P1529MORTALITY PREDICTORS IN INCIDENT HEMODIALYSIS PATIENTS - HAVE HEY CHANGED IN TEN YEARS?

Abstract

Abstract Background and Aims In the last years there has been an increase in elderly patients with multiple comorbidities inducing hemodialysis (HD). Since dialysis treatment itself may be associated with a further deterioration in functional status, nephrologists are increasingly careful in selecting these patients for HD. Concerned with this reality we tried to understood if early mortality predictors (in the first 6 months) in incident HD patients have changed in almost 10 years, in the same hospital HD unit. Method This is a retrospective observational study of incident HD patients between 01 January 2017 and 30 June 2019. We evaluated similar clinical, analytical and demographic data to those used to predict mortality in the same HD unit from 1 January 2010 to 30 September 2014. Logistic regression analysis was used to evaluate 6 month mortality predictors. Statistical analysis was performed using SPSS version 25 for Windows. Results The average age of 163 incident HD patients were 70.63±3.9 years (similar to the previous population: n= 235; 70.7 ± 14.9 years) and 57.1% were male. During this study we observed 26 (16%) deaths, 12 of which (46.15%) occurred in the first 6 months of hemodialysis. Pneumonia and cachexia were the major causes of mortality, unlike the previous population, in which majority of deaths were attributed to cardiovascular events. Between January 2010 to 30 September 2014 the strongest predictors of early mortality were dementia [adjusted odds ratio (OR) 15.94 (CI: 4.09–62.10)], central venous catheter use [(OR) 12.29; (CI: 3.54-42.65)], cancer [(OR) 4.64 (CI: 1.48-14.54)] and heart failure [(OR) 3.57 (CI: 1.08-11.75)]. Differently, in this study, the institutionalization and the presence of metastases were the predictors that showed a higher risk of death [p=0.005; adjusted odds ratio [(OR) 10.4 (CI: 2.017–49.9) and p=0.01; (OR): 14.9 (CI: 1.89-42), respectively]. Longer hospitalizations at the time of HD induction [(p=0.044; (OR):1.103; CI: 1.003-1.213)] and albumin values <2.5 mg/dL [(p=0.03; (OR): 3.8 (CI: 1.14-13)] were also strong mortality predictors. which were not previously observed. Conclusion Nowadays, nephrologists are less liberal in initiating dialysis to elderly patients with dementia and cardiovascular comorbidities. However, the population is getting older and our recent mortality predictors may reflect the aging of chronic kidney disease patients, who have multiple comorbidities as cachexia, requiring institutionalization and longer hospitalizations. It is increasingly important to evaluate patients prior to dialysis initiation, so our goal is to create an adjusted mortality score in our HD unit to help make the decision about inducing or not HD in our patients.