Ian Shrier (@IanShrier) profile picture

Sport medicine physician with research interests in synthesizing information, decision making and injury treatment/prevention. Activity is an exercise in fun!

Montreal, Canada
Joined January 2015
Photo by Ian Shrier (@IanShrier)Photo by Ian Shrier (@IanShrier)
For those interested in the history of causal inference in four very different disciplines, and three commentaries about how these individuals changed the way we approach analyzing data.
Observational Studies is excited to announce our new special issue "Rebels with a Cause: Monologues from Heckman, Pearl, Robins, and Rubin": muse.jhu.edu/issue/48885 These fascinating monologues are followed by insightful perspectives by Didelez, Mealli, and Tchetgen Tchetgen
2
5
1
17
Observational Studies is excited to announce our new special issue "Rebels with a Cause: Monologues from Heckman, Pearl, Robins, and Rubin": muse.jhu.edu/issue/48885 These fascinating monologues are followed by insightful perspectives by Didelez, Mealli, and Tchetgen Tchetgen
2
68
10
218
New BJSM article suggesting Load was not part of StARRT framework for RTP decision making had important misconceptions.BJSM editorial policy does not allow Letters to Editor See SportRxiv (doi.org/10.51224/SRXIV.177) for my response and suggested corrections.
1
10
2
35
Recent arXiv pre-print describing how we should think about analyzing recurrent events (such as injuries), with or without competing events. This is ground-breaking methods research. Not for the faint of heart. Please share with your statisticians. arxiv.org/abs/2202.08500
2
6
Very good article explaining how to create a causal directed acyclic graph using a real-world example. Barnard-Mayers et al. 2022 jclinepi.com/article/S0895-4…
6
If you like to learn from podcasts, the Society for Epidemiological Research has 2 platforms: Epidemiology Counts caters to a general audience. SERious EPI caters to practicing epidemiologists.” epiresearch.org/serlibrary/s…
5
“causal diagrams” also important for prediction!! arxiv.org/pdf/2011.02677.pdf. “applied statistics as the coordinated merging of the three essentials of logic, causation, and probability to provide a transparent foundation for sound study design, analysis, and interpretation.”
1
Corrected link for previous tweet on journal politics. statmodeling.stat.columbia.e…
An interesting bizarre story about journal and author politics on tactics to get published. statmodeling.sta.columbia.ed…
Key point: One enters theory-based causal effects between all variables and all interactions. It does not estimate these. Papers need 100% transparency on all hypothesized causal relationships used. Parametric g-formula uses both data and simulations. 2/2
1
#Editors are seriously misguided when they insist on non-causal language for #observational studies when the intent is clearly causal (possible intervention): often the case in nutrition, public health, and clinical areas. https://nitter.cf/t.co/g0AJgKWOSk
1
14
34
Causal inference article on youth collision sports not affecting cognition later in life. doi.org/10.1093/aje/kwab177. We previously showed properly treated first concussion doesn’t causally increase risk of subsequent concussion. https://doi:10.1136/bjsports-2018-099104.
1
New Society for Causal Inference. sci-info.org/about-us/. Great cast of leaders in the field. Would be great if sports medicine professionals with interest in causality research get involved. Will up everyone’s game.
1
5
bjsm.bmj.com/content/55/2/10… says load management in elite football doesn't prevent injuries. They never recorded load in control group so don't know what "load management" was compared to. Essential to describe interventions in both treatment groups in order to interpret studies.
1
3
19
Last reference is the free online book by Hernan and Robins for those who want to start using causal inference methods hsph.harvard.edu/miguel-hern…. First section has no statistics! :-) end/
4
Here are the easy to follow references on causal inference approaches I mentioned at #IOCprev2021. “The book of why” (Judea Pearl) covers general concepts for anyone. Miguel Hernan’s MOOC at edx.org/course/causal-diagra… is great for researchers who need an intro. 1/
6
You might also like this preprint to better report injury rates. Current method of athlete-exposures should be dropped except for rare research questions. osf.io/preprints/sportrxiv/3…
A very interesting pre-print on issues with consensus statements in the sport/exercise medicine community + a proposal for a different consensus statement model (based on Supreme Court function) that embraces dissenting opinions + increases transparency. osf.io/preprints/sportrxiv/8…
2
3