Relative risk versus Absolute risk reduction

Have you seen the viral post starting with “Plot twist: Numbers don’t lie?” The post then states we should be using absolute risk instead of relative risk…and, that the absolute numbers show the vaccines don’t work. And, there’s a lot of fancy math that is hard to follow. Confused? Me too.


✅ Short answer: There’s no plot twist. The numbers are not lying.


✅ Long answer – Here’s the details: ◾ The post going around states the vaccine efficacy data (the 94 and 95% numbers we have all heard about) for Pfizer and Moderna are the wrong numbers to look at – instead, the post says we should be calculating the absolute risk reduction.

◾ The post uses the wrong formula to calculate the 95% number.

◾ Then the post does more fancy gymnastics to calculate the absolute number – and ends up with 1.16%. (BTW, this number is calculated wrong in the post too).

◾ The post claims that we should be using the absolute risk reduction number which is only 1.16%.

◾ The post then concludes “How is almost every drug and vaccine study reported? Relative risk”. That ending is supposed to be a mic’ drop to say scientists are using the wrong number to show the vaccine works.

◾ Therefore, the post concludes the true vaccine effect is actually the low 1.16% number, not the 95% number. Ugh.


✅ Whoa. Uhm, get ready for some learning, web-peeps. Welcome to epidemiology 101.


✅ First some definitions.

1) Relative risk is the comparison between the placebo group RELATIVE to the vaccine group. It’s kindof like a ratio through division.

2) Absolute risk is the absolute difference between the groups. Like an addition or subtraction problem.


BOTH measures are calculated in epidemiology. BOTH measures are important and tell a certain part of the epidemiology-story. BOTH measures have value in certain studies and you have to use them appropriately when drawing conclusions in science. This depends on your study design, sampling methods, and endpoints – all fun epidemiology stuff.
For vaccine trials, both measures are important. BUT the most meaningful number is the relative efficacy numbers, not the absolute risk reduction.

✅ TAKE HOME MESSAGE: The vaccine efficacy is legit at 94-95%. No one is lying about that by not using or reporting the absolute risk reduction. The correct numbers were reported. There’s no plot twist.


-FNE

***I’m sorry if this post feels vague if you haven’t seen the viral bad-data post. I’m trying to not share that stuff even as a screenshot to do my part in reducing bad information.

This Post Has One Comment

  1. Dr Gary H Kramer

    it is regrettable that some people seemed to WANT to believe the worst, and then twist the facts to fit their distorted worldview.

Leave a Reply