What test to use?

[Rgemmell]

Hopefully a straightforward question!

I have a group of patients, and just under half were known to our service before their admission to hospital. I've grouped them according to disease type, and I want to show whether there's a significant difference, so for each disease type a p value. What should I be using?

 

[Karabiner]

Hopefully a straightforward question!

We'll see.

I have a group of patients,

How large ist this group?

and just under half were known to our service before their admission to hospital. I've grouped them according to disease type,

How many disease types are there? Does each patient have 1 only 1 disease type assigned to him, or possibly more than one?

and I want to show whether there's a significant difference,

Difference between what, exactely?

so for each disease type a p value.

What is your research question?

With kind regards

Karabiner

 

[Rgemmell]

We'll see.

How large ist this group?

How many disease types are there? Does each patient have 1 only 1 disease type assigned to him, or possibly more than one?

Difference between what, exactely?

What is your research question?

With kind regards

Karabiner

Thanks for your response!

I have 159 patients, divided into twelve disease types, with each patient only having one disease. My research question is related to improving referrals for all disease types, this is to look at baseline information, i.e. are all of the disease types as likely to have referred to our speciality before admission?

 

[Karabiner]

That means, you have 159 patients, divided into 24 groups (2 groups based on previous admission x 12 disease types)?
Is it perhaps possible to increase your sample size?

With kind regards

Karabiner

 

[Rgemmell]

Unfortunately not, this was an extra analysis I was asked to include in describing my demographics.