Table 4.
Multivariate Cox Proportional Analysis of Metformin, Statin, and Finasteride Effects on PCa Incidence Rate in Men with Type 2 Diabetes: Comparison between Patients Without Missing Covariates and Patients With Missing Covariates.
| No Missing HR |
Covariates p-value |
With Missing HR |
Covariates p-value |
|
|---|---|---|---|---|
| Metformin Use | 0.89 | 0.02 | 0.87 | <.001 |
| Statin use | 0.80 | <.001 | 0.75 | <.001 |
| Finasteride Use | 0.44 | <.001 | 0.50 | <.001 |
| Metformin*Statin | 0.82 | .001 | 0.84 | <.001 |
| Statin*Finasteride | 1.28 | 0.02 | 1.09 | 0.10 |
| Metformin* Finasteride | 1.60 | <.001 | 1.48 | <.001 |
| Age | 1.03 | <.001 | 1.03 | <.001 |
| Black | 1.82 | <.001 | 1.83 | <.001 |
| Maximum PSA | 1.003 | <.001 | 1.003 | <.001 |
| Charlson Comorbidity | 1.03 | <.001 | 1.02 | <.001 |
| ΔA1c | 1.16 | <.001 | 1.17 | <.001 |
| ΔLDL | 1.002 | <.001 | 1.003 | <.001 |
| ΔBMI | 1.01 | <.001 | 1.01 | <.001 |
Results adjusted for propensity scores of metformin use and propensity scores of having missing covariates.