Biomedical Engineering Reference
In-Depth Information
Table 1
Simulation Results: Non-proportional Rates Model
b
b
2
n
BIAS
ASE
ESD
ECP
BIAS
ASE
ESD
ECP
30
0
0.017
0.429
0.345
0.903
0.001
0.610
0.504
0.913
50
0.012
0.294
0.265
0.927
0.004
0.422
0.384
0.934
100
0.007
0.197
0.187
0.936
0.009
0.286
0.269
0.937
200
0.004
0.135
0.132
0.946
0.007
0.191
0.189
0.945
30
0.5
0.010
0.609
0.460
0.888
-0.002
0.718
0.549
0.887
50
0.010
0.420
0.361
0.904
0.009
0.481
0.416
0.913
100
-0.008
0.279
0.259
0.930
-0.005
0.315
0.294
0.932
200
0.002
0.190
0.185
0.946
0.000
0.215
0.207
0.945
30
1.0
0.010
0.790
0.544
0.868
0.004
0.827
0.587
0.881
50
0.015
0.513
0.428
0.896
0.001
0.529
0.442
0.905
100
0.005
0.335
0.312
0.927
0.000
0.346
0.316
0.928
200
-0.004
0.236
0.223
0.938
-0.004
0.234
0.224
0.940
30
2.0
-0.006
1.331
0.668
0.846
-0.004
1.204
0.653
0.862
50
0.015
0.677
0.522
0.880
0.004
0.627
0.489
0.883
100
0.001
0.436
0.384
0.913
-0.009
0.402
0.350
0.910
200
-0.001
0.292
0.280
0.937
-0.001
0.270
0.253
0.931
given by:
d
i
(t) = d
0
(t) expf
1
LBW
i
+
2
RDS
i
+
3
T T N
i
+
4
ASP H:mod
i
+
5
ASP H:sev
i
+
6
MALE
i
+
7
MALE
i
(t2:5yrs)g:
(14)
Having selecting the nal model, tted means and their corresponding
point-wise 95% condence intervals were computed for selected covariate
patterns.
A risk factor prole of the study population is presented in Table 2, with
respect to all covariates retained in the nal model. Each of the birth con-
ditions studied was relatively rare. A summary of the event history of the
Search WWH ::
Custom Search