I would like to calculate age adjusted incidence rates per 1000 person-years
for each category of my exposure variable, but I'm having difficulty getting
them and would be very grateful if anyone could point me in the right
direction.
This is what I have done so far. I have individual person data with years of
follow-up. I -stsplit- the data into seven agebands, and calculated the
number of person-years for each ageband. Then I fitted a poisson regression
model, including the outcome variable (disease/no disease), agebands,
exposure (4 categories) and person-years of follow-up (Stata output given
below). Then I used -adjust- with the -exp ci- options to get age-adjusted
incidence rates with confidence intervals for each of my four categories of
exposure. Is this the correct/best method of getting age-adjusted incidence
rates? The incidence rates given by -adjust- are per person-year, but I
would like incidence rates per 1000 person-years together with a 95%
confidence interval. Can anyone suggest how I could calculate these?
I also have a query about the output from -adjust-. I imagined that the
incidence rates given by the -exp- option were simply the exponentiated
values of the predicted values given by -xb-. But judging from my output,
this clearly isn't the case. Any insights into this would be much
appreciated.
Many thanks.
Anne-Helen
______________________________
. stset exit, failure(CC) id(id) origin(time entry) scale(365.25)
. . . . . . . . . . .
. stsplit ageband, at(45(5)70) after(time=dob)
. . .. . . . . . . . .
. gen pyear = _t - _t0
. xi: poisson CC i.ageband i.exp4, e(pyear)
i.ageband _Iageband_0-70 (naturally coded; _Iageband_0 omitted)
i.exp4 _Iexp4_1-4 (naturally coded; _Iexp4_1 omitted)
Iteration 0: log likelihood = -1037.4761
Iteration 1: log likelihood = -1037.4761
Poisson regression Number of obs =
13514
LR chi2(9) =
17.74
Prob > chi2 =
0.0383
Log likelihood = -1037.4761 Pseudo R2 =
0.0085
----------------------------------------------------------------------------
--
CC | Coef. Std. Err. z P>|z| [95% Conf.
Interval]
-------------+--------------------------------------------------------------
--
_Iageband_45 | 1.040127 1.069059 0.97 0.331 -1.055189
3.135444
_Iageband_50 | .3947027 1.030789 0.38 0.702 -1.625607
2.415012
_Iageband_55 | 1.056793 1.019859 1.04 0.300 -.9420931
3.055679
_Iageband_60 | 1.192617 1.017726 1.17 0.241 -.8020903
3.187324
_Iageband_65 | 1.324562 1.015631 1.30 0.192 -.6660373
3.315162
_Iageband_70 | .8593928 1.008605 0.85 0.394 -1.117437
2.836222
_Iexp4_2 | -.0992991 .2107329 -0.47 0.637 -.512328
.3137298
_Iexp4_3 | -.0720948 .2074217 -0.35 0.728 -.4786339
.3344442
_Iexp4_4 | -.4293346 .2291571 -1.87 0.061 -.8784743
.019805
_cons | -6.237414 1.009309 -6.18
0.000 -8.215623 -4.259204
pyear | (exposure)
----------------------------------------------------------------------------
--
. adjust, by(exp4) exp ci
----------------------------------------------------------------------------
----
Dependent variable: CC Equation: CC Command: poisson
Variables left as is: _Iageband_45, _Iageband_50, _Iageband_55,
_Iageband_60, _Iageband_65,
_Iageband_70, _Iexp4_2, _Iexp4_3, _Iexp4_4
----------------------------------------------------------------------------
----
----------------------------------------------
exp4 | exp(xb) lb ub
----------+-----------------------------------
1 | .01149 [.008532 .015472]
2 | .012355 [.009037 .016892]
3 | .010575 [.007791 .014355]
4 | .008619 [.006015 .012349]
----------------------------------------------
Key: exp(xb) = exp(xb)
[lb , ub] = [95% Confidence Interval]
. adjust, by(exp4) se ci
----------------------------------------------------------------------------
----
Dependent variable: CC Equation: CC Command: poisson
Variables left as is: _Iageband_45, _Iageband_50, _Iageband_55,
_Iageband_60, _Iageband_65,
_Iageband_70, _Iexp4_2, _Iexp4_3, _Iexp4_4
----------------------------------------------------------------------------
----
----------------------------------------------------------
exp4 | xb stdp lb ub
----------+-----------------------------------------------
1 | -7.66225 (.15185) [-7.95987 -7.36462]
2 | -4.78092 (.159579) [-5.09368 -4.46815]
3 | -4.56331 (.155906) [-4.86888 -4.25774]
4 | -5.91873 (.183514) [-6.27841 -5.55905]
----------------------------------------------------------
Key: xb = Linear Prediction
stdp = Standard Error
[lb , ub] = [95% Confidence Interval]
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
