| |
 |
|
|
Science Forum Index » Space - Consult Forum » per 100 person years formula
Page 1 of 1
|
| Author |
Message |
| Seema |
 Posted: Thu Jan 18, 2007 7:23 am |
|
|
|
Guest
|
Dear all
Please help me on the formula for per 100 person years.
How to calculate the same
For example:
Frequency of disease progression - 42.0 (95% CI 26.6-61.5) per 100
person years.
Thanks
Seema |
|
|
| Back to top |
|
| David Winsemius |
| Posted: Fri Jan 19, 2007 9:04 pm |
|
|
|
Guest
|
"Seema" <shah.seema@ranbaxy.com> wrote in news:1169119413.924104.28940
@m58g2000cwm.googlegroups.com:
Quote: Dear all
Please help me on the formula for per 100 person years.
How to calculate the same
For example:
Frequency of disease progression - 42.0 (95% CI 26.6-61.5) per 100
person years.
Perhaps a couple of examples could help. That CI suggests a std deviation
around the point estimate of (42-26.6)/1.96 = 7.86 and the variance of =
61.7 would suggest that you had someplace around 62 events. Person-years
is sum of the subject time people in the study were at risk of an event,
possibly more than or less than one year for any given person. If the
event is non-recurring as progression would be, then you need to stop
counting at the time of the event. If a subject is observed without an
event for the entire trial period, then all of that time is included.
Lets say it was 63 events, since the math works out better. If your
observed rate was 42/100 Pyrs, that could have come about with any of the
following:
300 persons being observed an average of 6 months with 63 events, or ...
150 persons being observed an average of one year with 63 events, or ...
75 persons being observed an average of 2 years with 63 events.
--
David Winsemius |
|
|
| Back to top |
|
| |
|
Page 1 of 1
All times are GMT - 5 Hours
The time now is Wed Dec 03, 2008 4:12 pm
|
|