Read on Twitter
A nice opportunity for me to show the usefulness of this:
https://twitter.com/dnunan79/status/1179345857470222336?s=20
It may surprise folk to know that many modern-day medical interventions [including 'ahem' exercise) are far from miracles (see examples in pic). But NNT of 8 & 25 is right up there w the best.
/1 https://twitter.com/CT_Bergstrom/status/1273021731817443328
https://twitter.com/dnunan79/status/1179345857470222336?s=20
It may surprise folk to know that many modern-day medical interventions [including 'ahem' exercise) are far from miracles (see examples in pic). But NNT of 8 & 25 is right up there w the best.
/1 https://twitter.com/CT_Bergstrom/status/1273021731817443328
Borrowing* figures from: https://twitter.com/dongdon49026332/status/1273019551769378816?s=20
(*herein lies one of the problems with press-released data raised by others:
https://twitter.com/MaartenvSmeden/status/1273131848969969667?s=20)
We can use @d_spiegel tool to explore how we can communicate the #RECOVERYtrial findings
https://understandinguncertainty.org/node/233
/2
(*herein lies one of the problems with press-released data raised by others:
https://twitter.com/MaartenvSmeden/status/1273131848969969667?s=20)
We can use @d_spiegel tool to explore how we can communicate the #RECOVERYtrial findings
https://understandinguncertainty.org/node/233
/2
Let's look at the group of patients who required 02 but not ventilation:
We first determine the proportion (risk) of people that died in the group NOT receiving dexamethasone (dexa) =
2750/550 = 0.20 (20%). Note that 80% of these people survived with no dexa anyway.
/3
We first determine the proportion (risk) of people that died in the group NOT receiving dexamethasone (dexa) =
2750/550 = 0.20 (20%). Note that 80% of these people survived with no dexa anyway.
/3
Next we determine risk of death (in time frame of the study period) in the group treated w dexa =
1375/220 = 0.16 (16%). So 84% of people in the dexa group survive, an increase of 4% compared w the no dexa group. That's one way to communicate the findings
There are others
/4
1375/220 = 0.16 (16%). So 84% of people in the dexa group survive, an increase of 4% compared w the no dexa group. That's one way to communicate the findings
There are others
/4
Here's a cates plot that shows what happens when we give dexa to 100 people (this is known as natural frequencies & is helpful to visualise the impact on people as opposed to %).
4 people (yellow) on dexa receive the benefit.
/5
4 people (yellow) on dexa receive the benefit.
/5
However, you will see that 80 people (green) will not have died anyway, and 16 people (red) will still unfortunately die (in the study time period) even when taking dexa.
The risk tool allows you to visually display the findings in different ways - here's a column chart.
/6
The risk tool allows you to visually display the findings in different ways - here's a column chart.
/6
The tool also allows you to express the findings using text only, and to frame the findings negatively ('experiencing the event') or positively ('being free from the event) - I really like this as it changes the whole perspective
Here is an example of positive framing
Here is an example of positive framing
And here's the same findings framed negatively:
/8
/8
