2 weeks ago
running a logistic regression model in SAS I found 6 predictive variables (risk factors) of a disease.
I would like to know which SAS procedure will allow me to:
run an AIC stepwise procedure so I can construct an optimal predictive model, for instance this predictive model will allow me to divide patients into low (0-2 predictive variables or risk factors), moderate (3-4) or high risk GROUPS (5-6) to get the corresponding cumulative incidence of the disease expressed as % (and 95% confidence interval). For example having 0 to 2 risk factor will result in 3.0% (95% CI 1.1-6.5%), 3 to 4 in 41.6% (31.0–51.8%), and 5-6 risk factors in 77.3% (51.4–90.5%), respectively for low, moderate and high risk groups.
all your help will be greatly appreciate it