Abstract
Objective To describe the provision of primary and specialist palliative care for older adults with advanced chronic kidney disease (CKD).
Design Population-based retrospective descriptive cohort study using electronic health records.
Setting St Michael’s Hospital Academic Family Health Team, including 5 clinics and an interdisciplinary home-based outreach team, in Toronto, Ont.
Participants Older adults who are 65 years of age or older with advanced CKD and undergoing community-based nondialysis kidney care between April 1, 2012, and April 1, 2022, with at least 1 year of follow-up.
Main outcome measures Prevalence of advance care planning, goals of care (GOC) discussions, access to specialized palliative care, frequency of hospitalizations, places of deaths, and mean survival time.
Results The study included 47 older adults with advanced CKD who were not undergoing dialysis and receiving primary care. Sixty-eight percent were female (n=32), and the mean (SD) age was 81 (8) years. The mean (SD) number of comorbidities was 5 (2). Almost half of the patients (n=22) lived alone. Seventy percent of the cohort (n=33) had a documented substitute decision maker. Approximately one-third of the cohort (n=16) had GOC conversations involving prognosis, hospital transfer, place of death, and treatment goals. Forty-seven percent (n=22) accessed specialist palliative care during follow-up. Seventy-four percent (n=35) had 1 or more hospitalization. Thirty-two percent (n=8) died at home. The mean (SD) survival time was 2 (2) years.
Conclusion A palliative approach to care including advance care planning and GOC conversations would be appropriate for older adults with advanced CKD given their high rates of comorbidities, hospitalizations, and mortality. Primary care providers might experience challenges facilitating conversations in this population, highlighting opportunities for enhanced training and point-of-care interventions, and facilitating access to specialist palliative care consultations when appropriate.
With the aging of our population and the growing demand for palliative care among those with nonmalignant diseases, there is increasing importance placed on chronic kidney disease (CKD) due to its prevalence and symptom burden. Chronic kidney disease, as defined by an estimated glomerular filtration rate (eGFR) less than or equal to 60 mL/min/1.73m2, or markers of kidney damage for 3 months or more,1 affects approximately 13% of adults and more than one-third of those aged 65 years or older in North America.2,3 It is associated with heightened morbidity, mortality, and substantial health care costs, especially in those with advanced stages of CKD.4-6
Primary care providers play a crucial role in the comprehensive care of individuals with CKD. This role involves assessing disease severity, preventing disease progression and associated complications, managing symptoms, and supporting future planning and preferences of care.7,8 However, in a sample of older adults across Canada, only 50% named their substitute decision maker for medical treatment decisions. Further, in the same sample of older adults who talked with someone about their medical treatment preferences at end of life, only 18% discussed it with their family physician.9 The extent to which individuals with advanced CKD have these conversations in the Canadian primary care setting remains to be elucidated.
In this study, we examined the prevalence of advance care planning (ACP), goals of care (GOC) discussions, and access to specialist palliative care in a population of older adults with advanced CKD who were not undergoing dialysis in a major urban primary health care centre. Advance care planning involves a capable person identifying 1 or more future substitute decision makers. It also involves the person sharing their wishes, values, and beliefs with the substitute decision maker or makers about their future care to prepare them to make health care decisions on their behalf should they become incapable of making their own decisions.10 Conversations about GOC might be influenced by ACP. These conversations between a provider and a capable person or, if the person is incapable, the substitute decision maker, focus on ensuring understanding of the serious nature of the patient’s illness and exploring the patient’s values to inform treatment decision making in the current care context.10,11 We also investigated the frequency of hospitalizations, places of deaths, and life expectancy. We aimed to provide insights that would guide future interventions to support primary palliative care for patients living with advanced CKD.
METHODS
Study population and cohort formation
We conducted a retrospective cohort study using electronic medical health records from the St Michael’s Hospital Academic Family Health Team (SMHAFHT) in Toronto, Ont. We assembled a cohort of patients aged 65 years or older with advanced CKD between April 1, 2012, and April 1, 2022. The study end date was April 1, 2023, to allow for at least 1 year of follow-up. The SMHAFHT is an urban interprofessional primary care team affiliated with the Department of Family and Community Medicine at the University of Toronto. It includes 5 clinics and an interdisciplinary home-based outreach team that serves homebound frail older adults. Medical records from all clinic sites and the home-based team were included in the study.
We defined advanced CKD by a series of at least 2 consecutive measurements of an eGFR less than or equal to 20 mL/min/1.73m2 (calculated by the CKD Epidemiology Collaboration equation12) over a period of at least 90 days, not on dialysis.13 The research team established an eGFR less than or equal to 20 mL/min/1.73m2 to be a reasonable threshold at which to initiate ACP and GOC discussions, as dialysis initiation commonly occurs at an eGFR less than or equal to 10 mL/min/1.73m2.14
Exclusion criteria
We excluded patients undergoing chronic dialysis, as they represent a distinct population more likely to receive multidisciplinary care from their nephrology team, potentially involving access to palliative care. We excluded individuals with private or inactive charts, defined as patients who have not been at the family health clinic in 3 years or are lost to follow-up. Patients who had only a single eGFR less than or equal to 20 mL/min/1.72m2 and individuals with recovered kidney function eGFR greater than 20 mL/min/1.72m2 within the study period were excluded.
Baseline characteristics
Age was calculated from the date of birth on the index date, which is derived from the initial eGFR measurement that met the study criteria. We collected data on notable comorbidities15 and living situations reported on the patient profile, at specialist consultations, and from hospital discharge summaries. We documented whether patients lived alone as a proxy for social isolation.
Discussions about ACP and GOC
As a proxy for ACP, we counted substitute decision makers that had their name, contact information, and relationship clearly identified on the medical record. For GOC conversations, we counted discussions on prognosis, preferences for hospital transfer, place of death, dialysis compared with nondialysis care, code status, and other treatment preferences. We also counted discussions about palliative care and types of referrals to specialist palliative care, such as consultation, shared care, or takeover models of specialist palliative care involvement.16
Hospital admissions and locations of deaths
We examined hospital discharge summaries to determine admission dates and places of deaths.
Data analysis
We summarized the data as percentages. We used Microsoft Excel software to calculate percentages and means and SDs. For confidentiality purposes, we did not report small cell sizes (n≤6). The Unity Health Toronto Research Ethics Board granted approval and waiver of patient consent.
RESULTS
The SMHAFHT cared for almost 70,000 patients over a 10-year period from April 1, 2012, to April 1, 2022. About 18% were older adults. Seventy-three patients had advanced CKD and did not undergo dialysis. Twenty-four patients were subsequently treated with chronic dialysis and 2 patients had inactive charts. We included 47 older adults who met study inclusion criteria in terms of advanced CKD not treated with dialysis (Figure 1).
Cohort formation flow chart
As shown in Table 1, 68% of patients were female and 32% were male. The mean (SD) age was 81 (8) years. The mean (SD) eGFR at index was 16.8 (2.8) mL/min/1.73m2. The index eGFR rate for most patients (over 90%) was between 15 mL/min/1.73m2 and 20 mL/min/1.73m2. The mean (SD) number of comorbidities was 5 (2). Most patients in the study (>85%) had hypertension. Seventy percent had diabetes, 60% had heart failure, and 30% had dementia. Almost half the patients (47%, n=22) lived alone. The SMHAFHT home visiting team cared for approximately 15% of the cohort.
Baseline characteristics of older adults with advanced CKD: N=47.
Seventy percent of the cohort had a documented substitute decision maker. Approximately one-third of the cohort (n=16) had a GOC conversation with their primary care provider regarding prognosis, preference for transfer to hospital, place of death, and treatment goals. Similar proportions of the cohort discussed their code status (n=18) and thoughts regarding dialysis versus nondialysis care (n=15). Thirty-four percent of discussions involved whether to refer the patient to specialist palliative care. The most responsible family physician facilitated almost all GOC conversations. A few cases were led by a family medicine resident or nurse practitioner. Most of the cohort were not seen by specialist palliative care (n=25). Among patients seen by specialist palliative care (n=22), 41% were seen more than 3 months before death and 59% 3 or fewer months before death (Table 2). Sixty-four percent of patients seen by a specialist palliative care provider received a consultation with suggestions to the primary care provider, and the remainder of the cohort received either a shared care or a takeover-type model of specialist palliative care (eg, in a palliative care unit).
Advance care planning, goals of care discussions, and access to specialist palliative care among older adults with advanced CKD: N=47.
Approximately 75% of the patients had at least 1 hospitalization (n=35); 69% of these patients had 2 or more hospitalizations; and 23% had 5 or more hospitalizations. Approximately half of the patients died during the study period (n=25); the mean (SD) days from index to death date was 2 (2) years (range 5 months to 6 years). Almost 25% of the cohort died in an acute care setting, 32% died at home, and 44% died in a palliative care unit. Eight patients were lost to follow-up (Table 3).
Hospital admissions and dispositions among older adults with advanced CKD: N=47.
DISCUSSION
The study found that most older adults living with advanced CKD not treated with dialysis and receiving primary care were female, aged 75 years or older, had more than 5 comorbidities, and lived alone. Although most patients had a documented substitute decision maker as part of ACP, approximately one-third did not, and most did not have a GOC conversation. Almost half of the cohort accessed specialist palliative care consultation, and most of this was for end-of-life care. This cohort had multiple hospitalizations and a life expectancy of up to 2 years. While most patients died either at home or in a palliative care unit, a substantial proportion died in an acute care setting.
We highlight a relatively low prevalence of GOC conversations despite high levels of morbidity and mortality found in this population of older adults with advanced CKD. There is growing recognition of the importance of a palliative approach in primary care. Primary care providers can be better supported in delivering primary palliative care for patients with multimorbidity at risk of unwanted hospitalizations.17 There are clinical resources developed in Canada for family physicians that are focused on building professional capacity to improve the quality of primary palliative care in Canada, as well as guidelines and decision aids on managing advanced CKD.18,19 Electronic medical records might have the potential to provide enhanced access to ACP and GOC documents and triggers for timely identification of patients requiring a palliative care approach.20,21 Robust infrastructure to support regular follow-up with family physicians, provision of after-hours care, and coordination with home care and home-visiting primary care providers would enhance primary palliative care delivery.22 In Canada, major barriers to facilitating ACP and GOC discussions in primary care include uncertainties surrounding prognosis, difficulties supporting patients and their families in accepting the severity of their kidney disease, and the need to clarify the expectations of nondialysis compared with dialysis modalities of care.23 Although maintenance dialysis might prolong survival over a short span of years among older adults with kidney failure, dialysis also increases the risk of spending more time in the hospital.24 Common reasons for not initiating dialysis include patient preferences for avoiding commutes to the hospital, the feeling of being a burden, or the risks of dialysis-related complications.25 Prediction tools, such as the Kidney Failure Risk Equation, can be used to estimate the likelihood of disease progression and support timing of referral to nephrology.26 Specialist palliative care consultation teams can provide an additional layer of support in future planning and in addressing physical and psychosocial needs.27
In this population of older adults, a large proportion were living alone. Living alone with serious illness can make it difficult to die at home and increases the risk of psychosocial distress and hospitalizations.28 With the growing proportion of older adults living alone, there will likely be an increased need for home-based supports to provide an equity-informed palliative approach to care.29
Limitations
There are some limitations to consider when interpreting the results of this study. We based this study on a large, well-resourced, multisite urban interdisciplinary family health team, which is likely to limit the generalizability of the findings. Regarding the ACP and GOC discussions, it is possible that some of the conversations took place and were not clearly recorded in the chart, though guidelines strongly advise their documentation. Continuity of care has been shown to enable person-centred care and lead to greater confidence in medical decision making.30 Future work might examine continuity and access to primary care to further elucidate potential barriers to or facilitators of primary care provider–led GOC discussions with patients in advanced stages of CKD. This study focused on patients being treated conservatively without dialysis. A palliative care approach is also relevant for patients being treated with kidney replacement therapy, such as dialysis or kidney transplantation, which warrants further study in primary care.31
Conclusion
There are opportunities to enhance guideline-concordant documentation of ACP and GOC discussions among patients with advanced CKD in the primary care setting and for more timely access to specialized palliative care consultation when appropriate.
Footnotes
Contributors
All authors contributed to conceptualizing and designing the study; to collecting, analyzing, and interpreting the data; and to preparing the manuscript for submission.
Competing interests
None declared
This article has been peer reviewed.
Cet article a fait l’objet d’une révision par des pairs.
- Copyright © 2025 the College of Family Physicians of Canada
