Abstract
This session focused on issues related to implementation of randomized clinical trials in palliative care studies. Topics discussed included what kinds of clinical sites and patient populations were suitable, what types of clinical investigators (clinical specialty) should be involved in or lead the studies, what multisite mechanisms could be used to conduct the trials, and what funding issues were related to these studies. A trial of operative versus nonoperative management for small bowel obstruction caused by recurrent intra-abdominal cancer was considered. The feasibility of such a trial was examined in terms of whether there was "equipoise" for a majority of likely investigators in the field around the trial question, what other issues might impact accrual to the trial, and how many patients would be required to answer which of these two treatment arms was better. This last question is related to selection of a primary endpoint for the trial and was a modestly contentious issue for the trial design group. Both sensible compromises in endpoint selection and the education of the community of investigators for a particular randomized trial in palliative care are crucial steps for successful implementation. A major conclusion of this session is that implementation considerations are intimately related to the architecture of a specific trial and should be addressed practically and early in the design phase of any randomized trial addressing a palliative care question. In this respect, randomized trials in palliative care are no different than in other fields.