Evidence in palliative care research: how should it be gathered?

To the Editor: Aoun and Kristjanson’s article highlighted some of the difficulties facing researchers in palliative care1 and questioned the role of randomised controlled trials (RCTs). These difficulties do not exempt palliative care from seeking to improve care through thoughtful research but, rather, highlight areas that need to be specifically designed to cope with these difficulties.

In our recently completed large RCT of palliative care in southern Adelaide, for which we recruited 461 patients,2 we were able to overcome many of the obstacles cited by these authors.

As they rightly stated, many trials in palliative care are pragmatic. This should not be seen as a bad thing, as pragmatic studies are designed to test clinically relevant interventions3 within diverse populations across different settings and to report on a broad range of health outcomes. Such studies are more in keeping with the realities of palliative care. We would argue that traditional RCTs that examine a specific intervention within a specific patient type are less applicable to a palliative care population. In such a population, unlike many other areas of health, the diagnosis and prognosis do not dictate care needs.

Among the methodological issues, recruitment and retention are perhaps the biggest hurdles to be overcome. We found that developing systematic, evidence-based protocols, which were pilot tested before initiating the trial, significantly enhanced recruitment.4 To minimise burden, we ensured that all data collection was kept to a minimum and, where possible, data were recorded by study staff or collected from other sources as part of routine clinical encounters. Sample size calculations need to allow for attrition caused by increasing severity of disease. This will, however, inflate the numbers required for a study. Use of multiple sites for recruitment is a strategy that we have used successfully to reach sample size goals.

The article states that “RCTs are seldom acceptable to patients and their families”.1 This has not been our experience. Patients are willing to participate in research and find it a meaningful way to give back to the community.5 The challenge is to ensure that this willingness to participate in research is not exploited by palliative care researchers. The other ethical concern raised by Aoun and Kristjanson is the use of a control arm in which patients “deliberately have support services withheld”. We would agree that, if that were to happen, it would be a serious ethical breach. For RCTs in palliative care, the control arm should consist of the standard care provided — the rationale for doing a trial is that there is equipoise regarding the benefit of the intervention.

We feel that RCTs are feasible and appropriate for palliative care research.