SP646FACTORS CONTRIBUTING TO ACUTE RENAL REPLACEMENT THERAPY INITIATION IN PATIENTS WITH END-STAGE RENAL DISEASE - OUTCOMES FROM A LARGE RETROSPECTIVE COHORT STUDY

INTRODUCTION AND AIMS: Acute initiation of chronic renal replacement therapy (RRT) is well known to have a higher mortality compared to a planned initiation. We sought to (1) identify the reasons and risk factors for acute dialysis initiation and (2) compare the mortality risk between the different groups of acute presenters. METHODS: Single-centre retrospective data collection from patients with end-stage renal disease commencing RRT between Jan 2013-Dec 2015. Patients were categorised to those starting dialysis electively (planned), patients known to nephrology services for > 90 days requiring acute dialysis commencement (unplanned) and those known for < 90 days needing (urgent) dialysis initiation. RESULTS: 825 patients commenced dialysis between Jan 2013-Dec 2015. The number of patients per group and reasons for acute RRT initiation are outlined in Figure 1a. Patient-related factors included non-adherence, prior decision for conservative management or refusal of dialysis (n=36, 4.4%). 55.3% of unplanned starts required acute RRT initiation prompted by systemic illness due to sepsis (48.1%) and cardiovascular disease (22.2%). Glomerulonephritis accounted for most urgent starts from acute illness (35.8%). Higher rates of eGFR decline per month 1 year preceding acute dialysis initiation (Figure 1b) were observed amongst groups of acute starters when compared to a planned start (Unplanned:with systemic illness, 1.2ml/min/1.73m2, p=0.000; with an unexpected decline, 1.4ml/min/1.73m2, p=0.001; Urgent:with systemic illness, 4.1ml/min/1.73m2, p=0.000; with an unexpected decline, 3.0ml/min/1.73m2, p=0.012; Planned: 0.7ml/min/1.73m2). The risks of acute RRT initiation from systemic illness or an accelerated decline in renal function were significantly higher in patients known to nephrology services for > 3 months who were older (> 75 years) with co-morbid diabetes and hypertension (Figure 2a). Systemic illness at presentation in unplanned (OR 1.96 (1.27-3.03), p=0.002) and urgent (OR 2.36 (1.57-4.12), p<0.001) starters conferred an increased risk of mortality on multivariate analysis (Figure 2b). CONCLUSIONS: A higher risk of death observed with an unplanned dialysis start is associated with systemic illness at presentation. A similar survival at 1 year following acute RRT initiation is observed between planned and urgent groups but does not account for the morbidity acquired from an acute start or quality of life. Our data identifies an increased likelihood of unplanned dialysis initiation in older patients with comorbid diabetes and hypertension known to nephrology services for > 3 months. Higher rates of decline in eGFR observed 1 year preceding an acute dialysis start can be used to risk stratify those patients in whom this may be avoided through increased surveillance.