+ Reply to Thread
Results 1 to 3 of 3

Thread: Who to analyse these data please

  1. #1
    Points: 883, Level: 15
    Level completed: 83%, Points required for next Level: 17

    Posts
    6
    Thanks
    3
    Thanked 0 Times in 0 Posts

    Unhappy Who to analyse these data please




    A very short question:
    I have a group of patients, followed after an operation, and I have collected a number of variables on these patients before surgery
    Independent variables: Age, Sex.... (ie both continuous and categorical)
    Intervention: Surgery
    The outcome: Time until discharge (in days)
    I am not sure how to analys these data:
    1- Survival analysis with Cox PH: The problem is that all patients will be discharged and hence the event of interest (discharge) will occur in all patients, ie no censoring at all.
    2- Logistic regression after collapsing the outcome variabel into two categories, but this approach has lower power that if I would analys the outcome variable as a continuous variable, and it is unclear how to dichotomise the outcome variable (which value to split at)
    3- Negative binomial regression (not sure) since the outcome variable is in whole days (1, 2, 4.....)
    4- Multiple linear regression(?) the problem is that I have some categorical independent variables, I have read that categorical variables can be included in multiple linear regression, but not sure if this is correct

    Any help would be very appreciated

    Kind regards

  2. #2
    TS Contributor
    Points: 17,749, Level: 84
    Level completed: 80%, Points required for next Level: 101
    Karabiner's Avatar
    Location
    FC Schalke 04, Germany
    Posts
    2,540
    Thanks
    56
    Thanked 640 Times in 602 Posts

    Re: Who to analyse these data please

    What makes you think that for a survival analysis you need censored cases?
    Censored cases are a nuisance (which can be dealt with), but not a prerequisite.

    But with no censored cases, you could indeed try linear regression. Categorical
    variables can be included into the model. You have to transform categorical variables
    with k categories into k-1 dummy-variables first.

    With kind regards

    K.

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

    Bio_Stat (08-27-2014)

  4. #3
    Points: 883, Level: 15
    Level completed: 83%, Points required for next Level: 17

    Posts
    6
    Thanks
    3
    Thanked 0 Times in 0 Posts

    Re: Who to analyse these data please


    Thank you Karabiner
    I wonder in that case if survival analysis has any advantages compared to linear regression?
    Regards
    AM

+ Reply to Thread

           




Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts






Advertise on Talk Stats