FP371RECURRENT HYPERKALAEMIA AND ASSOCIATION WITH LENGTH-OF-STAY AND MORTALITY FOLLOWING HOSPITALISATION: REAL-WORLD EVIDENCE FROM UK PATIENTS WITH CKD

Abstract

Introduction and Aims: Hyperkalemia is associated with adverse clinical outcomes in patients with chronic kidney disease (CKD). This real-world analysis investigated the incidence and recurrence of hyperkalemia in incident CKD stage 3+ patients in the UK, and the excess length-of-stay and mortality of these patients following admission to hospital. Methods: This retrospective observational cohort study utilised patient records from the Clinical Practice Research Datalink and Hospital Episodes Statistics (HES), from Jan 2006 to Dec 2015. Patients with EGFR Z60mL/min/1.73m2or a diagnosis/admission code for CKD stage 3+ during the study period were included. Patients with a history of CKD or heart failure prior to the study period were excluded, or if dialysis was their first renal event. A hyperkalemia episode was defined as a serum potassium measurement >5.0 mmol/L, without a similar measurement in the preceding seven days. Median time between successive hyperkalemia episodes was quantified among patients experiencing the given number of episodes (ignoring censoring). Hospital admissions in hyperkalaemic patients were matched to those in non-hyperka-laemic patients on age, gender and primary diagnosis. Differences in mean length-of-stay and odds ratios associated with in-hospital/30-day-post-discharge mortality between the two groups were estimated using Generalised Estimating Equations. Results: A total of 306,569 hyperkalemia episodes were observed in 191,964 patients, at a rate of 321.9 episodes per 1,000 patient-years (95% confidence interval: 320.7-323.0). Hyperkalemia was experienced by 48.4% of the cohort, with 28.8% experiencing multiple episodes. The probability of multiple episodes increased with renal function decline, and was greater for females than males, patients with diabetes than those without, and patients in receipt of RAASi than those not in receipt (all p<0.0001). The median time from initial CKD event to first hyperkalemia episode (among patients who experienced at least one episode) was 1.7 years, falling to 0.9 years from first to second episode, 0.6 years from second to third episode, and 0.5 years from third to fourth episode. HES records could be linked to 99,995 patients; mean length-of-stay among 55,439 hospital admissions of hyperkalaemic patients was 10.6 days, 3.5% longer than in the matched non-hyperkalaemic cohort (p<0.0001), and increas-ingto 12.9% (p=0.0003) for severe hyperkalemia (>6.0 mmol/L). The odds of in-hos-pital/post-discharge mortality was 20% greater for patients with hyperkalemia (p<0.0001). Conclusions: Recurrent hyperkalemia was common among incident CKD stage 3+ patients in the UK. Hyperkalemia was more frequent after each successive episode, with increasingly shorter intervals between episodes. Hospitalised patients with hyperkalemia tended to remain in hospital longer than those without hyperkalemia, therefore placing a greater burden on the healthcare system, and were more likely to die during or shortly after the admission.