Volume 39, Issue 1 , Pages e6-e8, January 2010
How Should We Conduct and Interpret Phase III Clinical Trials in Palliative Care?
Article Outline
To the Editor:
The article by Wildiers et al.1 raises some challenges in terms of ethical approaches to phase III end-of-life studies and their interpretation. Although the authors should be commended for undertaking a large, multisite, randomized, controlled trial in palliative care, there are fundamental questions that do need to be addressed before the first steps can be taken to adopt the study's findings into practice.
There is a need to understand the natural history of secretions in the terminal stages of a life-limiting illness.2 What happens when there is no pharmacologic intervention? What is the contribution of the cumulative effect of other medications with anticholinergic side effects?3 What is the contribution of comorbid illnesses, such as cardiac or respiratory failure? How many people were receiving parenteral fluids? How many had widespread edema or hypoalbuminemia? Given the large number of observational studies that have been done in hospice and palliative care, it is disappointing that the natural history of secretions at the end of life is still so poorly defined.
Combining all these factors, what are the characteristics of people who do not get secretions compared with those who do, and in light of this article, which factors are associated more strongly with a clinically meaningful response when pharmacologic therapy is introduced? This is brought into sharper focus when the data to support reduction in the volume of secretions in other settings using these medications suggest that they are relatively ineffective.4 The study by Wildiers et al.1 lacked a placebo arm. Given the current state of knowledge in this field, the high risk of side effects with each of these medications, and the background evidence that intervention and placebo may have the same effect on terminal secretions in this population, the lack of a placebo is a significant omission.
Also of concern is lack of information on standardization. Given the enormous effort to put this study together, why wasn't the intervention blinded? How was the assessment of the primary outcome standardized across multiple sites? Nurses are an exceptionally well-trained group of clinicians in hospice and palliative care, but we cannot necessarily depend on their ability to discern, for example, pulmonary edema requiring diuretics from terminal secretions. Retreatment decisions were also left to nurses' subjective decisions and could likely be influenced by worried family members or other patients in the same room.
As this was a multisite study, standardization across sites is an imperative for a primary outcome measure that is inherently subjective and measured by a third party. Although we commend the authors for trying to conduct a large prospective study addressing such a fundamental area to our discipline, the importance of the question increases the importance of ensuring that the study design and methods are able to be transferred across multiple sites.
These basic study design issues are magnified further by lack of informed consent in the first 81 participants. Ethically, how can people not be consented when it is unknown whether the net clinical benefit of these interventions (extrapolated almost exclusively from preoperative approaches in people who are well to normal physiological secretions) offer benefit that the patient or their caregivers will be able to perceive? There is sufficient equipoise to do the study. If an argument “that the therapy is in widespread use” was an acceptable yardstick, then almost any phase III study in hospice and palliative care of off-license prescribing could be justified without prior patient or proxy consent.
Consent in this setting is not a barrier to participation but rather an invitation by a researcher to a person or his/her family to engage in a process of exploration in an open and informed way. The communities in which we live have asked all clinical researchers, irrespective of how widespread the intervention being evaluated is in existing practice, to engage in a respectful and inclusive dialogue about an individual's involvement in research. Failure to engage in seeking consent risks compromising the community's willingness to allow hospice and palliative care clinical research in the future, even if that research is to directly improve the quality of care. The fact that the subject is unconscious does not revoke this fundamental right unless the intervention is immediately life saving.
Several options have been demonstrated in clinical research, including research in hospice and palliative care settings, to work well, and are acceptable to Institutional Review Boards and Research Ethics Committees.5 They include
How should these results be interpreted by clinicians? One could argue that the medications are equally ineffective in controlling secretions given a lack of knowledge of the natural history of the symptom, little basic science evidence to support their use in the pathologic state of dying and given an expected placebo response of up to 30% in such studies. The understanding of the natural history of secretions in this setting and the inclusion of a blinded placebo is not only desirable but also mandatory if the findings of this study are to influence practice.2
References
- Atropine, hyoscine butylbromide, or scopolamine are equally effective for the treatment of death rattle in terminal care. J Pain Symptom Manage. 2009;38(1):124–133
- . Interventions for noisy breathing in patients near to death. Cochrane Database Syst Rev. 2008;(1):CD005177
- Changes in anticholinergic load from regular prescribed medications in palliative care as death approaches. Palliat Med. 2009;23(3):257–265
- . Reducing gastric secretions—a role for histamine 2 antagonists or proton pump inhibitors in malignant bowel obstruction?. Support Care Cancer. 2009 Mar 17;[Epub ahead of print]
- . Approval by MREC of a modified patient information and consent form. Does this set a precedent for trials in palliative care?. Palliat Med. 2004;18(5):484–485
- . Novel consent process for research in dying patients unable to give consent. BMJ. 2003;327(7408):198
- . Ethical research in palliative care: A guide through the Human Research Ethics Committee process. Wollongong, Australia: Centre for Health Service Development, University of Wollongong; 2004;
PII: S0885-3924(09)00795-7
doi:10.1016/j.jpainsymman.2009.08.003
© 2010 U.S. Cancer Pain Relief Committee. Published by Elsevier Inc. All rights reserved.
Volume 39, Issue 1 , Pages e6-e8, January 2010
