DEVELOPING A QUANTITATIVE FIRST LINE SCREENING TOOL FOR PROTEIN-ENERGY WASTING IN MAINTENANCE HAEMODIALYSIS PATIENTS

C GODBER1, A WU1, A CLOSE1,2, M GALLAGHER1,3, A SEN1,3

1University Of Sydney, Sydney, Australia, 2Canterbury Hospital, Sydney, Australia, 3Concord Repatriation General Hospital, Sydney, Australia

Background: Protein energy wasting (PEW) affects 28-54% of maintenance haemodialysis patients and significantly increases the risk of cardiovascular complications and death. Current methods of screening and assessment for PEW rely on clinician involvement which is often impractical, and may delay access to intervention.
Aim: To develop a quantitative and automated first-line screening tool for PEW in maintenance haemodialysis patients
Methods: Subjective Global Assessments (SGA) were completed as part of standard care for 57 patients in 2014. These scores were retrospectively collated with a score of 5 or lower indicating the presence of PEW. A literature review identified seven relevant and routinely recorded parameters (albumin, c-reactive protein, transferrin, phosphate, cholesterol, creatinine and body-mass index) to be used in the screening algorithm. Using each patient’s data from the time of the SGA a binomial logistic regression was completed to derive an algorithm for calculating PEW risk on a 0-1 scale. The accuracy of the tool in this population was then evaluated.
Results: 29 patients had SGA defined PEW and 82.8% of these patients were identified as having a risk score over 0.5. Of the studied variables, BMI (p<0.001) and albumin (p<0.036) were most strongly correlated with SGA results. A receiver operating characteristic curve was drawn using the algorithm derived from the regression and the area under the curve was 0.895.
Conclusions: This study demonstrated that PEW risk may be predicted using routinely collected biochemical and anthropometric parameters. Further analysis would need to be conducted to determine the ideal threshold value to indicate PEW intervention, the most efficient selection of parameters and strategies for clinical implementation. Validation in a larger and more diverse cohort is also required.


Biography:
Charles Godber is a medical student at the University of Sydney. Prior to this he completed a Bachelor of Biomedicine at the University of Melbourne

Categories