event has arrived to YouTube. First up is @barbradickerman: https://t.co/AtfRyqUazL Subscribe for more.
19 Jan, 2022 - 04:40 PM UTC
CAUSALab
@CAUSALab
New from CAUSALab: "Effectiveness of Localized Lockdowns in the COVID-19 Pandemic" A novel synthetic control approach suggests that local lockdowns cannot control pandemic growth in the presence of effects from neighboring areas that don't have lockdowns. https://t.co/XFH1MSM2q4
14 Jan, 2022 - 09:21 PM UTC
Harvard Epidemiology
@HarvardEpi
Do you love working with code/programming, big data, and public health? Is SAS or R your second language? Come check out this Senior Programmer role with @CAUSALab in the Department of Epidemiology @HarvardChanSPH! https://t.co/JAYLcx4pF5
13 Jan, 2022 - 10:30 PM UTC
CAUSALab
@CAUSALab
Matching is widely used for confounding adjustment, but the matched population may end up being too small and unrepresentative. @bijan_niknam and José Zubizarreta at the CAUSALab explain how to optimize matching to address those problems. @JAMA_current https://t.co/7xawakuAHW
11 Jan, 2022 - 08:13 PM UTC
CAUSALab
@CAUSALab
Interested in #causalinference? Learn from top experts in the field. Summer courses offered at the Harvard T.H. Chan School of Public Health in June, 2022. 👉Key Topics in Causal Inference 👉Target Trial Emulation Stay tuned for more details. https://t.co/9SrCKvKSd9
20 Dec, 2021 - 05:50 PM UTC
daniele grasso
@danielegrasso
La entrevista de @oriol_guell a @_MiguelHernan da en el clavo: - Los antígenos deberían ser casi gratis y hacerse 3 veces por semana - La campaña de la 3ª dosis es floja - Se ha cronificado la presión hospitalaria - Falta Agencia de Salud Pública autónoma https://t.co/fWXrWuHrD5
19 Dec, 2021 - 12:22 PM UTC
Clara Pons Duran
@ClaraPonsDuran
I highly recommend the course on Causal DAGs from @_MiguelHernan @CAUSALab to anyone working in #epidemiology who wants to draw their assumptions and better understand causality. I really enjoyed it, and most important, I learned a lot! Available at https://t.co/h6yUyAhbZ2
06 Dec, 2021 - 01:27 AM UTC
CAUSALab
@CAUSALab
BREAKING @NEJM: Head-to-head comparison of two #COVID19 vaccines: Pfizer-BioNTech vs. Moderna. A collaborative project of investigators @CAUSALab and @VAResearch in Boston. When randomized trials don't exist, observational data can be used to emulate a #TargetTrial 👇 https://t.co/RswwRUng5h
01 Dec, 2021 - 10:54 PM UTC
Joy Shi
@joy_shi1
New paper out in @BMC_series! We propose methods for instrumental variable (IV) analyses of time-varying exposures and survival outcomes (with a focus on applications to Mendelian randomization studies) 👇: (1/n) https://t.co/B8yEoIEtW0
27 Nov, 2021 - 01:00 AM UTC
Oxford Medicine
@OUPMedicine
“To me, causal inference is not a separate set of tools for epidemiologists, because most of what epidemiologists do is causal inference anyway”: @_MiguelHernan discusses his influential 2002 paper. More @AmJEpi interviews here: https://t.co/rqlkAMLxwQ https://t.co/I7HHw6svIU
16 Nov, 2021 - 02:05 PM UTC
CAUSALab
@CAUSALab
The CAUSALab and @HarvardEpi invite you to this year's first Kolokotrones Circle Seminar open to the public: George Davey Smith: "Pre-registered triangulation of evidence." Please join us on Nov 22 virtually or in person. Registration open until Nov 19: https://t.co/BS5BOYdAKb
12 Nov, 2021 - 10:17 PM UTC
CAUSALab
@CAUSALab
If you're interested in #causalinference from real world data, join hundreds of colleagues who have already registered for the CAUSALab Inaugural Event on November 5. In-person registration almost sold out. Online registration extended through November 2. https://t.co/kX5lQ45Yml https://t.co/rbJgvsJ5ou
28 Oct, 2021 - 09:47 PM UTC
CAUSALab
@CAUSALab
Interested in comparing effect estimates from observational studies and randomized trials? Then you're interested in #benchmarking observational analyses that emulate a #TargetTrial. Don't miss Issa Dahabreh's talk. https://t.co/faOr54HpH0
23 Oct, 2021 - 11:57 AM UTC
Javier Muñoz Laguna
@JMunLaguna
"Explicit #targettrial emulation increases the transparency and replicability of observational effect estimates." @_MiguelHernan https://t.co/kuW9szE6dt
03 Oct, 2021 - 12:04 PM UTC

@_miguelhernan's

Tweets

57 tweets
37 retweets
106 replies
@_miguelhernan's

Twitter Client

182 Twitter Web App
18 Twitter for Android
@_miguelhernan's

Tweet Times

@_miguelhernan's

Tweet Days


@_miguelhernan's

Retweets

11 - CAUSALab
2 - EpiEllie
1 - barbradickerman
1 - NicoleBensonMD
1 - goodmanmetrics
1 - kerinalthoff
1 - cdsamii
1 - AlfredoMorabia
1 - H_Rossman
@_miguelhernan's

Replies

80 - _MiguelHernan
2 - MaartenvSmeden
2 - mendel_random
1 - NoahHaber
1 - maths_gee
1 - juanrichi
1 - dingding_peng
1 - H_Rossman
1 - califf001
@_miguelhernan's

Hashtags

22 - #COVID19
7 - #causalinference
6 - #observational
5 - #TargetTrial
5 - #SARSCoV2
4 - #gformula
3 - #randomized
2 - #IPweighting
1 - #hydroxychloroquine
@_miguelhernan's

Mentions

3 - @OurWorldInData
2 - @NEJM
1 - @Biometrics_ibs
1 - @SASsoftware
1 - @_tony_matthews
1 - @KatalinGemes
1 - @berglund_anita
1 - @karolinskainst
1 - @ADAlthousePhD