Hi All, Sorry if this is a terribly simple question but I have looked north and south and can't seem to find an answer. I have a very simple model with one continuous predictor and an outcome variable, and want to get a relative risk to say that for each unit increase in x the relative risk for y increases by z amount. Is there a way to get the point estimate and 95% confidence interval for a continuous variable equivalent to the 'estimate 'Beta' trt_group 1 -1 / exp' statement for categorical predictors? PROC GENMOD DATA=Have MODEL y = x/dist=bin link=log; RUN; Many thanks
... View more
Many thanks Koen, that information was interesting. Unfortunately, I still couldn't figure out a way to calculate the mean across the whole study. Is the only way to calculate the lsmeans at specific time points, save to an output, and take a mean of those means? Or is there a simpler way?
... View more
I am looking to do an exposure response analysis. The drug exposure readings are collected at weeks 0, 5, 10, 15, 20, 25, 30. I want to relate this to binary outcomes (eg cardiovascular event yes_no). What would be the best way to analyse this - proc glimmix? Any help greatly appreciated.
... View more
Hello all, I need some help with marginal effects in proc mixed. I am comparing two treatments from a clinical trial. I have tried the following code: PROC MIXED DATA=Data METHOD=ml; CLASS trt_group patient_id; MODEL y=visit_yrs|trt_group/solution residual; Random int visit_yrs/subject=patient_id; FORMAT trt_group trt_group.; RUN; What I'd like to do is get the marginal effects for the treatment group (ie mean of all readings for all participants across the study). I believe lsmeans only allows categorical variables. I did try putting visit_wks as a class variable to accommodate the lsmeans statement, however that seems to change all the parameters for the model. I hope I've got the terminology correct. Any help would be greatly appreciated!
... View more
Thanks Modeller, that makes sense. I guess what I'd like to be able to say is that conclusions from testing the N=150 population can be generalised to the N=200 population (because unfortunately not everyone in the N=200 had the assay). Is that still possible, even if the N=50 is significantly different from the N=150 in some ways?
... View more
I have a dataset (say N=200) of which a subset (N=150) have undergone a certain assay. I want to be able to say that the subgroup N=150 is not significantly different from the entire population (N=200). Is there a way to compare the two eg in terms of baseline characteristics. I tried comparing the N= 150 who have the assay with N=50 who do not have the assay (eg with ttests and chi square tests). However, for one variable (blood pressure), the difference between the N=150 and N=50 group was significant, however on visual inspection the value for the N=150 did not look very different from the N=200. Is there a way to compare the N=150 subgroup to the N=200 dataset rather than compare the N=150 and N=50 subgroups of the N=200. I hope that makes sense.
... View more