Effect of average annual NSAID dose (in DDDs/year; categorized using the quintiles of the distribution) on the risk of developing prostate cancer.

a) For each class, results from two separate models are reported. In one model, the dose quintiles were entered as an ordinal variable (a linear term). In the second model, each level (quintile) of the ordinal dose variable was represented in the model with a binary indicator variable. In the analyses shown in the left panel, the reference group is men who did not fill any prescriptions of the index class (never-users).b) Adjusted for ever having visited a urologist 1–11 years prior, SCREENED and volume of family physician visits in the 5 years prior to the index date, and for all NSAID classes listed in the table.c) iOR: Incremental OR. The ORs in the right panel are incremental ORs from models that contrast the effect at each level with that at the previous level. iORs consistently (at all levels) above (or below) 1.0 suggest a monotonic increasing (or decreasing) dose-response relationship.