Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • metan Hazard Ratio log scale

    Hello,
    I am running cox regression and I want to make forest plot of my hazard ratios using metan. I have been asked to produce my forest plot in log scale. If I remove "eform" from my command, the table of HRs with confidence intervals are also in log scale which makes it not readable. Here is a part of my data and the syntax I have used so far.

    clear
    input float RR float lci float uci str30 Condition str30 Period index

    1.24 1.07 1.43 "condition 1" "1-3 " 1
    1.21 1.02 1.44 "condition 1" "4-6 " 1
    0.92 0.71 1.20 "condition 1" "7-9 " 1
    1.33 0.92 1.93 "condition 1" "10-12" 1

    1.57 1.43 1.72 "condition 2" "1-3" 2
    1.21 1.07 1.38 "condition 2" "4-6" 2
    1.04 0.86 1.25 "condition 2" "7-9" 2
    1.06 0.77 1.48 "condition 2" "10-12" 2
    end

    gen double lRR=log(RR)
    gen double lRRlci=log(lci)
    gen double lRRuci=log(uci)


    metan lRR lRRlci lRRuci , by(Condition) eform effect(HR) xlabel(0.5,2) force nooverall stats nosubgroup nobox nowt label(namevar=Period) xtitle("Hazard ratio (95% CI)", size(small)) graphregion(fcolor(white))


    I thought this was the way to produce on log scale. So what do I do wrong? I really appreciate any suggestions
Working...
X