Group wise estimated absolute risk / effects after meta analysis for GRADE

[Vivek Gupta]

Good afternoon All
I am just tipping my topples with meta analysis in Stata 17. It was working smooth till I reached the step of developing a GRADE table.
GRADE tables require not just the pooled Risk Ratio /. Mean difference etc, but also the following

1. Number of participants in each comparison arm and number of outcomes. This bit is easily achieved using "summarise" and r(Summer) scalars.
2. RISK DIFFERENCE - Estimated by re-runninng meta with ,rd
3. Absolute Risk in Control or intervention Arm - This is where I am struggling

Any help would be much appreciated

My current approach is below.
My doubt is that the risk so obtained is unweighted.
Is this the right approach? If no, how can I obtain weighed risk
Many Thanks
Vivek

input study   intout   intnoout   contout   contnoout  
"Study 1"       22        873        27        885  
"Study 2"       88       2393        79       2397 
end
meta esize  intout intnoout  contout contnoout, studylabel(study) esize(lnrratio)
meta summarize , rr fixed

// FE meta-analysis with Risk Difference
meta esize  intout intnoout  contout contnoout, studylabel(study) esize(rd)
meta summarize ,  fixed
// Simple summations and divisions for n/N in Intervention and Control and Absolute risk in controls
quietly summ intout
local sintout = r(sum)
quietly  summ intnoout
local sintnoout = r(sum)
quietly  summ contout
local scontout = r(sum)
quietly  summ contnoout
local scontnoout = r(sum)
display "No. of patients" col(30) "Risk-Controls" _newline ///
    _col(1) "Control" _col(12) "Inter." _col(30) "(%)"_newline ///
    _col(1) `scontout' "/" `scontnoout'+`scontout' /// n/N in Control
    _col(12) `sintout' "/" `sintout'+`sintnoout' /// n/N in Intervention
    _col(30) %5.2f  (`scontout'*100) / (`scontnoout'+`scontout') // Risk