1. ## outcomes of surgery

Can someone point me into the right direction

I have a database of patients who have had surgery which is deemed a success or not(0,1).

Each patient has a number of variables under interest
before or after a date (1,2)
one of 4 types of procedure ( 1,2,3,4)

all of which may or may not impact on the success of surgery.

ontop of this I am looking at how long each operation took (continous data) - and whether or not there is a corelation between the time it took and the variables above.

I think I am looking at doing a linear regression but not sure

please can anyone point me in the right direction

Thanks

ed

2. ## Re: outcomes of surgery

If your dependent variable is 1 or 0, then you are likely looking at multiple logistic regression. What is your sample size and what percent had a successful surgery?

3. ## The Following User Says Thank You to hlsmith For This Useful Post:

edmundfitz (02-24-2015)

4. ## Re: outcomes of surgery

Originally Posted by hlsmith
If your dependent variable is 1 or 0, then you are likely looking at multiple logistic regression. What is your sample size and what percent had a successful surgery?
Hi -

there are 2 separate outcomes - success or failure ( 0,1)

and time for operation ( continous data ) 30 to 560 mins

the sample size is 1210
success is around 90% failure is 10%

i thought that you cant use a multiple logistic regression with a continous data outcome ?

Thanks

5. ## Re: outcomes of surgery

So you have got 2 analyses, one multiple logistic regression (success),
and one multiple linear regression (time) (or, ANOVA with 3 factors).
I would assume that perhaps a logarithmic transformation of "time"
could improve model fit.

With kind regards

K.

6. ## The Following User Says Thank You to Karabiner For This Useful Post:

edmundfitz (02-27-2015)

 Tweet