+ Reply to Thread
Results 1 to 3 of 3

Thread: Confounder analysis for LOGISTIC regression

  1. #1
    Points: 5, Level: 1
    Level completed: 9%, Points required for next Level: 45

    Posts
    1
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Confounder analysis for LOGISTIC regression




    Hi there,
    I have the following situation:
    I have a logistic regression analysis, in which I have 3 variables:
    The dependent variable- dichotomous- (complication/no complication).
    The independent variable- continuous variable- no. of days passed since hospitalization.
    The confounding variable- also continuous variable- CCI Index.
    How do I perform such an analysis in SPSS? How do I interpret the results? Is there anything to read or a YouTube video explaining that?
    Many thanks and best regards!

  2. #2
    Omega Contributor
    Points: 38,396, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    7,001
    Thanks
    398
    Thanked 1,186 Times in 1,147 Posts

    Re: Confounder analysis for LOGISTIC regression

    What do you think is the confounder?

    Is complication post-discharge?

    What is your sample size and how many complications?

    What is the range for CCI index and what exactly is the instrument use?

    CCI and days may be correlate.

    -sorry i dont use SPSS.
    Stop cowardice, ban guns!

  3. #3
    Points: 3,730, Level: 38
    Level completed: 54%, Points required for next Level: 70

    Posts
    155
    Thanks
    7
    Thanked 30 Times in 29 Posts

    Re: Confounder analysis for LOGISTIC regression


    I would also add, there are some general guidelines for confounding, some sources have more extensive definitions but the following two are generally consistent:
    1. The confounder should be associated with the exposure (predictor) but not a consequence of the exposure
    2. The confounder should be associated with the disease (outcome) but not a consequence of the disease

    The two types of conditions stated above can be applied based on the data you have, but can also take into consideration the body of evidence that exists (i.e. the literature on your specific topic). For assessing confounding based on your data, researchers typically calculate how much the adjusted and unadjusted odds ratio differ, usually using some arbitrary value such as 10% or 20% (i.e. the adjusted OR + 10% and - 10%). Sometimes this is frowned upon because this procedure is specific to your dataset, which is why the definition considering the body of evidence is also common (and is probably what informed the decision to collect data on the potential confounder anyway). In many cases both definitions should inform your decision on whether or not to adjust.

    In terms of interpreting the effect, compare the adjusted and unadjusted odds ratios, if they are meaningfully different this is generally taken to provide evidence of confounding.

+ Reply to Thread

           




Tags for this 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