#6904 EVALUATION OF KIDNEY FUNCTION GRAPH SURVEILLANCE ACROSS SOUTH WEST WALES, UK

Abstract

Abstract Background and Aims For people requiring Kidney Replacement Therapy (KRT) late presentation to a nephrologist (within 90 days of initiating KRT) is associated with adverse outcomes. The ASSIST-CKD kidney function graph surveillance (KFGS) quality improvement project was adopted by the Department of Nephrology in South West (SW) Wales, UK in November 2017. Between the end of 2017 and the end of 2020, late presentation rates reduced from 19.5% [1] to 13.5% [2] across the region, having reduced to 9.7% in 2019 [3]. During KFGS, graphs of eGFR over time are generated for patients <65 years with eGFR ≤50ml/min/1.73 m2 and ≥65 years with eGFR ≤40ml/min/1.73 m2 with significantly deteriorating kidney function. ‘Alerts’ are sent to the GP for review and/or referral to nephrology. Over 100,000 graphs have been reviewed in SW Wales. We present demographic and initial outcome data for patients identified by KFGS. Method An observational retrospective cohort study analysed data of patients identified through KFGS in SW Wales between 1/11/17 and 1/10/21. Data were extracted from ASSIST-CKD and renal (Vital Data) databases using SQL coding. Results Data pertaining to 4811 patients are included for analysis. 44% of patients were male. At initial eGFR the mean age of patients ‘flagged’ was 73 years. The mean [range] initial eGFR was 32ml/min/1.73 m2 [2-50]. The median time [range] from initial eGFR to KFGS alert was 8 days [2-113]. For those with results within 90 days of initial eGFR the mean initial haemoglobin (Hb) was 118 g/L. The mean serum potassium was 4.7 mmol/L. Data for 1144 (24%) patients were included for outcome analysis, where a clear timeline from initial eGFR to KFGS alert, first registration with the nephrology service (registration), and outcome could be established. 3667 patients were excluded from outcome analysis because they were registered before KFGS or were not registered with nephrology, had died or moved out of area after KFGS alert. The median time [range] from alert to registration was 22 days [0-1636]. 20% of patients (n = 232) received CKD education from a specialist nurse. The median time [range] from registration to first contact from a CKD education nurse was 349 days [8-1419]. 34% of patients received IV iron and 31% received an erythropoiesis-stimulating agent (ESA) after alert. The median time [range] from registration to first IV iron and first ESA was 257 [5-1705] and 277 [3-1650] days respectively. 6% (n = 71) of patients required KRT. The late presentation rate was 22.5%. 21% of patients needing haemodialysis (HD) received their first treatment via an arteriovenous (AV) fistula, 1% via AV graft, 59% via non-tunnelled line and 19% via tunnelled line. 44% (7/16) of patients presenting late died within 90 days of starting KRT. 33% (18/55) of patients died within 90 days of starting KRT if they presented to a nephrologist ≥90 days before starting KRT. Conclusion Late presentation rates have reduced in SW Wales since implementing KFGS. The relatively high late presentation rate and low rate of initial definitive access for HD in the present data requires further analysis. eGFR values that prompted alert may denote Acute Kidney Injury (AKI) as opposed to, or as well as CKD progression; increasing the urgency of referral to nephrology. eGFR values pre and post alert will be reviewed to elucidate this. Patients presenting late were more likely to die within 90 days of starting KRT than those presenting early, emphasising the importance of early identification of progressive kidney disease. We aim to compare present data with a cohort referred to nephrology independent of KFGS to inform renal centres of the benefits of KFGS beyond headline late presentation rates, including timely access to specialists. We will consider potentially unwelcome consequences of KFGS, such as inappropriate referral. We cannot draw conclusions of causality between KFGS and outcomes. The GP may not have seen the alert, and there are several reasons why a GP may decide not to refer to nephrology following a KFGS alert.