Forest plot for a single clinical trial study subgroup analysis

[Winfred Mwangi]

Hi I would like to generate forest plot as described above:

Am using this command:

ipdover, over( Age_protocol2 FacilityxVL2 timeDxHIV_cat2 Marital_status3 education_level3 hours_cat2 phone_own3 PregnancyOutcome ) ///
irr forestplot( (null xlabel( 0.10 (1) 6, force) boxsca(70) textsize(150) (pci)) ///
favours("<---------------------" "Favours SMS group" # "--------------------->" "Favours Control group" ) fp(6)) ///
: poisson gap33_binary arm2 if gap33b==1

I get attached graph

1. How do I generate p values and include a column at the end for p values?
2. How do I interpret the findings eg with the marital status -single status seem to favour intervention?
3. What the direction of my arrows correct?

Thank you.

NB/ I used poisson here but I know and it has been suggested I should use poisson. But please help me get the command or the procedure right.

See data example below:

----------------------- copy starting from the next line -----------------------

* Example generated by -dataex-. To install: ssc install dataex
clear
input str14 clinic_id float(gap33_binary arm2 Age_protocol2 FacilityxVL2 timeDxHIV_cat2 Marital_status3 education_level3 hours_cat2 phone_own3 PregnancyOutcome)
"K2123" 0 0 2 3 2 1 2 1 1 0
"K2140" 1 0 2 3 3 1 2 0 1 0
"K2200" 0 0 1 3 1 1 2 1 1 0
"K2183" 0 1 1 3 1 1 3 1 1 0
"K2185" 1 0 1 3 3 3 4 1 1 0
end
label values arm2 arm2_label
label def arm2_label 0 "Control", modify
label def arm2_label 1 "sms", modify
label values Age_protocol2 Age_protocol2_label
label def Age_protocol2_label 1 "18-29yrs", modify
label def Age_protocol2_label 2 "30-39yrs", modify
label values FacilityxVL2 FacilityxVL2_label
label def FacilityxVL2_label 3 "Kitale", modify
label values timeDxHIV_cat2 timeDxHIV_cat2_label
label def timeDxHIV_cat2_label 1 "≤1yr", modify
label def timeDxHIV_cat2_label 2 "1-5yrs", modify
label def timeDxHIV_cat2_label 3 ">5yr", modify
label values Marital_status3 Marital_status3_label
label def Marital_status3_label 1 "Married", modify
label def Marital_status3_label 3 "widow/separated", modify
label values education_level3 education_level3_label
label def education_level3_label 2 "Primary", modify
label def education_level3_label 3 "Secondary", modify
label def education_level3_label 4 "Post-secondary", modify
label values hours_cat2 hours_cat2_label
label def hours_cat2_label 0 ">1hr", modify
label def hours_cat2_label 1 "≤1hr", modify
label values phone_own3 phone_own3_label
label def phone_own3_label 1 "Yes", modify
label values PregnancyOutcome PregnancyOutcome_label
label def PregnancyOutcome_label 0 "Live birth", modify

------------------ copy up to and including the previous line ------------------

Thank you again!

[Winfred Mwangi]

Apologies for my typos
No 3 querry is Is the direction of the of arrows of attached graph correct?

It has been suggested I should use logistic regression

Thanks