Jon Huang

@[email protected]
608 Followers
244 Following
5 Posts

Epidemiology, Biostatistics, Public Health, Environmental Health. Child health and well being. Assistant Professor @ University of Hawaii at Manoa + Adjunct Assistant Professor Duke-NUS Medical School.

Causal inference; biomarkers; population health; social, environmental, and policy determinants of health and health equity; bioethics.

He/Him. Views mine.

http://www.jonhuang.org

NameJon Huang
LocationSingapore
Twitterhttps://twitter.com/jon_y_huang
Websitehttp://www.jonhuang.org
Jon HuangNov 13, 2022

The reflex to indiscriminately stratify and adjust on covariates is strong in biomedical research. I face this so often, especially with clinician-scientist collaborators.

The analyses and models that arise from this, in fact, often do not well represent how exposures and treatments exert their effect in reality.

A #TargetTrial approach crystallizes when, where, and how co-variates should be treated.

Take this great illustration from Target Trial originator @MiguelHernan where he explains the study design choices for study on COVID vaccine booster effectiveness:

https://fediscience.org/@MiguelHernan/109331025655365201

#epidemiology #biostatistics #epitwitter #statstodon #epiVerse @epiVerse

Miguel Hernan (@[email protected])

Attached: 1 image 1/ Using observational data, we estimated the short-term effectiveness of a 4th dose of #COVID19 vaccine: 68% for hospitalization, 74% for death in persons over 60 (compared with 3 doses) https://nejm.org/doi/10.1056/NEJMoa2201688 And then we received an interesting criticism: "You overestimated short-term effectiveness against hospitalization/death because only people infected with #SARSCoV2 during follow-up can be hospitalized or die due to #COVID19 and you didn't restrict the analysis to infected people."

FediScience.org
Jon HuangNov 12, 2022

https://uxdesign.cc/mastodon-is-antiviral-design-42f090ab8d51

This article captures so much what I love about mastodon

Thank you @clive !

#mastodon #antiViral #SlowSocialMedia

Show threadJon HuangNov 11, 2022

@simon

So maybe most productive is to just kind of overview epi causal inference domains/thrusts.

I'll put some names, but will be unavoidably biased towards colleagues / folks I know better. I'll edit later with tags or other names as I think:

1. Econ adjacent: A number of us that use RD, IV, DiD & collab w econs, often in familiar topics: policy, injury, ID, child development, etc. Sam Harper comes to mind, many others

2. Structural/mediation/sequential treatment: this is prob the largest most heterogenous group with development from HIV, cardiovascular, perinatal, psych, trying to understand how to deal with mechanisms/pathways or "channels" via your "bad controls." Methods are probably most foreign to econs (until double ML): marginal structural models, gformula, structural nested models. There are connections to trialist methods to recover effects from flawed trials, e.g. principal strata as well as, separately, target trials. Our Epi Star Ellie Murray could tell you more. I would also put genetic IVs here bc most have realize the assumptions are largely implausible. See work from fellow econ Carlos Cinelli.

3. ML age: Lines are perhaps most blurred now between the quant (epi/econ) and engineering disciplines. There is convergence on Double ML, policy learning, and mechanism learning from existing trials. Kara Rudolph for the latter. Related is ML synthetic controls for climate/enviro health effects.

This is a start! 🙏🏽

Jon HuangNov 8, 2022

I am still learning how to find my old twitter people over multiple instances and worrying about missing conversations given the whole "no alogorithm" thing, but slowly realizing:

was I actually seeing everything on twitter through my usual behavior (scrolling the TL), anyway? No!

With the mastodon web app, I can pin the hashtags I want and see ALL those posts and ONLY those posts without having to scroll through all the muck (unless you want to)! Truly a blessing!

Jon HuangNov 1, 2022

#introduction

Hi I'm Jon, my job is to work with data from studies that follow families over time (cohorts) to figure out what we need to do for child and family health and well-being.

My passion in this is to get both researcher and knowledge users to use data and analyses properly, including ideological blind spots like the pitfalls of over-interpreting molecular data (even when it's not genetics) or how research needs to move beyond blaming moms for the success of their kids

I am also a dad of two young kids and thus have no additional hobbies other than being online 😅

#publichealth #epidemiology #datascience #biostatistician #biostat #childhealthandbehaviour