Building an Evidence Base for Active Medical Management without Dialysis: Tale of Two Programs.

There is clinical equipoise about the utility of dialysis in older adults with multiple comorbidities, as it may not prolong life, and it is unclear if it improves quality of life.1 And yet, older adults are the fastest growing proportion of new dialysis patients and more than half will die within 1 year of initiating dialysis.2,3 The symptom burden in this population is high4,5 and equal to that of patients with terminal cancer.6 Patients on dialysis also struggle with existential and spiritual concerns7 and have poor quality of life.8 For older adults, dialysis also confers a 3–10-fold increased risk of frailty9 and a 3.5-fold increased risk of severe cognitive impairment.10 Although patients may expect their symptoms, quality of life, and function to improve with dialysis, quality of life is poor, overall symptom burden persists, and function declines.11,12 Palliative care (PC) improves the lives of patients with serious illness and their families through patient-centered care.13 In contrast to hospice, which serves patients in the final 6 months of life, PC provides support to patients at any stage of their disease and has been shown to improve survival,14 quality of life,15 and mood14–16 in patients with cancer and heart failure. Recent guidelines recommend that integrated PC for kidney disease17,18 and kidney palliative care (KPC) expertise can support active medical management with dialysis, but studies are limited. Two recently published Kidney360 original articles report data from their KPC programs, specifically addressing active medical management without dialysis or conservative kidney management (CKM) as a therapeutic option for patients approaching kidney failure. These reports lend understanding of who selects CKM and how those patients fare (Table 1). Table 1 - Comparison of features in kidney palliative care programs Feature Bursic et al. 19 So et al. 20 Source of patient referral No limitations Treating nephrologist, dialysis nurse Clinic staff Dual trained PC/nephrologists PC specialist, nephrology trainee, senior nurse, dietician, and social worker Proportion in KPC receiving CKM 67/16544 with eGFR<20 ml/min per 1.73 m2 144/604Mean eGFR of 13 ml/min per 1.73 m2 Topics/components of KPC program Decision-making supportSymptom assessment and managementReferrals to home healthCaregiver supportEncourage advance care planning and completion of physician orders for life-sustaining treatment Symptom assessment and managementMultidisciplinary support Bursic et al.19 described the delivery of active medical management without dialysis using an embedded KPC model. In this US-based clinic, two dual trained PC/nephrologists cared for 165 patients (84% CKD, 16% dialysis) over a 4-year period. They accepted referrals from any specialty or care setting with the most common reason for referral being decision-making support. The program was located within a traditional CKD clinic and included training for patient-facing staff (medical assistants and administrative assistants) on how to respond to patients requests for supportive care. This addresses the important role non-nephrologists can play in supporting patients and families. The nephrology clinic educational material was also modified to explicitly include information about supportive care and active medical management without dialysis. Patients in the CKM group were seen an average of four times in the study period. So et al.20 reported their KPC experience from St. George Hospital, Australia, caring for 604 patients over a 6-year period. Patients were referred by their nephrologist or dialysis nurse to a multidisciplinary team. They included 144 patients (23%) who had chosen CKM and excluded patients with CKD who were preparing for KRT but were not yet on dialysis in their analysis. This distinguishing factor precludes direct comparisons with the Bursic study, as the St. George patients had selected a pathway (CKM versus KRT) by time of referral, and the mean eGFR was approximately 10 ml/min per 1.73 m2 lower. The remaining patients were on hemodialysis (44%), peritoneal dialysis (9%), or had a kidney transplant (23%). Both programs exemplify that KPC is for both conservatively managed patients and those receiving KRT. Conservatively managed groups were older and had higher comorbidity, lower function, and lower quality-of-life scores at baseline compared with those who selected kidney replacement pathways. The focus of the St. George KPC clinic was symptom management and multidisciplinary support provided by a PC specialist, nursing, dietician, and social worker. The US-based KPC clinic not only included symptom assessment and management (opioid and analgesic prescription) but also explicitly reported in their study that they addressed decision-making support, referrals to home health, caregiver support, and advance care planning. Given the multidisciplinary nature of the St. George KPC clinic, these elements were also likely part of the program. Other components of a KPC clinic may complement the symptom management efforts by reducing caregiver burden and existential tension. In the study by Bursic et al., approximately half of the patients with CKD (48%) followed in the KPC program eventually selected active medical management as their disease progressed. Among this group two-thirds had an eGFR that fell below 20 ml/min per 1.73 m2. Of this group with declining kidney function, 57% died within the follow-up period with a mean time to death of 472 days (15.7 months). Advance care planning and designation of surrogate decision maker were common activities in the KPC, and 68% of all KPC patients who died within the study period received hospice services. This is three-fold higher than reported hospice utilization among dialysis patients.21 So et al. found that compared with patients receiving KRT, conservatively managed patients were more likely to report problems with mobility, self-care, and usual activities at baseline, but they were no more likely to report pain or anxiety/depression. At 12-month follow-up, the conservatively managed group had no significant decline in these domains. Although quality-of-life scores were lower for the conservatively managed group at baseline, these did not worsen at follow-up. Furthermore, when the analysis was restricted to the oldest group (age >75 years), there were no differences at baseline in quality of life, mobility, self-care, usual activities, or anxiety/depression. In fact, dialysis patients were more likely to report pain at baseline. At 12-month follow-up, there were also no differences in quality-of-life domains across the groups. Notably, the 80% of the conservatively managed group died or dropped out by 6-month follow-up and so only a fraction had data available at for evaluation. The median survival time of CKM patients was 14.1 (95% confidence interval, 9.5 to 18.6) months. It is likely that those who dropped out did not have stable trajectory of quality of life. As expected, patients without 12-month data in the KPC program were older and more commonly had congestive heart failure and lower function and quality of life at baseline. The analysis is limited in that individual trajectories were not analyzed. These studies build upon our prior knowledge of KPC programs with a conservative management pathway. Scherer et al.22 described their KPC program at New York University, caring for 55 patients (25% receiving CKM, 42% KRT, 20% undecided, and 7% planning for KRT). Patients who selected CKM were older and had lower function and lower eGFR compared with those receiving or planning for KRT. Activities in this KPC clinic were similar to the Bursic and So articles and included decision-making support, advance care planning, and symptom management, but quality of life and survival outcomes were not yet reported. Outcomes in patients managed on a conservative pathway have been previously reported on a smaller scale. The Lister Renal Unit in the United Kingdom previously reported their experience caring for 30 patients who selected CKM versus 124 patients who selected dialysis during the same period.23 CKM was offered within a low clearance clinic model separately from kidney supportive care. Patients completed quarterly quality-of-life assessments over approximately 15 months. Using a growth model analysis, they found no serial change in quality-of-life measures except life satisfaction, which decreased significantly after dialysis initiation and remained stable in CKM. The median survival in the CKM was closer to 30 months indicating that the population may not have been as sick as those in either the Bursic or So papers, despite a mean eGFR of 14 ml/min per 1.73 m2. In practice, patients diagnosed with advanced kidney disease often present to nephrology clinic looking for reassurance that they do not need dialysis. At the same time, they are fearful that death is the only alternative pathway to preparing for KRT. As nephrologists, we must carefully elicit our patients' goals and prognostication preferences before making a recommendation as to whether KRT or CKM is appropriate. These KPC programs provide us with further prognostic data that older, sicker, and more frail patients supported in a CKM pathway with symptom management and multidisciplinary support can maintain quality of life and of those followed may survive 14–16 months. Suffice to say, the outcomes of these programs are contingent upon PC expertise integrated into nephrology care. Future studies are needed to document unique trajectories of quality of life and physical and cognitive function among patients in CKM and KRT pathways that can inform decisions for patients with differing priorities. Identifying which components of KPC affect patient-centered outcomes is also critical for expanding the provision of KPC services. Finally, as KPC specialists are in short supply, alternative models of providing KPC should also be developed. In summary, these two articles offer data from their KPC programs that may guide our prognostic conversations with patients and highlight elements of their programs that might be replicated in other sites. Further study in larger cohorts, with consistent referral indications and longitudinal data collection would help to further expand our evidence base on CKM.