Multilevel Models for dependent variables

mydamj — Tue, 28 Feb 2023 16:58:07 GMT

It’s about Medical Project.

Our dependent variable which is preterm has 4 values/ groups (extreme preterm(Baby born <32 weeks of gestation ) Early preterm ( bon 32- 34 weeks of ga) late preterm( 35-36) and term (>=37 weeks of gestation)).

We are experienced in SAS in dealing with a binary (2 groups Preterm-YES & Preterm-NO) especially in our Proc-Genmod procedure to get our relative risks. Now we are stuck when we are trying subdivided preterm group/cohort( Extreme(<32) Early(32-34)Late(35-36)).

When we subdivided as above the SAS code/ Proc-Genmod code we already used for our binary outcome variable won’t work(which we know).

To resolve this troubleshoot we took an alternate/ shortcut path where we kept our outcome variable binary with meeting above subgroups by doing multiple models instead of single model (i.e., extreme Preterm vs Term- Model1; Early Preterm Vs Term - Model2; Late Preterm vs Term).

Our goal is to create a single model where our dependent variable has 4 values explained above my reference is term and I would like to get my RR(Relative Risk) for my 3 groups(extreme preterm, early preterm, Late preterm)

Need Suggestions,

Thanks

Re: Multilevel Models for dependent variables

StatDave — Tue, 28 Feb 2023 17:38:56 GMT

The note that I referred you to in your previous post shows how this can be done.

topic Re: Multilevel Models for dependent variables in Statistical Procedures

Multilevel Models for dependent variables

Re: Multilevel Models for dependent variables