J CHEN 1,2, M BROWN 3,4, M JOSE 5,6, F BRENNAN 3,4, D JOHNSON 7,8,9, M ROBERTS 10,11, G WONG 12,13,14, H CHEIKH HASSAN 1,2, A KENNARD 15,16, R WALKER 14,17, C DAVIES 18,19, N BOUDVILLE 20,21, M BORLACE 22, C HAWLEY 7,8,9, W LIM 20,21
1Wollongong Hospital, Wollongong, Australia, 2University of Wollongong, Wollongong, Australia, 3University of New South Wales, Sydney, Australia, 4St George Hospital, Sydney, Australia, 5University of Tasmania, Hobart, Australia, 6Royal Hobart Hospital, Hobart, Australia, 7Princess Alexandra Hospital, Brisbane, Australia, 8University of Queensland, Brisbane, Australia, 9Australasian Kidney Trials Network, Brisbane, Australia, 10Eastern Health, Melbourne, Australia, 11Monash University, Melbourne, Australia, 12The Children’s Hospital at Westmead, Sydney, Australia, 13Westmead Hospital, Sydney, Australia, 14University of Sydney, Sydney, Australia, 15Australian National University, Canberra, Australia, 16Canberra Hospital, Canberra, Australia, 17Eastern Institute of Technology, Napier, New Zealand, 18Australia and New Zealand Dialysis and Transplant Registry, Adelaide, Australia, 19University of Adelaide, Adelaide, Australia, 20University of Western Australia, Perth, Australia, 21Sir Charles Gairdner, Perth, Australia, 22Royal Adelaide Hospital, Adelaide, Australia
Aim: To assess the association between dialysis access and dialysis withdrawal in the first 12 months after dialysis initiation.
Background: Dialysis withdrawal is a major cause of early mortality in patients initiated on dialysis, and haemodialysis via central venous catheter (CVC) is known to be associated with early mortality. The association between dialysis modality and early mortality post-dialysis initiation is inconsistent.
Methods: Using data from the Australian and New Zealand Dialysis and Transplant Registry, we examined the association between dialysis access (categorised as peritoneal dialysis (PD), haemodialysis via arteriovenous fistula [AVF], and haemodialysis via CVC) and early death attributed to dialysis withdrawal (stratified by 0-6 months and >6-12 months) in incident adult dialysis patients in Australia between 2005 and 2018 using adjusted competing risk analyses.
Results: Of 31,579 incident dialysis patients, 1205 (4%) patients experienced death attributed to dialysis withdrawal within 12 months post-dialysis initiation, with 161 (13%) initiated on dialysis via PD, 223 (19%) on haemodialysis via AVF, and 821 (68%) on haemodialysis via CVC. Compared to patients on PD, the adjusted subdistribution hazard ratios (SHRs) for dialysis withdrawal between 0-6 months were 1.12 (95%CI 0.82-1.54) for haemodialysis via AVF, and 2.89 (2.20-3.78) for haemodialysis via CVC. The adjusted SHRs for dialysis withdrawal between >6-12 months were 0.92 (0.70-1.23) for haemodialysis via AVF, and 1.88 (1.48-2.40) for haemodialysis via CVC. No interaction was observed between dialysis access and late referral or body mass index.
Conclusions: Patients initiated on PD and haemodialysis via AVF had similar early dialysis withdrawal risk, whereas the risk significantly increased in patients initiated on haemodialysis via CVC with greater magnitude in the first six months post-dialysis initiation.
Jenny Chen is a consultant nephrologist at Wollongong Hospital, NSW. She completed her Master of Clinical Epidemiology through the University of Sydney in 2018 and spent a year in Vancouver, Canada, for home dialysis fellowship. She has a strong interest in kidney supportive care and home dialysis, especially peritoneal dialysis. She is an active member of the peritoneal dialysis working group of the ANZDATA Registry.