https://communities.sas.com/t5/Statistical-Procedures/Assessing-Modifying-Effects-of-City-Parameters-on-Heat-EMS/m-p/927137#M46109 <P>Hello SAS Community,</P><P>I'm currently working on a project where I'm investigating the modifying effects of city parameters on the relationship between heat and the number of EMS services across 25 cities. Here's a breakdown of my approach and where I need some guidance:</P><OL><LI><P><STRONG>Initial Stage</STRONG>:</P><UL><LI>I've estimated the heat effects adjusted for the year and weekday using a negative binomial model for each city separately. This step provided city-specific estimates of the heat effects.</LI></UL></LI><LI><P><STRONG>Second Stage</STRONG>:</P><UL><LI>Now, I want to use meta-regression to explain the variance of the heat effects across the cities.</LI><LI>My challenge lies in incorporating two different types of covariates: daily parameters and yearly parameters.</LI></UL></LI></OL><P>Here are my specific questions:</P><UL><LI>Is there a way to assess the daily covariates in the meta-regression without aggregating them into a single mean?</LI><LI>How can I properly incorporate both daily and yearly parameters into the meta-regression model while accounting for the city-specific estimates obtained in the initial stage?</LI><LI>How should the data structure look like to utilize the <CODE>PROC MIXED</CODE> function for this purpose?</LI></UL><P>Any advice or suggestions on how to approach this would be greatly appreciated.</P><P>Thank you!</P> Mon, 06 May 2024 09:12:29 GMT mortz 2024-05-06T09:12:29Z https://communities.sas.com/t5/Statistical-Procedures/Assessing-Modifying-Effects-of-City-Parameters-on-Heat-EMS/m-p/927137#M46109 <P>Hello SAS Community,</P><P>I'm currently working on a project where I'm investigating the modifying effects of city parameters on the relationship between heat and the number of EMS services across 25 cities. Here's a breakdown of my approach and where I need some guidance:</P><OL><LI><P><STRONG>Initial Stage</STRONG>:</P><UL><LI>I've estimated the heat effects adjusted for the year and weekday using a negative binomial model for each city separately. This step provided city-specific estimates of the heat effects.</LI></UL></LI><LI><P><STRONG>Second Stage</STRONG>:</P><UL><LI>Now, I want to use meta-regression to explain the variance of the heat effects across the cities.</LI><LI>My challenge lies in incorporating two different types of covariates: daily parameters and yearly parameters.</LI></UL></LI></OL><P>Here are my specific questions:</P><UL><LI>Is there a way to assess the daily covariates in the meta-regression without aggregating them into a single mean?</LI><LI>How can I properly incorporate both daily and yearly parameters into the meta-regression model while accounting for the city-specific estimates obtained in the initial stage?</LI><LI>How should the data structure look like to utilize the <CODE>PROC MIXED</CODE> function for this purpose?</LI></UL><P>Any advice or suggestions on how to approach this would be greatly appreciated.</P><P>Thank you!</P> Mon, 06 May 2024 09:12:29 GMT https://communities.sas.com/t5/Statistical-Procedures/Assessing-Modifying-Effects-of-City-Parameters-on-Heat-EMS/m-p/927137#M46109 mortz 2024-05-06T09:12:29Z