BACKGROUND:
Osteoarthritis (OA) leads to joint failure and total joint replacement (TJR, either hip (H) or knee (K)). Worsening of pain and joint space narrowing are believed to be surrogates for joint failure; however, we hypothesize that TJR, as a reflection of joint failure, can be used as an endpoint in event-driven clinical trials within a reasonable duration. We explored the incidence of TJR in the Prospective Epidemiologic Risk Factor (PERF I) study.

METHODS:
A total of 5855 Danish postmenopausal women aged 49-88 enrolled in the PERF I study during 1999-2001 (baseline). Three-, six- and twelve-year follow-up data from the Danish National Patient Registry was collected, including occurrence of TJR and OA diagnosis. At baseline the women were asked whether they had OA.

RESULTS:
The women with a TJR diagnosis before or after baseline were on average 1 year older (p < 0.001) and heavier (p < 0.001), compared to women with no TJR. The 3-, 6- and 12-year cumulative incidences were 1.1, 2.4 and 6.0% for TKR, and 2.1, 4.4 and 9.3% for THR. For those with an OA diagnosis at baseline the respective incidences were 2.7, 5.6 and 11.7% and 3.9, 7.2 and 13.6% CONCLUSIONS: Within 3, 6 or 12 years TJR incidences were double for women with an OA diagnosis compared to the all-comer population. TJRs are frequent amongst elderly women with OA and it is, therefore, feasible to conduct event-driven clinical trials where TJR is the endpoint demonstrating clinical benefit of a novel disease-modifying OA drug (DMOAD).