ASSESSING THE RISK AND SEVERITY OF HOSPITALISATIONS ACCORDING TO LEVEL OF KIDNEY FUNCTION: AN EXTEND45 ANALYSIS

M JUN1, L SUKKAR1,2, B NEUEN1, C FOOTE3,  K ROGERS1,  A KANG1, A SCARIA1, A CASS4,  C POLLOCK5,  D SULLIVAN6,  J KNIGHT1,  M GALLAGHER1, M JARDINE1
1The George Institute For Global Health, Newtown, Australia, 2University of Sydney, Camperdown, Australia, 3Concord Repatriation General Hospital, Concord, Australia, 4Menzies School of Health Research, Darwin, Australia, 5Kolling Institute for Medical Research, St Leonards, Australia, 6NHMRC Clinical Trials Centre, Camperdown, Australia

Aim: To determine the risk and severity of all-cause and cardiovascular hospitalisations by kidney function level in a community-based cohort study of NSW adults aged ≥45 years.
Background: Contemporary assessments of the relationship between chronic kidney disease (CKD) and health services use by level of kidney function have been limited. Prior studies typically used diagnosis codes to identify CKD which have relatively low sensitivity for ascertaining earlier stages of CKD.
Methods: Based on data from the EXTEND45 study (the 45 and Up Study linked to hospital and community pathology datasets by the Centre for Health Record and Linkage [CHeReL]), we identified a population-based cohort (2006-2014) of 41,099 people aged ≥45 who had a measure of kidney function (estimated glomerular filtration rate [eGFR]). The risk and severity of hospitalisations (defined by the hospitalisation length of stay [HLOS]) were assessed by eGFR category (≥90 [reference], 60-89, 45-59, 30-44 and <30ml/min/1.73m2) using multivariable Cox regression.
Results: Of 41,099 participants, 80.2% experienced ≥1 hospitalisation event over a mean follow-up of 5.8 years. All-cause hospitalisation risk increased as eGFR declined (HR 0.99[95% CI:0.97-1.03]; 1.05[1.00-1.09]; 1.17[1.10-1.24] and 1.57[1.43-1.72] for eGFR 60-89, 45-59, 30-44, and <30ml/min/1.73m2, respectively; p-trend<0.001) even after adjustment for age and sex. Cardiovascular events rose even more steeply than all-cause hospitalisations (HR 1.23[1.08-1.39]; 1.49[1.28-1.73]; 1.73[1.45-2.07] and 2.24[1.80-2.79], respectively). Mean HLOS also increased as eGFR declined (all-cause hospitalisation [2.3, 2.8, 3.9, 5.1 and 5.7days, respectively]; cardiovascular hospitalisations [3.7, 3.8, 5.7, 5.9 and 8.3days, respectively).
Conclusions: In contemporary NSW, declining kidney function is associated with higher illness burden, particularly cardiovascular burden. The increased mean HLOS suggests that discrete medical events are more severe as kidney function declines.


Biography:

Min Jun is a Senior Research Fellow at the George Institute for Global Health and Scientia Fellow at UNSW Sydney.He holds a PhD(2012) and MScMed(ClinEpi) in clinical epidemiology from the University of Sydney and has completed a 3.5-year international postdoctoral fellowship at the University of Calgary, Canada.Min’s research interests include the use of large trial and real-world, population-based data sources to better understand current and potential management strategies in kidney disease as well as exploring the associations between various risk factors and clinically important outcomes among individuals with chronic disease.

Categories