The pedant in me wants to note that this result is based on the number
of events being Poisson distributed.
Good thought Ronan
Tony
Peter A. Lachenbruch
Department of Public Health
Oregon State University
Corvallis, OR 97330
Phone: 541-737-3832
FAX: 541-737-4001
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Ronan Conroy
Sent: Friday, August 15, 2008 8:38 AM
To: [email protected]
Subject: Re: st: RE: Detection of disease
On 14 Aug 2008, at 16:57, On Behalf Of Carlo Georges wrote:
> For example i need to detect with 95% confidence the abscence of
> disease
> in
> a population where the presumed prevalence would be 20%. How lrge a
> sample
> size do I need to be 95% certain that the population is free from
> disease.
This is an impossible task, I think. A better approach would be to ask
what the maximum disease prevalence would be to result in zero
observed cases in a sample size N.
There was a lovely paper years ago in JAMA called
Hanley, J. A., & Lippman-Hand, A. (1983). If nothing goes wrong, is
everything all right? Interpreting zero numerators. JAMA, 249(13),
1743-1745.
Hanley and Lippman-Hand make the point that if zero events are
observed in N cases, then the upper limit is roughly 3/N. This means
that even if you observe no cases in 1,000 participants, the 95% CI
for the rate is zero to 3.7 per thousand (I cheated and did a -cii- on
this). So you can be 95% certain that the rate is no more than 3.7 per
thousand or less.
The topic is discussed in
Eypasch E, Lefering R, Kum CK, Troidl H. Probability of adverse events
that have not yet occurred: a statistical reminder. BMJ. 1995 Sep
2;311(7005):619-20.
which is accessible online.
http://www.bmj.com/cgi/content/full/311/7005/619
Ronan Conroy
=================================
[email protected]
Royal College of Surgeons in Ireland
Epidemiology Department,
Beaux Lane House, Dublin 2, Ireland
+353 (0)1 402 2431
+353 (0)87 799 97 95
+353 (0)1 402 2764 (Fax - remember them?)
http://www.flickr.com/photos/ronanconroy/sets/72157601895416740/
P Before printing, think about the environment
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
