Goals of Work:Multiple sites enable more successful completion of adequately powered phase III studies in palliative care. Audits of the frequency and distribution of the symptoms of interest can better inform research planning by determining realistic recruitment goals for each site. The proposed studies are to improve the evidence-base for registration and subsidy applications for frequently encountered symptoms where current pharmacological interventions are being used 'off-licence'.Methods: Six services participated in a standardized, retrospective, consecutive cohort audit of five symptoms of their inpatient populations to inform the design of double blind randomised controlled phase III studies to which each site would recruit simultaneously. The audit covered all deaths in a three month period for people who were referred to a specialist palliative care service who had at least one inpatient admission between referral and death regardless of when the person was referred to the service. The audits were based around inclusion and exclusion criteria for the proposed studies.Main Results: Of the 468 people whose medical records were reviewed, potential study participant rates varied by symptom having accounted for general and specific inclusion and exclusion criteria: pain 17.7%; delirium 5.8%; anorexia 5.1%; bowel obstruction 2.8% and cholestatic itch 0%. For those people with a symptom of interest, it was noted at the beginning of the inpatient admission more than half the time. Of all inpatients, fewer then one third would be eligible to participate in at least one study.
Conclusions:These data provide a baseline estimate of potential people to approach about clinical trials in supportive care but do not account for clinician 'gate-keeping', lack of interest in participating nor withdrawal from the study once initiated. The data are retrospective and therefore limited by clinical documentation. The audit directly informed an increase in the number of participating sites.
Page 2 of 11Archived at Flinders University: dspace.flinders.edu.au
INTRODUCTIONLarge scale trials in supportive, palliative and hospice care are feasible, especially if protocols can be designed appropriately for use in several sites simultaneously [6]. One reason for the failure of many studies in palliative care is an over-estimation of likely recruitment even for frequently encountered symptoms. A more comprehensive understanding of patterns of symptom occurrence for each participating site (given local variations in referral patterns) and general factors that may affect potential phase III study participation in a supportive and palliative care population need to be included in feasibility assessment. The design needs to take careful account of such findings if studies are going to be successful, and key performance indicators for each site can be tailored to local symptom patterns to monitor trial progress.