How predictive is it really? p-values and confidence intervals of hazard ratios only tell you the probability of comparisons to null effects on hypothetically r

How predictive is it really? p-values and confidence intervals of hazard ratios only tell you the probability of comparisons to null effects on hypothetically repeated data. What I really want to know is, under the prior of proportion of PD in the general population, what is confusion matrix for this method?