How to create a dataset for a Poisson Model to control confunders of age gender?

[HelenLi]

Greetings,

I want to assess the association between EmergencyVisitDueToInfection vs. Temperature. My question is how I can control age and gender in the model, please?

Table 1 is Emergency Visit per row per patient.

Table 2 is Daily temperature (per row per date).

I created table 3 which contains: my dependent variable Infection-Count per date, my independent variable temperature per date. I also want to include age-related or gender-related variable in table 3 before I feed the PROC GENMOD Poisson Model. However, I am not sure what kind format of gender/age I should create in table 3, so I can control age/gender confounders in PROC GENMOD.

Table 1:
ID  Date           Age  Gender Infection (1=Yes, 0=No) 
1   2018-01-01   20     M           1  
2   2018-01-01   30     F            0 
3   2018-01-02   40     F            1 
4   2018-01-02   50     F            1  

5   2018-01-02   60     M           1 

Table 2:

Date              Temperature

2018-01-01         5

2018-01-02       10
......

Table 3:
Date             Count of Infection  Temperature
2018-01-01            1                        5
2018-01-02            2                        10
......

PROC GENMOD DATA = Table3;

            MODEL CountOfInfection = Temperature  Gender? Age? /Link= Log Disk=P;

RUN;

How can I put age and gender in the table 3, please?

Thanks a lot,

Helen

Many thanks,
Helen

[StatDave]

Your response is binary at the individual level (table 1). Simply add the temperature to the individual data and fit a logistic model with temperature, gender, and age as predictors.