P1852REGISTRATION WITH PATIENT-FACING ONLINE HEALTH RECORDS IN CHRONIC KIDNEY DISEASE IS ASSOCIATED WITH A LOWER INCIDENCE OF HYPERKALAEMIA

Abstract Background and Aims PatientView (https://patientview.org) is a web-based health record available throughout the United Kingdom which allows patients to review blood results and clinical communication from renal services. PatientView registration may differ across demographic groups and may be associated with improved clinical end points. We investigated the demographic factors associated with PatientView registration, and the association with incidence of severe hyperkalaemia. Method A retrospective analysis was performed of a prospectively-acquired database of adults with chronic kidney disease (CKD) stages 3 – 5 in a single Scottish health board attending 2006–2017. Individual start date was point of first clinic attendance; end date was date of death, commencement of renal replacement therapy or point of data extraction. Covariates included age, sex, biochemistry, socioeconomic status and co-morbidity. The Scottish Index of Multiple Deprivation (SIMD) was used as a measure of socioeconomic status, where quintile 1 is most deprived. A survival analysis of time to severe hyperkalaemia (>6.5mmol/L) was performed with PatientView registration as a time dependent covariate. Results Of 9935 patients, 967 (9.7%) were registered users of PatientView. The PatientView group had lower age in years (54.7 [42.7-65.6] vs 71.5 [62.1-78.4], p<0.001), estimated glomerular filtration rate (eGFR) (27.1 [18.6-38.8] vs 38.5 [28.3-51.8] ml/min/1.73m2, p<0.001), and higher SIMD (indicating lower levels of deprivation) (23 vs 13% in least deprived quintile, 25 vs 37% in most deprived quintile, p<0.001); fewer patients were female in the PatientView group (46 vs 51%, p<0.001). PatientView users had lower body mass index (29.1±7.6 vs 29.8±7.3 kg/m2, p<0.001), and lower prevalence of diabetes (28 vs 33%, p<0.001), heart failure (5 vs 7%, p<0.001), myocardial infarction (13 vs 22%, p<0.001) and stroke (3 vs 6%, p<0.001). On Cox regression, PatientView was associated with reduced risk of hyperkalaemia (HR 0.43, 95% CI 0.32,0.58) after adjustment for baseline eGFR, baseline serum potassium, SIMD, sex, diabetes and heart failure. Conclusion PatientView use is more prevalent amongst male, middle-aged patients with higher socioeconomic status. Registration with PatientView is associated with reduced likelihood of adverse clinical events including hyperkalaemia. Limitations include lack of ethnicity data, retrospective design and drug therapy. PatientView usage may associate with concordance to a low potassium diet, reflecting general health literacy. However, it is plausible that a causal relationship exists between Patientview usage and hyperkalaemia, by facilitating feedback on, and engagement with, lifestyle changes encouraged by healthcare providers. Whilst there remains a lack of PatientView registration in certain demographic groups, these should be targeted via quality improvement initiatives to enhance renal outcomes in underserved groups.