Conditional Logistic Regression Model for Odds of Amenorrhea
Clinical Trial of Contracepting Women

The LOGISTIC Procedure
 
Conditional Analysis

Model Information
Data Set WORK.AMEN
Response Variable y
Number of Response Levels 2
Number of Strata 1151
Number of Uninformative Strata 664
Frequency Uninformative 1810
Model binary logit
Optimization Technique Newton-Raphson ridge

Number of Observations Read 4604
Number of Observations Used 3616

Response Profile
Ordered
Value
y Total
Frequency
1 1 1231
2 0 2385

Probability modeled is y=1.



Note: 988 observations were deleted due to missing values for the response, explanatory, or strata variables.

Strata Summary
Response
Pattern
y Number of
Strata
Frequency
1 0
1 0 1 144 144
2 1 0 54 54
3 0 2 82 164
4 1 1 50 100
5 2 0 23 46
6 0 3 31 93
7 1 2 27 81
8 2 1 15 45
9 3 0 11 33
10 0 4 261 1044
11 1 3 145 580
12 2 2 136 544
13 3 1 114 456
14 4 0 58 232


Newton-Raphson Ridge Optimization


Without Parameter Scaling

Convergence criterion (GCONV=1E-8) satisfied.

Model Fit Statistics
Criterion Without
Covariates
With
Covariates
AIC 1367.057 949.276
SC 1367.057 974.049
-2 Log L 1367.057 941.276

Testing Global Null Hypothesis: BETA=0
Test Chi-Square DF Pr > ChiSq
Likelihood Ratio 425.7810 4 <.0001
Score 384.0000 4 <.0001
Wald 301.3065 4 <.0001

Analysis of Maximum Likelihood Estimates
Parameter DF Estimate Standard
Error
Wald
Chi-Square
Pr > ChiSq
time 1 0.7587 0.3766 4.0577 0.0440
time2 1 0.0367 0.0742 0.2442 0.6212
time*trt 1 1.2107 0.5324 5.1720 0.0230
time2*trt 1 -0.2235 0.1040 4.6126 0.0317