01-26-2018 12:52 PM
I’d like to assess the bias due to human mobility on the incidence rate of a rare disease, let’s say schizophrenia (SP). Incidence rate formula below shows that number of people who moved in (inflow) and moved out (outflow) of the study area units (county) would affect the incidence rate (for a single year in this example).
Incidence rate=new cases /number of people who don't have SP yet
Number of new cases, inflow and outflow is known. However, the risk of SP among movers is not known but its probability must range from 0 through 1. I’d like to run this simulation exercise in SAS.
I’d like to compute all possible value of number of people with SP among people moved out and moved out given risk of SP ranges 0-1 and also compute all possible values for Net=n of SP among moved in-n of SP among moved out. The result can be plotted as [SP risk among movers*bias in incidence rate] scatter plot. Just to clarify, if risk=1 for SP among movers then all who moved out and moved in would have SP.
Can you please help what SAS procedure or statistics help run this simulation study? Any specific example or demo will be highly appreciated. I have never done it before.
Thanks for your precious time!
See attached is the sample data.
proc import datafile="...\support.csv" out=data dbms=csv replace; guessingrows=max; run;
01-26-2018 12:56 PM
Do you have a full license for SAS? If so, that likely includes SAS Simulation Studio, though it may not be installed.
You should verify that.
If you're going to program it in BASE, first read the paper by David Cassell - Don't Be Loopy, Simulation the SAS Way
And another good reference is Rick Wicklins book on Simulation.
Need further help from the community? Please ask a new question.