sample size for sensitivity/specificity

[Evie]

I have a clinical test (positive/negative) and surgery (disease/not-
diseased) and plan to calculate sensitivity/specificity. How do I
estimate the sample size? Any advice would be greatly appreciated.

[Steve Simon, P.Mean Consulting]

The traditional sample size formulas based on power don't work here. The
simplest thing is to select a sample size that insures that the
confidence intervals for sensitivity and specificity are reasonably narrow.

I talk about this at several places on my old and new websites. Here's a
couple of good starting points:

http://www.pmean.com/category/DiagnosticTesting.html
http://www.pmean.com/category/SampleSizeJustification.html
--
Steve Simon, Standard Disclaimer
The Monthly Mean is celebrating its first anniversary.
Find out more about the newsletter that dares
to call itself "average" at www.pmean.com/news

[Pedro Emmanuel Alvarenga Americano do Brasil]

If you would like something very simple and straight I suggest this reading!

http://emj.bmj.com/cgi/content/full/22/3/180

It implments a graph wich no formulars or computers is necessary. Very good for starters.

Abraço forte e que a força esteja com você,

Dr. Pedro Emmanuel A. A. do Brasil
Instituto de Pesquisa Clínica Evandro Chagas
Fundação Oswaldo Cruz
Rio de Janeiro - Brasil

2009/11/16 Steve Simon, P.Mean Consulting <n...@pmean.com>

[Thomas Fröjd]

Just for your information, if I remember it correctly there is a typo
in the graph suggested below in the pdf version. The correct graph can
be had at the BMJ website.

On 16 Nov, 19:27, Pedro Emmanuel Alvarenga Americano do Brasil

[Evie]

That's great, thank you all for your help.

Evie

> > > to call itself "average" atwww.pmean.com/news- Hide quoted text -
>
> - Show quoted text -

[Barry Brown]

> > Evie wrote:
> > > I have a clinical test (positive/negative) and surgery (disease/not-
> > > diseased) and plan to calculate sensitivity/specificity. How do I
> > > estimate the sample size? Any advice would be greatly appreciated.
>
> > The traditional sample size formulas based on power don't work here. The
> > simplest thing is to select a sample size that insures that the
> > confidence intervals for sensitivity and specificity are reasonably narrow.
>

I have written a program called CONFINT that calculates the sample size necessary to
have a specified probability that the confidence interval length  is not longer than some
specified size.   The program is free and includes Fprtran90 source, Win32 and Mac
executables.  Download from
          http://biostatistics.mdanderson.org/SoftwareDownload

--
Barry W Brown
Professor, Biostatistics and Applied Mathematics
M. D. Anderson Cancer Center

[Martin Holt]

If I understand correctly, Evie will need to be able to have confidence intervals simultaneously for sensitivity and specificity. Googling, I found this (it comes with no guarantees). Beware the wrap-over.

 

http://74.125.47.132/search?q=cache:a0Idf5bGvCkJ:www.kck.usm.my/ppsg/statistical_resources/samplesize_forsensitivity_specificitystudiesLinNaing.xls+sensitivity+specificity+sample+size+formula&cd=6&hl=en&ct=clnk

 

In the past I have used a "simple" formula found in a fairly unknown journal but it made sense..it could be fairly easily derived...but it's not to hand. If you'd like me to search for it just let me know.

 

Best Wishes,

Martin

[Evie]

Thank you all for your advice which has been really helpful

On 17 Nov, 17:52, "Martin Holt" <m861h...@btinternet.com> wrote:
> If I understand correctly, Evie will need to be able to have confidence intervals simultaneously for sensitivity and specificity. Googling, I found this (it comes with no guarantees). Beware the wrap-over.
>

> http://74.125.47.132/search?q=cache:a0Idf5bGvCkJ:www.kck.usm.my/ppsg/...

>
> In the past I have used a "simple" formula found in a fairly unknown journal but it made sense..it could be fairly easily derived...but it's not to hand. If you'd like me to search for it just let me know.
>
> Best Wishes,
> Martin
>
>
>
>   ----- Original Message -----
>   From: Barry Brown
>   To: meds...@googlegroups.com
>   Sent: Tuesday, November 17, 2009 4:43 PM
>   Subject: {MEDSTATS} Re: sample size for sensitivity/specificity
>
>     > > Evie wrote:
>     > > > I have a clinical test (positive/negative) and surgery (disease/not-
>     > > > diseased) and plan to calculate sensitivity/specificity. How do I
>     > > > estimate the sample size? Any advice would be greatly appreciated.
>
>     > > The traditional sample size formulas based on power don't work here. The
>     > > simplest thing is to select a sample size that insures that the
>     > > confidence intervals for sensitivity and specificity are reasonably narrow.
>
>   I have written a program called CONFINT that calculates the sample size necessary to
>   have a specified probability that the confidence interval length  is not longer than some
>   specified size.   The program is free and includes Fprtran90 source, Win32 and Mac
>   executables.  Download from
>            http://biostatistics.mdanderson.org/SoftwareDownload
>
>   --
>   Barry W Brown
>   Professor, Biostatistics and Applied Mathematics

>   M. D. Anderson Cancer Center- Hide quoted text -