Is modified creatinine index really superior to geriatric nutritional risk index for predicting malnutrition and clinical outcomes in hemodialysis patients?