Both Sides Now: Disclosure of Prognosis on Dialysis

Long before its publication in 2000, I was honored to be asked by the Renal Physicians Association/American Society of Nephrology (RPA-ASN) Working Group to review the draft clinical practice guideline, “Shared Decision-Making in the Appropriate Initiation of and Withdrawal from Dialysis” (2). Generally, I believed these were important, thoughtful, and clearly articulated. The draft document included the recommendation that all patients requiring dialysis should have their chances for survival estimated, and that these estimates should be discussed with the patient and their family. I can remember reacting rather viscerally to this particular proposal; these are the exact comments I submitted back to the Working Group: “The recommendation that discussions about life expectancy and quality of life should occur will be controversial, since it is likely not done now by most nephrologists, and since there is no data to support or reject this approach. An argument can be made that this may be too much information, and/or too distressing for patients, or families. I think this weakness should be acknowledged, with a call for research into this area.”

Looking back, my words ring with paternalism. Paternalism describes a view of the doctor–patient relationship oriented toward therapy, reassurance, and avoiding harm. Disclosure and sharing of information may be ignored if it is felt to aid the goals of therapy. This model was the norm in North America until late in the 20th century (3).

I know that I am not alone in harboring some residual paternalism about complete disclosure of the prognosis with end-stage renal disease (ESRD). I believe the roots of this discomfort are deep. From the ancient times of the Hippocratic oath to the core principles that are the foundation of modern medical bioethics, beneficence and non-maleficence are sacred duties for physicians. Beneficence is the duty of a health care provider to be of benefit to the patient, while the principle of non-maleficence means that health care providers will not intentionally create needless harm or injury to a patient through acts of commission or omission. In other words, doctors are expected to protect their patients from harm. If the disclosure of distressing information that a patient did not specifically request causes pain or emotional distress to that patient, then a physician might feel that they have violated a fundamental tenet of the doctor–patient relationship.

At the very least, it is apparent that full disclosure can lead to a perceived conflict between non-maleficence and another fundamental moral principle of medical ethics: respect for the autonomy of the individual. The rise of the patient rights movement and the evolution of bioethics have led to major changes concerning the duty to disclose. Indeed, modern bioethics favors full disclosure as a means of respecting and enhancing patient autonomy and decision-making (4).

Reservations about full disclosure are still common among nephrologists. Fine and his colleagues report in this issue of Peritoneal Dialysis International (PDI) that, at an ASN-ISN session in San Francisco in 2001, most nephrologists did not favor voluntary disclosure of outcomes to patients starting dialysis (5). Similarly, preliminary data from a survey of Canadian Society of Nephrology (CSN) members about chronic kidney disease (CKD) care show this ambivalence clearly. When asked whether they tell patients that dialysis is a life-extending treatment and that they will die of kidney failure in weeks or months after the kidney function falls below a certain level, 93% said “always” or “usually.” Similarly, 80% said they usually or always give patients who specifically ask for it detailed and blunt information about prognosis, for example, that life expectancy is reduced on average to 1/3 that of another person of the same age without kidney disease. However, if patients do not specifically ask for this detailed and blunt prognostic information, only 38% always or usually give it (Abstract submitted to CSN annual general meeting, Calgary, May 2005).

As a mid-career nephrologist with some formal ethics training, I do not usually utilize the paternalistic approach that was the decision-making style of the past. For example, I discuss prognosis and all treatment options with patients diagnosed with glomerulonephritis and I disclose all treatment options for ESRD, including palliative care and a trial of dialysis. I encourage advanced care planning (ACP) and advance directives. Ironically, and perhaps analogously, 10 years ago many physicians had reservations that ACP would be too distressing for patients. It is noteworthy that empirical research has shown that this is not the case (6-11).

In this issue of PDI, Fine et al. report on a survey of general nephrology patients with mild kidney disease (5). These patients are absolutely and unambiguously clear about the disclosure issues that trouble physicians so much. They want to know actual life expectancy on dialysis and they want as much information as possible, both good and bad. Finally, they say that the doctor should volunteer life-expectancy information without having to be prompted by a patient.

Given the trends toward disclosure and empowering patient autonomy, it seems regressive that the Health Research Ethics Board at the University of Manitoba refused to allow Dr. Fine to apply his study instrument to seek the opinions of patients with advanced CKD. It now behooves other investigators to extend these findings into the more relevant population in the late stages of CKD. However, it seems likely that these patients will also want all the prognostic information. It is not correct for nephrologists to think that, because they disclose that without dialysis a patient will die (one side of the coin), this then supersedes disclosure of information about suboptimal survival and reduced quality of life (the other side of the coin).

One important benefit may flow from the discomfort that nephrologists will feel as they tell more patients and their families about the limited life expectancy on dialysis, and the significant compromises in quality of life that occur with the imperfect technologies we now employ. The constant reminders of these deficiencies in dialysis therapies may be an unpleasant blow to the ego of physicians who work in this life sustaining, high cost, and high technology niche of medicine, but if this serves to stimulate renewed efforts to enhance the pool of transplant donors (live and cadaveric) and to markedly improve all renal replacement therapies, it will lead to tangible benefits for future patients.

There are some who believe that biomedical ethics is too soft and philosophical a discipline, that its arguments are not easily resolved, and that its fundamental questions can never be answered. Fine and colleagues did not accept this stalemate and should be congratulated for recognizing the challenge of validating or refuting an opinion-based ethical guideline. Their study is an excellent example of applying an empirical approach to examine a controversial recommendation about an uncomfortable clinical situation. It strongly suggests that any residual paternalism that exists within the renal community on this important issue should be exposed and then extinguished. We must not actively or passively create an illusion about health and quality of life on dialysis. In 2005, there is a clear duty to disclose both sides of the coin, as patients and their loved ones anxiously contemplate an uncertain future with ESRD.