Objective: To determine the rate of conversion of prostate-specific antigen (PSA) to values of > 4 ng/mL in a normal male population.
Patients and methods: In men with an initial PSA level of < 4 ng/mL the probability of having a PSA of > or = 4 ng/mL and its relationship with the age at first PSA test, baseline PSA level and time to the determination was evaluated. Statistical methods were used to correct for the within-subject variability across many PSA determinations.
Results: In all, 45433 PSA tests in 21169 men with an initial PSA of < 4 ng/mL were available. Overall, in 960 men (4.5%) the PSA level changed to > or = 4 ng/mL. The lowest chance was for men aged < 55 years with an initial PSA of < 1 ng/mL and a new PSA determination during the first 18 months (0.952 conversions per 1000 men) while the highest was for men aged < 55 years with an initial PSA level of > 2.5 ng/mL (395 conversions per 1000 men). There was a significant and independent relationship between conversion and the interval between determinations, with conversion being 3.33 times more frequent during 18-42 months than with the reference category (< 18 months) and 5.23 times more frequent during > 42 months. The probability of conversion independently increased by 3 lambda for every additional year of age. Every unit of basal PSA multiplied by 6.48 the probability of conversion.
Conclusions: The risk for conversion to a PSA of > or = 4 ng/mL can be anticipated from the patient's age and initial PSA level. Re-screening before 4 years after the first round can be useless for men with an initial PSA of < 1 ng/mL. Four-year intervals could be inadequate for men with an initial PSA of > 2.5 ng/mL if conversion to > 4 ng/mL is to be avoided.