LB031 - VALIDATION OF PATIENT-REPORTED NUTRITIONAL RISK SCREENING IN OUTPATIENT ONCOLOGY DEPARTMENT

LB031

VALIDATION OF PATIENT-REPORTED NUTRITIONAL RISK SCREENING IN OUTPATIENT ONCOLOGY DEPARTMENT

L. Teichmann¹, A. Albers^1,*, C. Hofbauer¹, N. Knauthe¹, M. Eichler¹

¹NCT/UCC, University Hospital Carl Gustav Carus Dresden, Dresden, Germany

Rationale: Patient-reported digital screening may support early identification of nutritional risk in outpatient oncology. This study aims to validate an electronic patient-reported outcome (ePRO)-based adaptation of the Nutritional Risk Screening 2002 (NRS-2002) against established NRS-2002 classification.

Methods: This prospective validation study enrolled 150 oncology outpatients at the National Center for Tumor Diseases (NCT/UCC) Dresden. After plausibility filtering, 149 were analyzed. Participants completed a self-administered ePRO-based nutritional risk screening on tablets and underwent a standard NRS-2002 assessment by trained clinical staff 10–13 days later. Matched classifications were compared using a cut-off ≥3 for increased nutritional risk. Agreement and diagnostic performance were assessed by Cohen’s kappa, ROC analysis, sensitivity, and specificity.

Results: Nutritional risk was identified in 46 patients (30.9%) by ePRO-based screening and in 54 (36.2%) by standard NRS-2002 assessment. ePRO-based screening showed 66.7% sensitivity, 89.5% specificity, 81.2% accuracy, moderate agreement (κ=0.580), and good discrimination (AUC=0.859). Discordant classifications occurred in 28 cases: 9 reflected nutritional status scoring,10 disease severity assessment, and 9 both components.

Conclusion: ePRO-based screening showed moderate agreement, high specificity, and good discrimination compared with established NRS-2002 classification. Discordance across nutritional status scoring and disease severity assessment suggests that variability may be attributable to recall bias in patient-reported historical data, especially previous body weight and recent weight loss, as well as differences in the clinical interpretation of NRS-2002 components, rather than to the patient reported assessment mode.

Further evaluation should focus on standardizing disease severity assessment and the timing of repeated measurements.

Disclosure of Interest: None declared