Abstract: PO1375
Predictors of Treatment Discussions in Geriatric Dialysis Patients Who Died
Session Information
- Geriatric Nephrology: New Insights
November 04, 2021 | Location: On-Demand, Virtual Only
Abstract Time: 10:00 AM - 12:00 PM
Category: Geriatric Nephrology
- 1100 Geriatric Nephrology
Authors
- Bywater, Laura, Eastern Health, Box Hill, Victoria, Australia
- Douglass-Molloy, Hannah, Eastern Health, Box Hill, Victoria, Australia
- Roberts, Matthew A., Eastern Health, Box Hill, Victoria, Australia
- Mcmahon, Lawrence, Eastern Health, Box Hill, Victoria, Australia
- Ducharlet, Kathryn, Eastern Health, Box Hill, Victoria, Australia
Background
Geriatric patients on maintenance dialysis often have a high burden of symptoms and comorbidities with limited prognosis. Therefore, in the setting of declining clinical status, goals of treatment (GOT) discussions are important to inform illness expectations and balance the benefits and burdens of ongoing treatment. This study aimed to identify mortality risk factors that prompted nephrologists to have GOT discussions including dialysis withdrawal.
Methods
A cohort of 95 adult patients ≥65 years, cared for by Eastern Health (Victoria, Australia), who died between 1/1/2016-31/12/2019 was analysed using Fischer’s exact tests to identify psychosocial variables associated with functional decline and mortality.
Results
Mean age was 78 (SD 7.3), with mean dialysis vintage 5 years and average 3.7 admissions in the 12 months preceding death. Mean Charleston comorbidity index (CCI) was 10 (SD 2.3), with hypertension 78%, T2DM (56%), IHD (56%), cardiac failure (51%), malignancy (34%), peripheral vascular disease (32%) and cognitive impairment (28%) major comorbidities. Almost one third (29%) of patients lived in a nursing facility, 74% used gait aids, and 55% needed assisted transport to dialysis. Median time between needing transport and death was 1.3 years [IQR 0.5, 2.2].
A GOT discussion in the preceding year was documented in 65% of patients, in renal clinics (37%) or on dialysis (12%). Median time between GOT discussion and death was 5 months [IQR 1, 11]. Discussions were more likely if patients had high comorbidity (CCI>5, 57% vs CCI<5, 8%, p<0.05) with no association between GOT discussions and patient age, dialysis vintage, functional status, or specific comorbidities. Advance care plans were completed in 26% and were more likely if a GOT discussion had already transpired (23%v4%, p<0.05), or the patient lived in a nursing facility (16 v12%, p<0.05).
Deaths occurred in hospital (50%), hospice (19%), or at home (7%). Dialysis was withdrawn median 8 days before death [IQR 6, 11].
Conclusion
Older patients died with significant comorbidities and functional dependency, though only the former prompted GOT discussions. Despite a long-term relationship, nephrologists could improve documentation of future treatment planning with patients to promote patient centered end of life experiences.