Lightnews — Scholar-powered news

Meg Forrest 🌳🌲🌳

@meghanforr.bsky.social

73 followers 99 following 12 posts

Epi research associate + information specialist Former neuro critical care nurse Interested in stroke 🧠 epi research (with/&) causal inference methods she/her 📍 Berlin #EpiSky #Epidemiology #CausalInference #PublicHealth #AcademicSky

Posts Media Videos Starter Packs

Pinned

Meg Forrest 🌳🌲🌳 @meghanforr.bsky.social · Feb 6

📣 New paper!

We investigated the use of stacked proportional bar graphs (aka "Grotta bars") in observational neurology research & made infographics 🎨📊 to guide proper use & interpretation of these figures.

🔓 Read the OA paper: www.neurology.org/doi/10.1212/...

🙈 Spoilers below 👇

Neurology® Journals

www.neurology.org

1 4 6

Reposted by Meg Forrest 🌳🌲🌳

Peter Tennant @pwgtennant.bsky.social · 22d

Excited to be back in Berlin for the second gathering of the Einstein Circle on Causal Inference from Observational Health Data!

With @jlrohmann.bsky.social, @meghanforr.bsky.social, @chisatoito.bsky.social, @mpiccininni3.bsky.social, Venessa Didilez, Toivo Glatz, Rodrigo Huerta, & others!

#EpiSky

4 32

Reposted by Meg Forrest 🌳🌲🌳

Peter Tennant @pwgtennant.bsky.social · Jul 24

SAVE THE DATE: The 2026 IEA European Congress of Epidemiology and 70th @socsocmed.bsky.social Annual Conference will take place in London, UK on 8th-11th September 2026!

#EpiSky #EuroEpi2026

1 18 37

Reposted by Meg Forrest 🌳🌲🌳

Julia M. Rohrer @dingdingpeng.the100.ci · Feb 26

Causal inference iceberg!
What's missing?

Ice berg meme template. From top to bottom:
Correlation does not imply causation
Third variable adjustment
Just run an experiment
Confounding
Collider bias
Selective mortality
Observational longitudinal data
Mediation analysis is messed up
Posttreatment bias in experiments
Generalization as a causal inference problem
Missing data as a causal inference problem
Measurement as a causal inference problem
Causal foundations of applied probability and statistics

46 33 190

Reposted by Meg Forrest 🌳🌲🌳

Anne-Marie Hanff @ahanff.bsky.social · Jul 12

Before #CausalInferenceIntro:
"Hmm, this association looks unusual" 🤔

After #CausalInferenceIntro:
"Have you considered the possibility of collider bias?" 🤓

@pwgtennant.bsky.social @georgiatomova.bsky.social @jlrohmann.bsky.social @meghanforr.bsky.social
@laurieberrie.bsky.social
Rodrigo

4 4 11

Reposted by Meg Forrest 🌳🌲🌳

Peter Tennant @pwgtennant.bsky.social · Jul 10

70 people having a curry at the amazing Bengal Brasserie restaurant in Leeds City Centre!

My favourite part of the #CausalIntroCourse!

1 4 15

Reposted by Meg Forrest 🌳🌲🌳

Peter Tennant @pwgtennant.bsky.social · Jul 10

I *love* teaching collider bias with dice. And seeing the shocked (and angry) faces when the (seemingly) paradoxical results are revealed 😂

#EpiSky #CausalSky

Jess Rohmann @jlrohmann.bsky.social · Jul 9

No one teaches collider bias like @pwgtennant.bsky.social in the #CausalIntroCourse.

Lighthearted dice rolling 🎲 before the big reveal. Puzzled looks. Then 🤯 shock, as the implications for their research fields sank in. You could hear a pin drop during the compelling lecture that followed! 👏🏻

2 6 31

Reposted by Meg Forrest 🌳🌲🌳

Peter Tennant @pwgtennant.bsky.social · Jul 7

I love the Monday evening social! 😎 #CausalIntroCourse

2 6

Reposted by Meg Forrest 🌳🌲🌳

Society for Epidemiologic Research @societyforepi.bsky.social · Jun 3

We are very excited to be welcoming 1500 epidemiologists to Boston, next week for our conference!

Share your plans, and connect with other delegates, on LinkedIn or Bluesky using the #SER2025 hashtag or via the Whova conference app.

We are looking forward to seeing you all very soon!

#EpiSky

Simple flyer for the forthcoming #SER2025 conference

15 33

Reposted by Meg Forrest 🌳🌲🌳

Berlin Epi Methods Colloquium @bemcolloquium.bsky.social · May 2

NEXT TALK: We're excited to be joined next week by Martin Lajous, who will be presenting "Estimating effects on all-cause mortality in the presence of COVID-19 deaths"

📅 May 7, 2025
🕓 4pm CEST (Berlin time)
📍 hybrid (Charité Mitte or Zoom)

Registration links 👇

1 1 1

Meg Forrest 🌳🌲🌳 @meghanforr.bsky.social · Feb 28

We ♥️ Methods!
This afternoon, Corinna Dressler and I will host our talk “Evidence Synthesis: Types, reproducibility, and best practices“

⏰ 3 pm Berlin CET (9am US EST)
📍online (Zoom)

Info + registration: www.bihealth.org/de/aktuell/l...

#LoveMethods2025

#systematicreview #scopingreview #ebm

Meg Forrest 🌳🌲🌳 @meghanforr.bsky.social · Feb 6

🙌 Big shout out to my co-authors
Emma Lieske
Elena Tamayo Cuartero
Elena Fischer
Lydia Jones

🙌 Huge thanks to the team of supervisors & co-authors for the excellent guidance on this project - it was my master’s thesis!
@jlrohmann.bsky.social
@tweissgerber.bsky.social
@mpiccininni3.bsky.social

1 3

Meg Forrest 🌳🌲🌳 @meghanforr.bsky.social · Feb 6

In observational, causally-aimed neurology research, we investigated 🔎
❓ How often Grotta bars are used
❓ How often are they adjusted
❓ What adjustment methods are used

We also created infographics to help readers interpret these graphs, and guide authors while generating them

#stroke #causalsky

Meg Forrest 🌳🌲🌳 @meghanforr.bsky.social · Feb 6

When an observational study has causal aims, unadjusted Grotta bars will show confounded comparisons. This may lead to incorrect interpretation of observational study results.

Therefore…

1 1

Meg Forrest 🌳🌲🌳 @meghanforr.bsky.social · Feb 6

Simply showing a graph with observed outcome data only (unadjusted visualization) can be very misleading! Like we adjust the effect measures of interest for confounding, so too should we present adjusted versions of these bars alongside the crude/unadjusted ones.

#metaresearch #academicsky #episky

Figure comparing modified Rankin Scale (mRS) distributions for patients discharged home vs. not discharged home, before and after adjustment using inverse probability of treatment weighting.

The figure consists of two stacked bar charts:
• Panel A (Unadjusted): Shows the unadjusted distribution of mRS scores for two groups—patients discharged home and not discharged home.
• The discharged home group has a higher proportion of patients with lower mRS scores (0–2, represented by dark blue to light blue), while the not discharged home group has a greater proportion of patients with higher mRS scores (3–6, represented by yellow, orange, and red).
• The numbers inside the bars indicate percentages of patients in each mRS category.
• Panel B (Adjusted using inverse probability of treatment weighting): Shows the same comparison after statistical adjustment.
• The differences between groups are reduced, with more similar distributions of mRS scores between the discharged and not discharged home groups.

The color legend on the right corresponds to different mRS scores, ranging from 0 (dark blue, no symptoms) to 6 (dark red, death).

1 1

Meg Forrest 🌳🌲🌳 @meghanforr.bsky.social · Feb 6

Because randomization (ideally) works wonders, the contrast shown here is a causal one, and shows that tPA was (and still is) a very effective treatment. But what about when randomization is not possible and we are interested in answering causal questions using observational data?

Meg Forrest 🌳🌲🌳 @meghanforr.bsky.social · Feb 6

Grotta bars are a powerful tool to visualize and compare ordinal variables. They’re often used to depict functional outcomes in #neurology research. Perhaps the most memorable (and their namesake) is from the original #thrombolysis trials

www.nejm.org/doi/full/10....

Four stacked proportional bar charts comparing the outcomes of stroke patients treated with t-PA versus a placebo. Each of the four graphs has a different functional outcome (NIHSS, Barthel Index, mRS, or Glasgow Outcome Scale). Each chart represents patient percentages across different severity categories, comparing the intervention "t-PA" to placebo.

1 1

Meg Forrest 🌳🌲🌳 @meghanforr.bsky.social · Feb 6

Neurology® Journals

www.neurology.org

1 4 6

Meg Forrest 🌳🌲🌳 @meghanforr.bsky.social · Jan 8

Happening now! Hope you can join for Vanessa Didelez's talk "Statistical Methods for Causal Inference with Time-to-Event Data in Epidemiology" @bemcolloquium.bsky.social

Registration: eu01web.zoom.us/webinar/regi...

1 1

Meg Forrest 🌳🌲🌳 @meghanforr.bsky.social · Dec 14

It was such a nice visit!!

Peter Tennant @pwgtennant.bsky.social · Nov 29

Great #EpiSky Christmas album cover from my recent visit to the Charité Institute of Public Health!

With
@georgiatomova.bsky.social,
@jlrohmann.bsky.social, @meghanforr.bsky.social, @chisatoito.bsky.social,
@rhuerta.bsky.social,
@mpiccininni3.bsky.social,
Toivo Glatz, and others)

Picture of some epidemiologists stood outside a Christmas market!

1 1 5

Meg Forrest 🌳🌲🌳 @meghanforr.bsky.social · Dec 2

I was really lucky to have been able to participate in this class in January. It has a great program with great instructors!

Looks like spots are filling up quickly for the July 2025 course!

#EpiSky #CausalInference

Peter Tennant @pwgtennant.bsky.social · Dec 2

PLEASE SHARE: Registration is now open for the next Introduction to Causal Inference Course for Health & Social Scientists (7-11 July 2025, Leeds, UK).

See more info & register here: www.causal.training

Note, we are not planning any other courses until 2027.

#CausalIntroCourse #EpiSky #CausalSky

Flyer for the Introduction to Causal Inference Course for Health & Social Scientists, taking place on 7th-11th July 2025

1 3

Reposted by Meg Forrest 🌳🌲🌳

PubPeer @pubpeer.com · Nov 26

We’re encouraging meaningful commentary with $1,000 rewards for selected PubPeer comments. bsky.app/profile/pubp...

Now imagine if @hhmi.bsky.social , NIH, NSF, etc also awarded outstanding public reviews.

These could boost careers and CVs while building a vital layer of scientific evaluation.

PubPeer @pubpeer.com · Nov 25

And now for the announcement:
While many of our plans for the award funds will take time to implement, one can start immediately.
To thank our users and encourage scientific debate, we’re introducing $1,000 rewards for selected PubPeer comments!
Yes students and postdocs you read that right: $1000

2 25 43

Meg Forrest 🌳🌲🌳 @meghanforr.bsky.social · Nov 21

Super stoked for this talk! Hope to see some of you there next week :)

#EpiSky #CausalInference #Epidemiology #AcademicSky #PublicHealth

Berlin Epi Methods Colloquium @bemcolloquium.bsky.social · Nov 21

Announcement: Next week, Georgia Tomova will present "Understanding compositional data with fixed and variable totals using DAGs and data simulations". @georgiatomova.bsky.social

🗓️ Wednesday, November 27
🕔 16:00 - 17:30 Berlin time
📍 Hybrid (Charité Mitte or Zoom)

Registration links 👇
#EpiSky