Table 3. Effect of ever filling a prescription of an NSAID class on the risk of developing prostate cancer.

Variable	Unadjusted ORa (95%CI)	P-value	Adjusted ORb (95%CI)	P-value
Aspirin	1.13 (1.08–1.18)	<0.001	1.01 (0.95–1.07)	0.816
Arylacetic acids	1.11 (1.06–1.17)	<0.001	0.94 (0.88–1.00)	0.043
Butylpyrazolidines	1.10 (1.04–1.17)	0.002	0.99 (0.92–1.07)	0.776
Oxicams	1.10 (1.03–1.17)	0.002	0.96 (0.89–1.04)	0.368
Propionates	1.10 (1.05–1.15)	<0.001	0.89 (0.84–0.95)	<0.001
Coxibs	1.09 (0.70–1.69)	0.712	Excludedc
Fenamates	1.12 (0.97–1.30)	0.119	Excludedc
NA-NSAIDs	1.17 (1.11–1.24)	<0.001	0.88 (0.82–0.94)d	<0.001
NSAIDs	1.21 (1.13–1.28)	<0.001	0.87 (0.80–0.94)	<0.001

a) ORs from unadjusted conditional logistic regression models for comparison.

b) Adjusted for ever visited a urologist 1–11 years prior, SCREENED and volume of family physician visits in the 5 years prior to the index date and, when appropriate, for use of other NSAID classes.

c) Fenamates and Coxibs were excluded from this model because of small numbers.

d) From an adjusted model that included terms for NA-NSAIDs and aspirin in addition to screening predictors as above.

Note: Effect estimates throughout the paper have been rounded to two decimal digits. This is not meant to imply that our results are accurate to two decimal digits (most certainly they are not). However, rounding to one single digit would have made it difficult to spot any trends in the data.