Sampling ratio | Method I† | Method II‡ |
---|---|---|
(Controls:case) | Bias | Bias |
Values obtained after 100 replications of nested case-control analyses conducted using two methods for incidence density sampling with 10, 50, and 100 controls per case. | ||
Note the value reported as bias is the ratio of the average parameter estimate derived via conditional logistic regression (obtained after 100 iterations of the nested case-control analysis) to the parameter estimate derived via Cox proportional hazards regression using data for the full cohort. | ||
†The incidence density sampling program shown in the Appendix of this paper. | ||
‡The incidence density sampling program proposed by Pearce5 with follow up time partitioned in units of person-years. | ||
10:1 | 1.00 | 1.01 |
50:1 | 1.00 | 1.14 |
100:1 | 1.00 | 1.32 |