Dialysis
Written 29th August 2020 - Last Update 30th August 2026
Dialysis
CKD stage 5 needs some form of kidney replacement therapy, or comprehensive conservative care. Options should be discussed with the patient and family, since all require significant ongoing commitment.
KDIGO’s practice points are to inform people with CKD about the options for kidney replacement therapy and comprehensive conservative care, and to support comprehensive conservative management as an option for those who choose not to pursue it.
When to start
Start maintenance dialysis on the presence of signs or symptoms of kidney failure, not on a number. Education should extend to family members and carers.
Options
Comprehensive conservative kidney care — no dialysis or transplant; managed in the community with palliative or renal supportive care. It is planned, holistic, person-centred care built on detailed communication including prognosis and advance care planning, shared decision-making, active symptom assessment and management, psychological, social, family, cultural and spiritual support, and interventions to delay progression, but not dialysis. In most people life expectancy is lower than with dialysis or transplant — but in elderly patients with two or more comorbidities, dialysis may not offer a survival or quality-of-life advantage, and symptom management alone can improve quality of life.
Peritoneal dialysis (home) — continuous ambulatory peritoneal dialysis (CAPD), with four or more manual daytime bag exchanges. Needs a PD catheter; simple, gentle and portable; about a week of training in Australian practice (StatPearls describes 2-3 weeks of technique training at the unit); infection risk around one episode per two years; allows freedom to work and travel with generally good quality of life. PD is contraindicated with an uncorrected abdominal wall hernia, a pleuroperitoneal shunt, or abdominal adhesions.
Home haemodialysis — daytime (3-5 sessions per week, 4-6 hours each) or overnight (3-5 nights per week, 8 hours each). Needs a fistula created at least 3 months before use, and about 3 months of training on average. Flexible daily routine, more dialysis hours than centre-based haemodialysis, and better health outcomes. The home must be structurally sound and free of damp, mould and excessive pollution before a program is set up.
Centre-based haemodialysis — hospital or satellite centre, three times a week for 4-6 hours, individualised, with occasional overnight-session clinics. No training required, but a strict routine and diet plus transport to the centre are needed. Higher infection risk and poorer outcomes than home-based dialysis.
Transplant — living or deceased donor. Needs surgery and lifetime immunosuppression; average wait for a deceased donor is 3-7 years, shorter with a compatible living donor. Once function stabilises, patients regain freedom to work and travel with only dietary moderation, and good survival — but higher infection and cancer risk from long-term immunosuppression.
Access matters more than the modality label
An absolute contraindication to haemodialysis is the inability to secure vascular access. The “Fistula First” initiative encourages creation of an arteriovenous fistula in most patients. Starting haemodialysis through a tunnelled central venous catheter carries a substantially higher early mortality — one Canadian cohort found an 80% increased risk of death in patients starting HD with a central venous catheter, but no difference in mortality between PD and HD using a fistula or graft.
Large observational studies show a survival advantage for PD over HD in the first 1 to 3 years of dialysis.
Mortality among dialysis patients is predominantly cardiovascular.
Shared decision-making
Prognosis is the key information for shared decision-making in CKD G5, and it requires unbiased information on survival and on the outcomes people with CKD actually care about: quality of life, symptom burden, and support from family and clinicians. It should be done in a culturally appropriate way with attention to health literacy.
It is not a one-off conversation but an ongoing process guiding decisions over time. Various decision-support tools exist.
Five questions to guide it:
- What will happen if we watch and wait?
- What are your test or treatment options?
- What are the benefits and harms of these options?
- How do the benefits and harms weigh up for you?
- Do you have enough information to make a choice?
Advance care planning sits under the conservative-care umbrella and clarifies preferences for end-of-life care. Functional and cognitive decline as CKD progresses makes those conversations harder later, so start them early.
Depression in CKD/dialysis
Depression is common, under-recognised and under-treated in CKD and ESRD — around 20% of CKD patients experience a major depressive episode, a higher prevalence than in diabetes or heart failure. It is independently associated with increased morbidity and mortality: dialysis patients with depression are roughly twice as likely to die or be hospitalised within a year, and have about a 30% increase in cumulative hospital days. It does not reliably remit spontaneously.
Treatment considerations:
- Evidence is limited (small trials, few placebo-controlled), but SSRIs appear the safest antidepressant class in advanced CKD/ESRD
- Sertraline generally needs no dose adjustment, though its active metabolite is renally excreted — a reasonable first-line choice
- Citalopram needs no routine dose adjustment but use cautiously given its long half-life; escitalopram needs a lower initial dose and is not recommended below eGFR 20 mL/min
- Paroxetine needs dose reduction if CrCl under 30 mL/min
- TCAs are generally avoided given QTc prolongation, arrhythmia and orthostatic hypotension risk
- Venlafaxine needs a 25-50% dose reduction in mild-moderate renal impairment
- Non-drug options with the most promising evidence: more frequent haemodialysis, and CBT
Given the already elevated cardiovascular mortality in CKD, treating depression here matters beyond mental health alone — though whether treatment reduces morbidity and mortality, rather than just symptoms, is not well established.