Number needed to treat vs Number needed to harm - same right???

[aboveliquidice]

Just a bit of confusion concerning NNT or NNH...

NNT is 1/ (Probability of control - Probability of intervention)

If it comes out as a negative number... it is no longer considered an NNT correct? It would then be termed an NNH or number needed to harm. Is this correct, or do you see negative NNTs (which makes no sense to me) out there???

---

Members don't see this ad.

 

[Farmtastic]

I think this is correct.

But I thought it's NNT = 1/(Exp. grp rate - control grp. rate)

 

[aboveliquidice]

I think this is correct.

But I thought it's NNT = 1/(Exp. grp rate - control grp. rate)

I thought you assumed the intervention is for the better... thus the incidence with the control group would be higher.

If you take the absolute value (which is the crux of my question) - the value is the same.

 

[BMBiology]

Number needed to treat (NNT): The number of patients needed to treat with a specified therapy in order for one patient to benefit from treatment. The NNT is the inverse of the absolute risk reduction (1 divided by absolute risk reduction).

 

[Farmercyst]

Just a bit of confusion concerning NNT or NNH...

NNT is 1/ (Probability of control - Probability of intervention)

If it comes out as a negative number... it is no longer considered an NNT correct? It would then be termed an NNH or number needed to harm. Is this correct, or do you see negative NNTs (which makes no sense to me) out there???

If the number's negative you've got a worthless drug. (assuming the only way to get a negative number is to have a control that does better than your experimental.

As for NNH I believe you take the negative outcomes and do the same thing as opposed to the positive experimental outcomes. (How many heart attacks caused by Vioxx vs How patients with increased pain-free days)

 

[J ROD]

Dont worry about it....just get to the conclusion and discussion and forget the stats.....👍