C-REACTIVE PROTEIN TO ALBUMIN RATIO AND 6-MONTH MORTALITY IN INCIDENT HAEMODIALYSIS PATIENTS

Abstract BACKGROUND AND AIMS The prevalence of chronic kidney disease (CKD) is growing worldwide and ranges from 8% to 16%. Mortality rates are higher in the first few months of haemodialysis (HD). Protein-energy malnutrition has been demonstrated to be a major risk factor for mortality in this population. The C-Reactive Protein to Albumin ratio (CAR) has been associated with increased mortality risk. We aimed to evaluate if CAR could be used to predict 6-month mortality in incident HD patients. METHOD Retrospective analysis of CKD patients who initiated chronic HD between January of 2014 and December of 2019 in a tertiary-care hospital in Portugal. CAR was calculated at HD start. We analyzed 6-month mortality. Variables were submitted to univariate and multivariate analysis to determine factors predictive of 6-month mortality after HD start. We assessed the logistic regression method of the CAR to predict 6-month mortality and the discriminatory ability was determined using the receiver operating characteristic (ROC) curve. RESULTS A total of 787 patients were analyzed (mean age 68.34 ± 15.5 years and 60.6% male). The 6-month mortality was 13.8% (n = 109). Patients who died were significantly older [76.50 ± 11.39 versus 67.29 ± 15.52 years; P < 0.001, OR: 1.055 (1.035–1.074); P < 0.001, aOR: 1.058 (1.030–1.086); P < 0.001], had more frequently cardiovascular disease [65.1% versus 46.1%; P < 0.001, OR: 2.192 (1.437–3.342); P < 0.001, aOR: 2.210 (1.210–4.037); P = 0.010], central venous catheter at HD start [83.5% versus 58.3%; P < 0.001, OR: 3.622 (2.136–6.142); P < 0.001, aOR: 3.090 (1.584–6.026); P < 0.001], lower PTH [229.44 ± 170.50 versus 365.95 ± 415.80; P = 0.006, OR: 0.998 (0.996–0.999); P < 0.001, aOR: 0.998 (0.997–1.000); P = 0.014] and higher CAR [2.85 ± 3.85 versus 1.36 ± 2.44; P < 0.001, OR: 1.159 (1.086–1.236); P < 0.001, aOR: 1.126 (1.023–1.239); P = 0.015]. The AUC for mortality prediction was of 0.706 [95% confidence interval (0.65–0.76); P < 0.001]. The optimal CAR cut-off was >0.5, with an odds ratio of 5.362 (95% CI 3.208–8.963; P < 0.001). CONCLUSION In our study, we demonstrated that higher CAR was independently associated with a higher mortality rate in the first 6 months of starting HD, highlighting the prognostic importance of malnutrition and inflammation in patients starting chronic HD.