Important: statin NNT scores

[skyhawk]

I have been trying to come up with a simplified explaination of NNT scores so that everyone from every walk of life can easily understand.

Here is my best shot:

If 100 people take a statin, and heart attacks are reduced from 2% down to 1%, thats a 50% relative risk reduction, even though only one person was spared a heart attack.

But 100 people had to take the drug for over 3 years to prevent just one person from having a heart attack. That is a 1% absolute risk reduction.

In other words, if you take the drug, 99 other people will have to take it with you for over three years ( lipitor) for just one person to benefit.

The number needed to treat for one person to benefit is 100.

NNT =100

[Gaelen]

NNT=Number Needed to Treat, a statistical calculation relevant to clinical trial results, can also be used/described in a couple of other ways, so a single definition may not be such an easy task. ;)

There's also pretty much no way around the math as part of the explanation, and dumbing down the math to 'simplify' it could lead to inappropriate conclusions.

Even the Wiki article had to insert some level of math:
http://en.wikipedia.org/wiki/Number_needed_to_treat
Another explanation and the math required to come up with the NNT statistical calculation (number needed to treat) statistics is right here:
http://www.ptjournal.org/cgi/content/full/80/12/1214

[skyhawk]

Thanks. I guess I was just looking for a simple way to explain the difference between realative risk reduction, absolute risk reduction, and NNT scores, so that everyone, no matter what level a person is on, they can understand. Its tough to keep the explaination simple.