Biostatistician at BC Children's Hospital Research Institute. Scientist who works with data. Opinions expressed are my own.
Vancouver, British Columbia
Joined December 2013
- Tweets135
- Following214
- Followers199
- Likes144
Mike Irvine retweeted
Forgive me, for I am about to Bayes. Lesson: Don't trust intuition, for even simple prior+likelihood scenarios defy it. Four examples below, each producing radically different posteriors. Can you guess what each does? Revealed in next tweet >>
Our new study modelling the urgent need for #naloxone in the US. Using a #Bayesian model we developed for BC we estimated the probability of its use during a witnessed overdose. naloxoneneededtosave.org provides estimates for each state @DanCoombsUBC @TraciCGreen1
NEW—Naloxone: substantial increase in life-saving #opioid overdose intervention urgently needed in almost every U.S. state, finds study in @TheLancetPH.
hubs.li/Q013Y3Cl0
Mike Irvine retweeted
Excellent discussion with @tedfieldglobal from @GlobalBC on our @ubcmedicine study – visits during the COVID19 peak pandemic @BCCHresearch
globalnews.ca/news/7445146/b…
Academic lemma: the number of requests and emails will go towards infinity as the date goes towards when you're supposed to go on holiday.
Mike Irvine retweeted
Our team @CDCofBC & @ubcspph in #Vancouver is recruiting a quantitative #epidemiologist with focus on #COVID-19 and a #postdoc infectious disease #modelling to work with COVID-19 response team @CDCofBC. #epitwitter. DM me if interested. #BigData #YVRjobs @UBCmedicine
Mike Irvine retweeted
Long time frames to detect the impact of changing COVID-19 control measures. Tag line: it takes longer than 2 weeks to see the impact of a change in physical distancing (or other NPIs)
with @MostlyStatic @CarolineColijn
medrxiv.org/content/10.1101/…
Mike Irvine retweeted
Pleased our work on the use of mixture density networks for emulation in epidemiology has recently been published in @PLOSCompBiol. Check it out: doi.org/10.1371/journal.pcbi… @MikeAIrvine @DeirdreHoll @QuentinCAUDRON
Mike Irvine retweeted
Yesterday's talk for those who missed it, or don't have Netflix: Real-time modelling of the 2020 coronavirus epidemic. youtu.be/dLVLEZMzIOk via @YouTube. Shoutout to @UBC cosmologist Douglas Scott for the invitation to speak in the physics colloquium.
I created a blog post that allows you to explore how interventions might impact the covid-19 pandemic sempwn.github.io/blog/2020/0…
Lots of talk at the moment surrounding the R0. I recently created an app to explore the concept for educational purposes only: sempwn.github.io/vaccine-dyn…
Great to see this published before I move from @CDCofBC Thanks to the whole team for getting this one out!
Mike Irvine's paper on PrEP efficacy for HIV prevention among GBMSM (with Mark and BCCDC folks) is available online: Predicting the impact of clustered risk and testing behaviour patterns... sciencedirect.com/science/ar… @CDCofBC @MikeAIrvine @CoombsLab @mpjgilbert
Mike Irvine retweeted
#UBC researchers are working to find solutions to the overdose crisis, "the public health challenge of our time.” bit.ly/2Yy9aKq
Mike Irvine retweeted
Overdose deaths would be 2.5 times as high without emergency harm reduction in BC. “This is among the first evidence to show that a combination of harm reduction and treatment interventions saves lives,” says #UBC researcher Mike Irvine ow.ly/43v850uxo0i
Mike Irvine retweeted
Rapid expansion of B.C. harm reduction services likely saved about 3,000 lives, province's Centre for Disease Control says. Study finds overdose death rate between April 2017 and December 2018 could have been as much as 2.5 times higher. Background: cbc.ca/1.5156279
Mike Irvine retweeted
B.C.'s scale-up of harm reduction initiatives (overdose prevention sites, naloxone distribution, OAT) believed to have averted 3,000+ deaths over 20-month period, according to new @CDCofBC
study: bit.ly/2WRoOj4
Mike Irvine retweeted
Overdose deaths down Q1 in B.C. but experts reserved in their optimism.
"While the numbers are down compared to this time last year, we have maybe become inured to the fact that these are still incredibly high rates."
theglobeandmail.com/canada/a…
Visualization of mathematical modelling can help in public understanding of health risks.
An average person with the measles is likely to infect 12 to 18 others in an unvaccinated population.
So what starts as an isolated case can rapidly become a full-blown epidemic.
53eig.ht/2scZHrQ
Mike Irvine retweeted
New blog post: "A Recipe for Training Neural Networks" karpathy.github.io/2019/04/2… a collection of attempted advice for training neural nets with a focus on how to structure that process over time