How to interpret a HR of 0.79 (0.63-1.00) p=0.026?

jj11

New Member
#1
Hi!
A RCT study reported in the lancet reported the following hazard rate: 0.79 (95%CI 0.63-1.00) with a p=0.026.
I have always ''learned'' to interpret a 1 in a HR as non-significant, however the p-value is below 0.05.
Can I conclude a difference between the two study arms?

thanks!
 

jj11

New Member
#3
The article does not state that. The only information that was given was the following:
"the Hochberg approach was used to adjust for multiple statistical testing of overall survival across the two populations to maintain an overall type I error rate of 5%."
If it also is applicable to this analysis is not clearcly stated unfortunately.
 

hlsmith

Not a robit
#4
Post a link to the article and reference exactly which statistics you are examining.

Hochberg is typically a stepwise correction, so it is likely difficult to tease out the order of pvalues corrected. It may be likely that the author only corrected the pvalue and not the alpha used on the HR (or possible the other way around). Though seeing the article may help to resolve this.

Yes, if the HR CI includes one it means that you can't rule out the hazards between the groups may not be the same.
 

hlsmith

Not a robit
#6
I only skimmed the abstract but this is your sentence of interest:

" a significant difference was defined as p<0·025"

Think about this and let us know if you have questions.

You messed up above when writing "95%" they report: "HR 0·79 [97·5% CI 0·63–1·00]; p=0·026)"