FP410MULTI-CENTRE ANALYSIS OF FRACTURE RISK IN RENAL REPLACEMENT THERAPY PATIENTS

Abstract

Introduction and Aims: Introduction Bone fractures are an important cause of morbidity and mortality in patients on renal replacement therapy (RRT). Aim The aim of this multicentre observational study was to quantify the incidence of radiologically proven bone fracture by site, in prevalent RRT groups and study its relationship to associated risk factors. Methods: A retrospective analysis of patient electronic records was performed from 7 July 2010 to 4 September 2013. All radiology reports (X-ray, computerised tomography (CT), magnetic resonance imaging (MRI), Ultrasound (USG), nuclear medicine scans, angiographic reports and others) from all hospitals acrossWest of Scotland, attached to patients records were obtained. These were analysed by one individual to identify all fracture events. If multiple anatomically distant sites were involved in one event they were recorded as separate fracture. Classification of anatomical fractures sites was pre-defined. Covariates identified as potential risk factors for fractures were age, gender, RRT modality, primary renal diagnosis (PRD), duration of RRT, biochemical parameters (haemoglobin, serum albumin, phosphate, adjusted calcium, alkaline phosphatase, parathyroid hormone (PTH)) and medications affecting bone metabolism (prednisolone, cinacalcet, phosphate binders and Vitamin D analogues). For calculation of biochemical parameters mean of last three results prior to inception was used with pre-dialysis values used for patients on haemodialysis. RRT duration was defined as date from first RRT until date of inception. PRD was grouped as familial, primary glomerulopathy, tubulo-interstitial, diabetes, systemic, other. Univariate and multivariate Cox proportional hazard models were created to estimate the risk of all fractures. Results: We identified 2096 patients on RRT at the start of the study. 907 were on haemodialysis (HD), 108 on peritoneal dialysis (PD) and 1081 had a functioning transplant. There were 340 fractures in the three-year study period with an overall incidence of 62.8 per 1000 patient years. The incidences were 37.6, 99.2, and 57.6 per 1000 patient years in the transplant, HD and PD groups respectively (p <0.05). Radial, foot and hip fractures were the 3 commonest sites (n=53, 47 and 46 respectively). In the multivariate model, age (HR 1.02, p <0.05) and HD (HR 5.25, p<0.05) were independently associated with increased risk of fractures and systemic group of PRD (HR 0.42, p<0.05), high serum albumin (HR 0.96, p=0.06) and being on alfacalcidol (HR =0.51, p<0.05) and lanthanum (HR 0.41, p<0.05) were associated with decreased risk. In a multivariate model of only HD patients age (HR 1.03, p<0.05) was independently associated with increased risk of fractures and systemic PRD (HR 0.36, p<0.05), high serum albumin (HR 0.95, p <0.05) and being on alfacalcidol (HR =0.54, p<0.05) and lanthanum HR 0.46, p<0.05) were associated with decreased risk. In a multivariate model in transplant patients there were no significant independent predictors. Conclusions: The risk of fracture is higher in HD patients than in transplant patients even when controlling for other risk factors. The apparent protective association with alfacalcidol and lanthanum in HD patients deserves further exploration.