Nephrologists’ Reported Preparedness for End-of-Life Decision-Making

Materials and Methods

In 2005, the 2500 nephrologist members of the RPA were invited by e-mail to participate in an online survey of their attitudes and practices toward end-of-life decision-making. The invitation was sent in January 2005. Monthly reminders to participate in the online survey were sent by e-mail in February, March, and April 2005. The 325 members of the Canadian Society of Nephrology also were invited to participate in the same online survey. All methods adhered to the Declaration of Helsinki. The West Virginia University School of Medicine Office of Research and Graduate Studies reviewed and approved the Institutional Review Board for the Protection of Human Subjects application for exemption.

Results

A total of 360 nephrologists responded: 296 American nephrologists (12% of RPA membership), 61 Canadian nephrologists (19% of the Canadian Society of Nephrology membership), and three unknown (Table 1). Compared with Americans, Canadian nephrologists were younger, more likely to be affiliated with a medical school (89 versus 46%; P < 0.001), and not practicing in for-profit units (0 versus 57%; P < 0.001). Canadian nephrologists reported stopping dialysis of a higher number of patients in the preceding year (6.3 versus 4.2; P = 0.018) and reported referring fewer to hospice (2.5 versus 3.8; P < 0.001) than American nephrologists. They practiced in units that were less likely to have policies on dialysis withdrawal (7 versus 30%; P < 0.001) and cardiopulmonary resuscitation (CPR; 47 versus 86%; P < 0.001). Canadians were slightly less aware of the RPA/ASN guideline (49 versus 61%; P = 0.075) and RPA/ASN statement Quality Care at the End of Life (41 versus 57%; P = 0.026). However, of those who were aware, there was comparable usage. Canadian nephrologists reported being more likely to stop dialysis of a patient with severe dementia (71 versus 53%; P = 0.009; Figure 1). Their practices regarding withholding dialysis from a permanently unconscious patient or stopping dialysis at the request of a competent patient were similar to that of American nephrologists.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Practices to withhold or stop dialysis: Comparison between American and Canadian nephrologists (A) and between nephrologists who are very well prepared and less than very well prepared for end-of-life decision-making (B).

Thirty-nine percent of overall respondents perceived themselves to be very well prepared to make end-of-life decisions. Compared with the less than very well prepared, nephrologists who reported being very well prepared were older, had been in practice longer, had withdrawn more patients from dialysis in the preceding year (5.6 versus 3.8; P < 0.001), were more likely to use time-limited trials of dialysis (87 versus 74%; P = 0.003), were less likely to consult an ESRD Network Ethics Committee for difficult patient treatment decisions (40 versus 57%; P = 0.002), and were more aware of the RPA/ASN clinical practice guideline (70 versus 52%; P < 0.001) and statement on Quality Care at the End of Life (62 versus 48%; P = 0.007; Table 2). Nephrologists who felt very well prepared to make end-of-life decisions reported being more likely to stop dialysis if a patient develops permanent and severe dementia (66 versus 50%; P = 0.003; Figure 1). Withholding dialysis of a permanently unconscious patient and stopping dialysis if requested by a competent patient did not vary by level of perceived preparedness. Multiple logistic regression analysis identified age ≥46 yr (odds ratio [OR] 2.77; 95% confidence interval [CI] 1.70 to 4.53; P < 0.001), six or more patients withdrawn from dialysis in the preceding year (OR 2.58; 95% CI 1.43 to 4.66; P = 0.002), and awareness of the RPA/ASN guideline (OR 1.78; 95% CI 1.10 to 2.90; P < 0.001) to be independently associated with the highest level of self-reported preparedness (Table 3).

Discussion

With the significant symptom burden, comorbidity, and limited survival of the aging dialysis population, nephrologists who care for patients with ESRD increasingly are deciding with patients and/or families whether to withhold or withdraw dialysis. In fact, withdrawal from dialysis precedes death in up to 30% of chronic dialysis patients (8). The Institute of Medicine Committee for the Study of the Medicare End-Stage Renal Disease Program recommended that guidelines that could assist nephrologists, patients, and families in reaching clinical decisions on whether to initiate or to stop dialysis be developed (9,10). In the past 10 to 15 yr, there has been an increase in the research that is devoted to these issues, and the RPA and the ASN published a clinical practice guideline, Shared Decision-Making in the Appropriate Initiation of and Withdrawal from Dialysis, and a position statement, Quality Care at the End of Life (5), to address the issues of appropriate initiation of and withdrawal from dialysis. Despite these efforts, most nephrologists (61%) do not feel very well prepared to make these difficult end-of-life decisions (Table 1).

Awareness of the RPA/ASN guideline concerning end-of-life decision-making was an independent predictor for nephrologists’ reporting greater preparedness in dealing with these issues. These nephrologists also were more likely to report making end-of-life decisions in accordance with the guideline such as offering time-limited trials of dialysis, withdrawing dialysis from a patient with severe dementia, and referring patients to hospice. Despite the existence of this guideline to aid decision-making, there continues to be considerable variation in reported practice, especially concerning the patient with dementia. Even among nephrologists who report being very well prepared, only 66% were comfortable withdrawing dialysis from a patient with severe dementia when no clear previous wishes had been established. Recommendation 7 in the RPA/ASN guideline notes that it is reasonable to consider withdrawing dialysis for patients who have “all forms of severe irreversible dementia” if they are “unable to cooperate with the dialysis process…” (11). Although only 66% of the surveyed nephrologists would feel comfortable following this guideline, in 1990, only 32% of surveyed nephrologists reported being comfortable withdrawing dialysis in such a patient (7). Comfort with stopping dialysis in competent patients who request to withdraw from dialysis and withholding dialysis from a permanently unconscious patient seem to be more uniform, with the vast majority (90 to 95%) supporting such decisions.

Nephrologists who reported greater preparedness had been in practice longer and were caring for a greater number of patients. This suggests that clinical experience rather than education during training prepares nephrologists to deal with the complex end-of-life decision-making that is required when caring for patients with ESRD. This is not surprising given that end-of-life care is not well addressed in renal fellowship training (12). Second-year fellows report feeling unprepared to care for their dying patients (12). Our findings also are supported by other survey data that suggest that personal experience with end-of-life care affects health professionals’ comfort in discussing end-of-life issues with patients (13). A core curriculum in palliative care that deals with pain and symptom management, psychosocial and spiritual support of patients and families, advance care planning, and ethical issues surrounding initiation and withdrawal of dialysis was developed recently for nephrology fellows (14). In conjunction with the published guideline, this core curriculum may help to address the deficiencies of training programs in this area of clinical practice by overcoming a lack of training and/or experience with end-of-life issues and facilitating the acquisition of skills in this arena.

There were differences in the reported practices of American and Canadian nephrologists. Canadian nephrologists reported withdrawing more patients from dialysis and seemed more comfortable withdrawing from dialysis a patient with severe dementia. This finding confirms a similar difference among Canadian and American nephrologists on an earlier survey that used hypothetical cases (15). This study cannot ascertain the reasons, but these differences may reflect a health care system with less financial incentive to continue dialyzing patients. Worries about medical malpractice also may contribute to these observed differences among American and Canadian nephrologists (15). Canadian nephrologists also more often are associated with a medical school rather than with for-profit organizations. This potentially could be another factor that contributes to ethical decision-making. However, Canadian nephrologists reported being less likely to refer to hospice. Differences in economic, cultural, and medical practice would have to be explored to discern possible factors for these observed differences.

Despite the significant symptom burden and the high mortality, hospice use among dialysis patients is low in both Canada and the United States. Recommendation 9 in the RPA/ASN clinical practice guideline stresses a palliative care approach that allows patients to decide whether they wish to die in a health care facility or at home with hospice care. Hospice care emphasizes supportive care of the patient and the family, enabling families to continue caring for their loved ones at home while they die. Hospice care provides for home nursing and pastoral care and focuses on treating symptoms while neither prolonging life nor hastening death. Factors that play a role in hospice underuse likely include a lack of physician knowledge about the benefits of hospice, patient and family refusal of hospice referral, availability of hospice services, and reimbursement policies to hospices that accept dialysis patients (16). A recent study demonstrated that 41.9% of patients who have ESRD and die after withdrawal of dialysis and 3.4% of patients who die while still receiving chronic dialysis have been referred to hospice programs before death (16). The RPA, the Forum of ESRD Networks, and the National Hospice and Palliative Care Organization are working with the Center for Medicare and Medicaid Services to improve understanding of the Medicare hospice coverage for patients with ESRD and dialysis providers so that patients can receive hospice while continuing dialysis. The lower rate of hospice referrals in Canada may underlie a philosophical difference in the provision of hospice care between Canada and the United States. The costs of dialysis care in Canada are borne by the provincial governments and would not be the responsibility of the hospice program. Nonetheless, Canadian hospice programs, as a rule, will not admit a dialysis patient while he or she is continuing with dialysis, despite a prognosis of <6 mo. Although providing hospice care for patients who remain on dialysis remains controversial, hospice programs, in conjunction with the nephrology team, can help patients come to terms with their impending death, help resolve conflicts, and aid in end-of-life decision-making that, for some, ultimately may result in the withdrawal of dialysis.

CPR policies in dialysis units vary tremendously and range from recognizing “no codes” to mandating CPR should an arrest occur in the dialysis unit, even if unwanted by the patient. Although beyond the scope of this survey, further research is required to determine how various CPR policies affect end-of-life care. Why so few Canadian dialysis units have policies on “do not resuscitate” and withdrawing dialysis is not clear. However, given that litigation is not as prevalent in Canada, there has not been the same legal incentive to establish these policies as there is in the United States. Nephrologists who practice in units where CPR is discussed routinely feel more prepared to discuss end-of-life issues, regardless of the existence of a unit policy. It is likely that the practice environment, as well as time in practice and practice experiences, affects nephrologists’ comfort with end-of-life decision-making.

There are several limitations to this study. The sample size was relatively small, and response rates for American and Canadian nephrologists were only 12 and 19%, respectively. However, response rates from online surveys are known to be lower than those for postal surveys and are considered a cost-effective administration method that provides equivalent results to mailed surveys with samples from professional organizations (17,18). Also the number of respondents in this study was comparable to that in other studies of nephrologists’ attitudes and practices that have yielded useful results (7,15). Nephrologists’ perception of their preparedness to make end-of-life decisions is not necessarily the same as their actual preparedness, training, and skill in this area. However, physicians’ comfort with dealing with these difficult issues remains an important factor in providing quality end-of-life care. Also, self-report of end-of-life practices may vary from actual practices because of difficulties in recall and/or nephrologists’ desire to provide socially acceptable answers. These results may overestimate the extent of nephrologists’ comfort with end-of-life decision-making because nephrologists who are most interested in end-of-life care may have been the ones who primarily answered this survey.

The results of this study suggest that increased clinical experience and knowledge of the RPA/ASN guideline and position statement seem to promote comfort with end-of-life decision-making and clinical practice in conformance with the recommendations in the guideline. Because only approximately 50% of the respondents were aware of the RPA/ASN clinical practice guideline and position statement, additional educational efforts likely are needed to inform American and Canadian nephrologists and nephrology fellows of them. While comfort with end-of-life decision-making is, in and of itself, an important outcome for nephrologists, actual end-of-life practices and skills and ultimately the quality of end-of-life care that is delivered by nephrologists who are knowledgeable of the guideline and position statement need to be assessed to determine their full impact and benefit.