Re: Absolute risk of HRT
From: Rafael Haciski MD (haciski@earthlink.net)
Thu Jul 18 23:07:04 2002
Interesting layout, but I am not sure I understand the NNT - number of pts
needed to be treated to prevent one case of WHAT?
--
Rafael Haciski, MD
Baltimore, Md
> From: arthurfree@aol.com (Arthur Freeland, MD)
> Reply-To: ob-gyn-l@obgyn.net
> Date: Wed, 10 Jul 2002 23:04:55 -0500
> To: Multiple recipients of list OB-GYN-L <ob-gyn-l@mail.medispecialty.com>
> Subject: Absolute risk of HRT
>
> (apologies for sending this twice, but the table was truncated the first
> time so I'm trying again with it pulled back a little) Arthur
>
> When looking at the risk of any intervention, it is helpful to look at
> the
> NNT (number needed to treat - i.e. the total number of people treated
> in order to save one life, or prevent one case) and the NNH (the number
> needed to harm)i.e. from table below: one would have to treat 250 women
> with combination HRT for five years to see ONE stroke due to treatment.
> These are an effective measure of absolute risk/benefit. Since the
> Wyeth letter had some absolute numbers it is possible to calculate the
> NNT and NNH. The only problem I saw with their numbers is that the
> relative risk was given with a denominator of five woman years (the
> duration of the study) and the absolute risk was given per single woman
> year. For direct comparison, I've calculated the NNT and NNH for FIVE
> years. (In parentheses are the absolute increased risk per woman year
> given in the Wyeth letter)
>
> Outcome Relative Risk NNH NNT
>
> Heart attack 1.29 286 (7/10,000/yr)
> Stroke 1.41 250 (8/10,000/yr)
> Breast Cancer 1.26 250 (8/10,000/yr)
> Thromboembolism 2.11 111 (18/10,000/yr)
> Colon Cancer 0.63 333 (6/10,000/yr)
> Hip Fracture 0.66 400 (5/10,000/yr)
>
> As you can see, the absolute numbers, even over five years, are very
> small but not negligible.
>
> --
> Arthur Freeland, MD
> Kirksville, Missouri
