This is an interesting and, to me, new definition of Number
Needed to Harm (NNH)
I use NNH quite differently as one over the difference
in adverse event (side effect) rates and I think it is
much more useful this way (e.g., one person can both improve
and have a side effect -- or just one or, of course, neither.
I would prefer to see the situation that Ronan is talking
about as one where neither treatment is clearly better
(from just the benefit side)
Rich Goldstein
Ronan Conroy wrote:
On 1 Aug 2007, at 13:55, Svend Juul wrote:
NNT = 15.230769 (95% CI: 5.0701399; -15.169883)
- the negative numbers meaning that exposure to x may prevent low
birthweight.
The negative number has the interpretation of Number Needed to Harm
(with its minus sign removed), just as a negative Attributable fraction
is more usually written as "Prevented fraction in the exposed/population"
Thus the confidence interval in Svend's example runs from an NNT of 5
to an NNH of 15.
The other useful statistic that I recommend is the NTN (number treated
needlessly). It is the NNT-1. So if you have to treat 400 people to
prevent one event, then the number treated needlessly is 399. A
salutary reflection for the average doctor.
P Before printing, think about the environment
=================================
Ronan Conroy
[email protected]
Royal College of Surgeons in Ireland
120 St Stephen's Green, Dublin 2, Ireland
+353 (0)1 402 2431
+353 (0)87 799 97 95
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
