Stratified odds ratio following MICE procedure

[Sejal Patel]

Hi,
I am using Stata 14.
I imputed data using MICE procedure. Then I would like to estimate stratified odds ratio for smoking using logistic model (below).

. mi estimate: logit unconbp age ldl i.dm##i.smoking i.bmic i.race_eth

I tried to use lincom statment for odds ratio for smoking categories but got error msg.

requested action not valid after most recent estimation command
r(321);

Please let me know the code to get the stratified odds ratio for smoking categories,
Thank you!

[Clyde Schechter]

Post-estimation after multiple imputation is severely limited. The -lincom- command is not available. You can, however, emulate it from available commands. Here is an example, adapted from the -help file- for -mi estimate postestimation- and -help mi test-:

. clear*

. webuse mhouses1993s30
(Albuquerque home prices Feb15–Apr30, 1993)

.
. mi estimate (diff: _b[tax]-_b[sqft]): ///
>   regress price tax sqft age nfeatures ne custom corner

Multiple-imputation estimates                   Imputations       =         30
Linear regression                               Number of obs     =        117
                                                Average RVI       =     0.0648
                                                Largest FMI       =     0.2533
                                                Complete DF       =        109
DF adjustment:   Small sample                   DF:     min       =      69.12
                                                        avg       =      94.02
                                                        max       =     105.51
Model F test:       Equal FMI                   F(   7,  106.5)   =      67.18
Within VCE type:          OLS                   Prob > F          =     0.0000

------------------------------------------------------------------------------
       price | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         tax |   .6768015   .1241568     5.45   0.000     .4301777    .9234253
        sqft |   .2118129    .069177     3.06   0.003     .0745091    .3491168
         age |   .2471445   1.653669     0.15   0.882    -3.051732    3.546021
   nfeatures |   9.288033   13.30469     0.70   0.487    -17.12017    35.69623
          ne |   2.518996   36.99365     0.07   0.946    -70.90416    75.94215
      custom |   134.2193   43.29755     3.10   0.002     48.35674    220.0818
      corner |  -68.58686    39.9488    -1.72   0.089    -147.7934    10.61972
       _cons |   123.9118   71.05816     1.74   0.085    -17.19932    265.0229
------------------------------------------------------------------------------

Transformations                                 Average RVI       =     0.1200
                                                Largest FMI       =     0.1100
                                                Complete DF       =        109
DF adjustment:   Small sample                   DF:     min       =      92.10
                                                        avg       =      92.10
Within VCE type:          OLS                           max       =      92.10

         diff: _b[tax]-_b[sqft]

------------------------------------------------------------------------------
       price | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        diff |   .4649885   .1863919     2.49   0.014     .0948037    .8351733
------------------------------------------------------------------------------

[Sejal Patel]

Thank you Clyde.
I was able to modify this code to estimate odds ratio for smoking categories.