🧵 THREAD on BIAS in MEDICINE #MedTwitter #FOAMed

1a/

BIAS:

➡️ is integral to how we see people/make diagnoses: can't be removed

➡️ affects research questions that get asked, who gets funded, what gets published and in which journal

@FutureHealthJ: rcpjournals.org/content/future…
1b/

Bias training:

➡️ increases bias awareness

➡️ doesn't decrease biases (no evidence base)

➡️ bias training ignores communication, provides no debiasing interventions, doesn't measure motivation

➡️ doesn't address structural problems

Read: ncbi.nlm.nih.gov/pmc/articles/P…
2/

Shoutout to the amazing team:

@UlaChetty

@patrickod9

@returnofcam

@jodieblackadder

And Prof @damienridge who shaped the manuscript.

Hundreds of papers, hours of lectures/books, many drafts to get here.

Thank you 🙏🏽
3/

We saw outside medicine everyone was talking about bias. But we were not. Why?

Organisations make bias training mandatory for staff: qz.com/1144504/the-wo…

In California training is COMPULSORY for people-facing staff: leginfo.legislature.ca.gov/faces/billNavC…

Can they decrease biases?
4a/

2 types of bias:

➡️ Implicit: we are not aware of (fast thinking) - thought to be necessary to make decisions but problematic as thought to cause 'systemic inequalities', e.g. unequal pay/treatment

➡️ Explicit: aware of (slow thinking) - overt prejudice/discrimination
4b/

How does bias work?

➡️ Type 1: 'fast' implicit, unconscious processes due to learnt experiences/emotions/etc

➡️ Type 2: 'slow' explicit, conscious

💡 Why don't we all think slowly? - we wouldn't function as people / society. We process lots every second w/o thinking.
4c/

It is thought that you can measure your biases by taking a test @ProjectImplicit:

implicit.harvard.edu/implicit/

BUT whilst it is used widely, some psychologists don't believe it measures bias. Is it automation or perception?

How do you measure the unconscious mind? 😳
5a/

Why is it important now?

Politicians might buy up ventilators than implement public health measures - doing something is better than nothing (action bias)

Similar to experimental treatment at the beginning of pandemic - doing something is better than nothing (action bias)
5b/

People talk about bias training since resurgence of the #BlackLivesMatter movement.

This ignores wider structural problems that individuals face.

Problematic data exclusions:

Skin:
blackandbrownskin.co.uk/mindthegap

Underestimation of Oxygen sats: nejm.org/doi/10.1056/NE…
5c/

@crageshri @Dr2NisreenAlwan @soniakumar100

Problematic data inclusions:

Using race as a biological category in algorithms may very well perpetuate health inequalities:

nejm.org/doi/full/10.10…

nejm.org/doi/pdf/10.105…
6a/

Examples in medicine:

Huge anti-obesity/fat bias - attributes to lack of willpower, ignoring poverty/insomnia.

It's what Dr @DrSdeG has been talking about for ages - in numbers.

journals.plos.org/plosone/articl…

onlinelibrary.wiley.com/doi/10.1002/ob…

nature.com/articles/ijo20…
6b/

Lots of data to back up female doctors being mistaken for nurses. (I'm not saying anything wrong with being a nurse.)

liebertpub.com/doi/full/10.10…

jamanetwork.com/journals/jamai…
6c/

Data to suggest female doctors are thought to be subjectively less competent despite same objective competency

onlinelibrary.wiley.com/doi/full/10.11…

journals.lww.com/greenjournal/f…
6d/

Thanks @CCriadoPerez @DrNighatArif

Women are less likely to be believed when they have myocardial infarctions: ahajournals.org/doi/full/10.11…

Not mentioning worse mortality when in hospital:

onlinelibrary.wiley.com/doi/full/10.10…

ahajournals.org/doi/full/10.11…

sciencedirect.com/science/articl…
6e/

Causes of racial disparities in maternal mortality: npeu.ox.ac.uk/mbrrace-uk/rep…

Could cultural competency help here?

cochrane.org/CD009405/COMMU…

cc @teekayoki @KMeghrawi @MelaninMedics
7a/

Bias affects which research gets funded:

Small institutions > large institutions: journals.plos.org/plosone/articl…

AND funders when unblinded prefer large institutions in the global North

healthaffairs.org/doi/10.1377/hl…

…balizationandhealth.biomedcentral.com/articles/10.11…

…rchintegrityjournal.biomedcentral.com/articles/10.11…
7b/

If research only comes from a certain number of institutions, and so do those peer reviewers and journal editors...not mentioning the actual research questions

We are missing so much research because we believe that research from certain people/places is better than others.
7c/

@angeladsaini "Science is always shaped by the time and the place in which it is carried out."

Science is not free from outside influence including society.

Shoutout to @MedCrisis who covers bias in the actual research in much more detail:
8/

How about policy? @justinparkhurst: library.oapen.org/bitstream/hand…

➡️ Technical bias: data ‘cherry-picking’

➡️ Issue bias: use of evidence shifts political debate in a certain direction, e.g. presenting policy with evidence from one side of debate
9a/

💡 Is bias linked to diagnostic error ❓ Short answer can't tell as the best data is from surveys

bmcmedinformdecismak.biomedcentral.com/articles/10.11…

But clearly checklists have been shown to show human factors are important
9b/

💡 How do clinicians make diagnostic errors ❓

Here's a list:
10/

💡 Does bias training work in medicine ❓

NO - maybe only to increase awareness but not to decrease biases

qualitysafety.bmj.com/content/25/10/…

Outside medicine - does increase awareness: @DrDoyinA @EHRC

equalityhumanrights.com/en/publication…

@KUMEC4KCL @juheepahuja
11/

Why does bias training not work?

We don't:

➡️ measure motivation

➡️ give people ways to decrease biases

➡️ give people good non/verbal communication training with bias training: communication is the mediator between bias and patient outcomes.

ncbi.nlm.nih.gov/pmc/articles/P…
12/

Dilemma:

Bias is problem AND we can't really decrease bias

So we teach it better? (tweet above)

Maybe @ProjectImplicit - with a caution too.
13/

Thank you the thousands of researchers working on bias overall.

Our team pulled everything together that's all.

@helenamacarena @TheKingsFund @yvonnecoghill1 @whynotadoc @lexmakesthings - your/appearances on podcasts is cited. 🙏🏽

ENDS
@threadreaderapp unroll please, thank you

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with Dipesh Gopal

Dipesh Gopal Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @DipeshGopal

20 Mar
🧵 on #COVID19 in Pregnancy / Breastfeeding

1/

Prof @thangaratinam:

Pregnant women at increased risk of severe COVID (ITU admission), thus earlier delivery

Risk factors: obesity, advancing maternal age, non-white ethnic origin #GlobalPregCov

🔗: bmj.com/content/372/bm…
2/

Animation here:

Read more about the research here: birmingham.ac.uk/research/who-c…

#GlobalPregCov
3/

Prof Paul Heath @StGeorgesUni

Efficacy in slide in general population:

Pfizer/Moderna/AstraZeneca/Gamaleya from clinical studies.

Novovax, Janssen, Sinovac, Bharat Biotech from press release.

#GlobalPregCov
Read 12 tweets
19 Mar
🧵 THREAD on "good men" #NotAllMen #MeToo

🗣️ What is the pass we give ourselves when we call ourselves "good men"?

Full video from @wayfarerstudios @justinbaldoni:

feat. @TonyPorterACTM @acalltomen @jameyjaz
🗣️ The language used to objectify women is problematic.

Full video from @wayfarerstudios @justinbaldoni:

featuring @TonyPorterACTM @acalltomen
🗣️ What should men do? (watch from 56 mins)

Listen
Believe women
Don't use assumptions
Interject in the locker/boardroom

Be intentional about educating boys/young men

Set an example

Look inwards

Full video from @wayfarerstudios @justinbaldoni:
Read 4 tweets
3 Mar
🧵

1/

THREAD on @BMJLeader study: bmjleader.bmj.com/content/early/…

➡️ Gender and racial CONFERENCE PANEL KEYNOTE SPEAKER representation at Royal College conferences 2015-2019

➡️ Publicly available data: 20 of the 70 conferences
2/

At 60% (n=12) of conferences in our sample, there
was a predominance of male speakers.

Median percentage of female speakers: 35 - 46%.

In 15% (n=3) of conferences, there were an equal number of male and female keynote speakers.

No #manels detected! 🤯 @medicalwomenuk
3/

At 20% (n=4) of conferences examined, none of the keynote speakers were from ethnic minorities.

Median % ethnic minority speakers 9 - 18%.
No increase from 2015 to 2019.

@crageshri @RammyaMathew @Dr2NisreenAlwan @DrRakN @DrJessPotter
Read 10 tweets
25 Feb
🧵THREAD about public/healthcare funding

1/

🗣️ "We must not resort to infighting...instead of placing the focus of our rage where it deserves to be placed – at the feet of those who have been managing the NHS in recent years."
2/

It's not only mismanagement. It's consistent underfunding.

When you look at the govt's own figures, it looks like public spending is increasing: gov.uk/government/sta…

Note the y-axis is £ billions but not accounting for inflation (nothing about real terms).
3/

When you look in detail it looks like health spending is increasing and social care is constant.

Note the y-axis is £ billions but now IS accounting for inflation.
Read 12 tweets
4 Jan
1/

🧵 THREAD about probably one of the BEST discussions about social value of work, economic wealth and fairness in a while.

40 mins with @amolrajan, @PJTheEconomist @elerianm @KGerlich777 Louise Casey

@davidgraeber 🗣️ "The more your job helps others, the less you get paid"
2/

Reminded me of @euan_lawson for @BJGPjournal quoting Michael Sandel on meritocracy: bjgp.org/content/70/701…

🗣️ "In an unequal society, those who land on top want to believe their success is morally justified. In a meritocratic society, this means the winners must believe...
3/

🗣️ "...they have earned their success through their own talent and hard work.… at a time when racism and sexism are out of favor (discredited though not eliminated), credentialism is the last acceptable prejudice."
Read 10 tweets
28 Nov 20
2/

Editors' note: Dr @HDambhaMiller (Editor in Chief), Dr Umar Chaudhry (@StGeorgesUni), Dr Peter Adams (@uwicave_hill)

They explain how they will be taking on diversity/inclusion by looking at:

outreach / mentoring students
diversity at submission

bjgpopen.org/content/early/…
3a/

Our group work looks at how race and racism affects the working life of a GP: bjgpopen.org/content/early/…

When it comes to issues of race / racism, is it comfortable to ignore them?

@v_silverwood @EbMulla @olamide_dada @salmanWaqar
Read 39 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!

Follow Us on Twitter!