I need to calculate the sample size for a medical study: the aim is to
determine the cutoff point for the number of cells found in the esophagus of
‘normal’ patients and in patients with a certain condition.
There is no reference cell count available for “normal patients”. What is
known is that patients with this condition usually present a much higher
cell count. The study aims to determine a cutoff point in the cell count.
The researchers intend to work with 3 groups:
1) normal patients,
2) patients without symptoms and with the condition,
3) patients with symptoms and with the condition.
How many patients should I use in each group?
It looks like a Poisson distribution (I’d have to test), but I could work
with ANOVA if I the distribution were approx. normal, or otherwise Kruskal
Wallis.
ANY help would be greatly appreciated.
Thanks,
Binha
--
View this message in context: http://www.nabble.com/Calculating-sample-size-to-determine-a-cell-count-cutoff-point-tp16452219p16452219.html
Sent from the AllStat mailing list archive at Nabble.com.
|