st: Analysis of differences between two surgical methods in three timeperiods

[roland andersson <rolandersson@gmail.com>]

I am analysing the risk of complications after laparoscopic
appendectomy compared to open appendectomy in three timeperiods with
different usage of the laparoscopic technique. The model looks like
this

xi:logistic perpostkirkomplik  i.lapscopiintention*i.period i.appdgn +
other covariates.

The result is

perpostkir~k    Odds Ratio      Std. Err.       z       P>z     [95%
Conf.      Interval]

_Ilapscopi~1    1.591794        .4547446        1.63    0.104
.9093168        2.786497
_Iperiod2       1.029331        .1664423        0.18    0.858
.7497497        1.413169
_Iperiod3       1.334   .2121616        1.81    0.070   .9767403        1.821933
_IlapXper_~2    .7850812        .2947154        -0.64   0.519
.3761646        1.638518
_IlapXper_~3    1.561686        .5171758        1.35    0.178   .816029 2.988697
_Iappdgn2_1     1.958182        .2390147        5.51    0.000
1.541543        2.487426
_Iappdgn2_3     1.533735        .3206868        2.05    0.041
1.018061        2.310612

I want to analyse the difference in komplikation between open and
laparoscopic operation at each of the timeperiods. It seems that there
is a slight increase in complications for open appendectomy as
suggested by the estimates for _period2 and _period3.

Is it correct to say that the difference in complikations is OR 1.59
at timeperiod 1 and to use

lincom  _Ilapscopii_1 + _IlapXper_1_3 -  _Iperiod3, or

to obtain the estimate of the difference open vs laparoscopy at timeperiod 3?

Roland Andersson

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/