Dr Melissa Levy1, Dr James Sindone4, Bree Shi2,3, A/Professor Mark Dennis4, Professor Raj Puranik4, Professor Steve Chadban1,2,3, Dr Tracey Ying1,2,3
1Kidney Centre, Royal Prince Alfred Hospital, Camperdown, Australia, 2Kidney Node, Charles Perkins Centre, The University of Sydney, Camperdown, Australia, 3Australia and New Zealand Dialysis and Transplant (ANZDATA) Registry, Adelaide, Australia, 4Department of Cardiology, Royal Prince Alfred Hospital, Camperdown, Australia
Biography:
Melissa Levy is a final-year Renal Advanced Trainee at Royal Prince Alfred Hospital, with a strong interest in clinical epidemiology and the intersection between chronic kidney disease and metabolic health. She is passionate about improving cardio-metabolic outcomes in patients at all stages of kidney disease through the provision of patient-centred and evidence-based care.
Aim:
To determine the accuracy and completeness of cause of death (COD) recorded by ANZDATA, by assessing concordance with blinded clinicians.
Background:
Registry data informs clinical care, health policy, and research, however attribution of death to a single cause in kidney transplant recipients (KTR) and lack of standardisation limits data quality.
Methods:
KTR transplanted and managed by Sydney Local Health District clinicians until their death were studied. Pre-death medical data was extracted from clinical records and provided to a consultant panel. Primary and secondary COD were determined by 3 nephrologists and 2 cardiologists independently. Cohens Kappa statistic was calculated to assess concordance with the COD recorded on ANZDATA.
Results:
Between 2010–2023, 199 patients met inclusion criteria. Mean age at death was 63 years, 66.1% were male, and 80.3% were deceased donor recipients. 127 (64%) had clinical information relating to their death. Agreement for the primary COD was moderate (61.4%, kappa 0.50) and was highest for cancer (63.9%). The greatest discrepancy was ‘withdrawal’, which accounted for 12.6% of deaths in ANZDATA and 3.1% on panel assessment. Cardiovascular, infection, and cancer comprised 33.9%, 32.3% and 26.0% of clinician-adjudicated primary COD respectively, compared to 26.0%, 24.4% and 20.5% on ANZDATA. The most common secondary COD was graft dysfunction, present in 56.7% of deaths, followed by infection, which was present in 48.5% of cardiovascular and 48.8% of cancer deaths.
Conclusions:
Death in a KTR is often complex with multiple contributing causes. Given only moderate agreement between ANZDATA recorded and clinician-adjudicated primary COD, recording secondary COD may improve the accuracy and quality of data in the ANZDATA registry. Graft dysfunction and infection are underappreciated as contributors to death amongst KTR.
Presentation Slides PDF – Click Here