Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
st:ODDSRISK module and continous covariates
From
Nanlesta Pilgrim <[email protected]>
To
[email protected]
Subject
st:ODDSRISK module and continous covariates
Date
Fri, 4 Feb 2011 00:44:51 -0500
I'm new to the ODDSRISK module. My dataset is longitudinal and I'm
using log binomial regression (specifically GEE) to model my outcome
which is not rare. The problem is that the adjusted model does not
converge. As such, we decided to do logistic regression and then
convert the ORs to RRs using Zhang's formula. In reading Zhang's
paper and the description of ODDSRISK, all the examples are based on
the predictors being binary or categorical. When I compare the ORs
from the logistic regression to that of the ODDSRISK they change
somewhate in the reprint (see below). As such, I'm wondering if
anyone can point me to sources on how to appropriately transform ORs
to RRs for continous predictors in multivariate models?
. xi:xtgee marrphys disalcdays i.rdisoccup disageyrs i.rdiseducyrs
i.disipvatt manypartner ///
> i.rcrossgenrltn i.newsaidsrsk i.rdiswealth newdispoly disfamsize i.marrstructure i.newmarrparstatus, nolog f(bin) l(logit)
> corr(uns) vce(robust) eform
i.rdisoccup _Irdisoccup_0-15 (naturally coded; _Irdisoccup_0 omitted)
i.rdiseducyrs _Irdiseducy_0-1 (naturally coded; _Irdiseducy_0 omitted)
i.disipvatt _Idisipvatt_0-3 (naturally coded; _Idisipvatt_0 omitted)
i.rcrossgenrltn _Ircrossgen_0-1 (naturally coded; _Ircrossgen_0 omitted)
i.newsaidsrsk _Inewsaidsr_0-7 (naturally coded; _Inewsaidsr_0 omitted)
i.rdiswealth _Irdiswealt_0-2 (naturally coded; _Irdiswealt_0 omitted)
i.marrstructure _Imarrstruc_0-1 (naturally coded; _Imarrstruc_0 omitted)
i.newmarrpars~s _Inewmarrpa_0-3 (naturally coded; _Inewmarrpa_0 omitted)
GEE population-averaged model Number of obs = 1109
Group and time vars: id time1 Number of groups = 983
Link: logit Obs per group: min = 1
Family: binomial avg = 1.1
Correlation: unstructured max = 2
Wald chi2(23) = 46.59
Scale parameter: 1 Prob > chi2 = 0.0025
(Std. Err. adjusted for clustering on id)
------------------------------------------------------------------------------
| Semirobust
marrphys | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
disalcdays | 1.633986 .3207118 2.50 0.012 1.112188 2.400593
_Irdisoccu~1 | 1.112781 1.476963 0.08 0.936 .0825341 15.00327
_Irdisoccu~3 | 1.282093 1.700855 0.19 0.851 .0952129 17.26408
_Irdisocc~15 | 1.652254 2.21541 0.37 0.708 .1193305 22.87718
disageyrs | .9908942 .0936334 -0.10 0.923 .823368 1.192506
_Irdiseduc~1 | 1.329675 .3183714 1.19 0.234 .8316442 2.125952
_Idisipvat~1 | 3.318917 1.477306 2.70 0.007 1.387114 7.941099
_Idisipvat~2 | 3.637085 1.603931 2.93 0.003 1.532425 8.632326
_Idisipvat~3 | 3.059167 1.465666 2.33 0.020 1.196154 7.823824
manypartner | 1.419517 .4760212 1.04 0.296 .7356958 2.738941
_Ircrossge~1 | .887271 .1854241 -0.57 0.567 .5890743 1.336419
_Inewsaids~1 | 1.798441 .7136119 1.48 0.139 .8263116 3.914249
_Inewsaids~2 | 2.751564 1.126913 2.47 0.013 1.233011 6.140338
_Inewsaids~3 | 2.001107 .9327574 1.49 0.137 .8026158 4.989224
_Inewsaids~7 | .7764457 .5858674 -0.34 0.737 .1769443 3.407105
_Irdisweal~1 | 1.504489 .368856 1.67 0.096 .9304643 2.432643
_Irdisweal~2 | 1.417855 .3204714 1.54 0.122 .9104129 2.208132
newdispoly | 2.026248 .728694 1.96 0.050 1.00133 4.100227
disfamsize | .9496166 .0639217 -0.77 0.442 .8322451 1.083541
_Imarrstru~1 | 2.023349 .7418727 1.92 0.055 .9862177 4.151154
_Inewmarrp~1 | 1.144711 .3113246 0.50 0.619 .6717321 1.950721
_Inewmarrp~2 | 2.019597 .6692795 2.12 0.034 1.054831 3.866754
_Inewmarrp~3 | 1.196413 .3741348 0.57 0.566 .6481846 2.208328
------------------------------------------------------------------------------
.
.
. xi:oddsrisk marrphys disalcdays i.rdisoccup disageyrs i.rdiseducyrs
i.disipvatt manypartner ///
> i.rcrossgenrltn i.newsaidsrsk i.rdiswealth newdispoly disfamsize i.marrstructure i.newmarrparstatus
i.rdisoccup _Irdisoccup_0-15 (naturally coded; _Irdisoccup_0 omitted)
i.rdiseducyrs _Irdiseducy_0-1 (naturally coded; _Irdiseducy_0 omitted)
i.disipvatt _Idisipvatt_0-3 (naturally coded; _Idisipvatt_0 omitted)
i.rcrossgenrltn _Ircrossgen_0-1 (naturally coded; _Ircrossgen_0 omitted)
i.newsaidsrsk _Inewsaidsr_0-7 (naturally coded; _Inewsaidsr_0 omitted)
i.rdiswealth _Irdiswealt_0-2 (naturally coded; _Irdiswealt_0 omitted)
i.marrstructure _Imarrstruc_0-1 (naturally coded; _Imarrstruc_0 omitted)
i.newmarrpars~s _Inewmarrpa_0-3 (naturally coded; _Inewmarrpa_0 omitted)
---------------------------------------------------------------------
Incidence for unexposed risk group = 0.1373
---------------------------------------------------------------------
Predictor Odds Ratio Risk Ratio [95% Conf. Interval]
---------------------------------------------------------------------
disalcdays 1.6031 1.4805 1.0533 2.0248
_Irdisoccup_1 1.1745 1.1470 0.1436 4.6239
_Irdisoccup_3 1.3855 1.3158 0.1650 4.9334
_Irdisoccup_15 1.8436 1.6523 0.2130 5.3963
disageyrs 1.0244 1.0210 0.8633 1.2021
_Irdiseducy_1 1.2594 1.2161 0.8245 1.7434
_Idisipvatt_1 3.8656 2.7744 1.4022 4.4693
_Idisipvatt_2 3.9932 2.8304 1.4420 4.5213
_Idisipvatt_3 3.1882 2.4518 1.1189 4.2725
manypartner 1.6944 1.5470 0.9356 2.4063
_Ircrossgen_1 0.8891 0.9029 0.6262 1.2783
_Inewsaidsr_1 1.9766 1.7430 0.8445 3.1325
_Inewsaidsr_2 3.1465 2.4305 1.2359 4.0136
_Inewsaidsr_3 2.1753 1.8731 0.8036 3.5800
_Inewsaidsr_7 0.8583 0.8753 0.2206 2.7236
_Irdiswealt_1 1.4401 1.3581 0.9031 1.9714
_Irdiswealt_2 1.4018 1.3286 0.9154 1.8733
newdispoly 2.1276 1.8424 1.0139 3.0218
disfamsize 0.9578 0.9634 0.8597 1.0774
_Imarrstruc_1 1.7927 1.6168 0.8463 2.7852
_Inewmarrpa_1 1.2046 1.1717 0.7470 1.7724
_Inewmarrpa_2 2.1030 1.8265 1.0965 2.8210
_Inewmarrpa_3 1.2156 1.1807 0.7026 1.8907
---------------------------------------------------------------------
.
Thanks!
Nanlesta
*
* 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/