How long should long be? Long-term trials in rheumatic diseases

In the interests of the safety of patients it is necessary to collect long-term clinical data. It is often assumed that the longer the trial the better. However, the longer the trial the more scientific compromises are necessary if it is to be carried out in a practical way. Because the therapeutic efficacy of and tolerance to the nonsteroidal anti-inflammatory drug diclofenac was established after 6 months' trial the optimal duration for a comparative long-term trial of this preparation is 6 months. The small amount of information gained from trials of diclofenac longer than this did not justify the reduction in quality of trial design and performance that they required. Thus it would appear that at least for 1 antirheumatic drug the assumption that 'the longer the trial the better' is not true.