Meta-analysis of dose-response

[Paula Iso-Markku]

Hello,

I am trying to create a dose-response graph for a meta-analysis from risk ratio data using drmeta command from SSC in Stata 18.0. The commands I am trying the use are the following:

. mkspline METs= MET, nk(3) cubic

. mat knots = r(knots)

. drmeta b METs1 METs2, se(SE) data(N Cases) id(id) type(type) ml
variable type can take value 1 (odds ratio), 2 (hazard ratio), 3 (risk ratio), 4 (mean difference), or 5 (std mean difference)
                       *:  3200  conformability error
                glsest():     -  function returned error
                 <istmt>:     -  function returned error
r(3200);

Here is an example of my data:

* Example generated by -dataex-. For more info, type help dataex
clear
input double(b MET SE) int(N Cases) byte(id type)
                  0  120                   0  564 160 1 3
-.00502516792675071  420 .054867056211547936  853 146 1 3
 -.1178611667605349  840  .06224312905563363 2837 407 1 3
                  0    0                   0  292  31 2 3
 -.5194638117686279  540  .24544126039932823  928  48 2 3
  -.625001381111016 1260    .251750701640219 1249  50 2 3
                  0    0                   0  448  55 3 3
-.16251892949777494  630  .21943909827120192  591  66 3 3
-.35667494393873245 1080    .269671570857489  415  29 3 3
 -.3973058342603201  210   .3629646995849802  182  13 3 3
end

I don't understand what is wrong with my variable "type".