J CHEN1,2, M BROWN3,4, M JOSE5,6, F BRENNAN3,4, D JOHNSON7,8,9,10, M ROBERTS11,12,  G WONG13,14,15, H CHIEKH HASSAN1,2, A KENNARD16, R WALKER13,17, N BOUDVILLE18,19, M BORLACE20, C HAWLEY7,8,9,10, W LIM18,19

1Wollongong Hospital, Wollongong, Australia, 2University of Wollongong, Wollongong, Australia, 3St George Hospital, Sydney, Australia, 4University of New South Wales, Sydney, Australia, 5University of Tasmania, Hobart, Australia, 6Royal Hobart Hospital, Hobart, Australia, 7Princess Alexandra Hospital, Brisbane, Australia, 8University of Queensland, Brisbane, Australia, 9Translational Research Institute, Brisbane, Australia, 10Australasian Kidney Trials Network, Brisbane, Australia, 11Monash University, Melbourne, Australia, 12Eastern Health, Melbourne, Australia, 13University of Sydney, Sydney, Australia, 14Centre for Kidney Research at The Children’s Hospital at Westmead, Sydney, Australia, 15Centre for Transplant and Renal Research at Westmead Hospital, Sydney, Australia, 16Canberra Health Services, Canberra, Australia, 17Eastern Institute of Technology, Napier, New Zealand, 18Sir Charles Gairdner Hospital, Perth, Australia, 19University of Western Australia, Perth, Australia, 20Royal Adelaide Hospital, Adelaide, Australia

Aim: To examine the temporal trends and risk factors of cause-specific withdrawal mortality within 12 months of dialysis initiation.
Background: Mortality risk is high soon after dialysis initiation in patients with kidney failure, and withdrawal from dialysis is a major cause of early mortality. The temporal evolution and risk factors of cause-specific early withdrawal mortality remain uncertain.
Methods: Data from the Australian and New Zealand Dialysis and Transplant Registry were used to examine the temporal trends and risk factors associated with early withdrawal mortality attributed to psychosocial and non-psychosocial reasons in incident adult dialysis patients in Australia between 1998 and 2018 using adjusted competing risk analyses.
Results: Of 42,409 incident dialysis patients, 3639 (9%) experienced early mortality within 12 months after dialysis initiation. The proportion of early mortality attributed to dialysis withdrawal increased from 30% between 1998-2004 to 37% between 2012-2018, predominantly related to the increase in withdrawal for non-psychosocial reasons from 17% to 23%. For early psychosocial dialysis withdrawal, the covariates associated with subdistribution hazard ratios (SHR)>1.5 were older age, underweight, late nephrologist referral, presence of peripheral vascular disease, and state at dialysis initiation. The covariates associated with SHR<0.5 were Asian ethnicity, obesity, peritoneal dialysis as initial dialysis modality, and kidney failure attributed to polycystic kidney disease. For early non-psychosocial withdrawal, recent era was the only additional risk factor with SHR>1.5.
Conclusions: Early withdrawal mortality accounted for over 30% of early deaths with a temporal rise, predominantly due to non-psychosocial dialysis withdrawal. Several modifiable and non-modifiable risk factors were associated with early withdrawal mortality due to psychosocial and non-psychosocial reasons. Comprehensive pre-dialysis education for the high-risk patients may help better inform shared decision-making.

Dr Jenny Chen is an early-career nephrologist with a strong interest in home dialysis and kidney supportive care. She spent a year in Vancouver, Canada, for her home therapies fellowship in 2018 and completed Masters of Clinical Epidemiology in the same year. Her current clinical work includes expansion of home dialysis and kidney supportive care services in South Coast, NSW.

Recent Comments