PD741 - NUTRITIONAL RISK ASSESSMENT IN ALLOGENEIC HSCT: LACK OF CONCORDANCE BETWEEN NRI AND NUTRISCORE CHALLENGES CLINICAL PRACTICE

PD741

NUTRITIONAL RISK ASSESSMENT IN ALLOGENEIC HSCT: LACK OF CONCORDANCE BETWEEN NRI AND NUTRISCORE CHALLENGES CLINICAL PRACTICE

J. Sicchieri^1,*, P. N. Bezan¹, N. A. Alves¹, T. C. M. Costa¹, F. Traina², A. M. Navarro²

¹Hospital das Clínicas, Ribeirão Preto Medical School, University of São Paulo, ²Ribeirão Preto Medical School, University of São Paulo, Ribeirão Preto, Brazil

Rationale: In post-allogeneic Hematopoietic Stem Cell Transplantation (HSCT), accurate nutritional risk assessment is crucial. The Nutritional Risk Index (NRI) and Nutriscore are frequently used, but their diagnostic concordance is controversial. This study compared their concordance in post-allogeneic HSCT patients using specific cut-offs to verify if they provide equivalent risk classifications.

Methods: A cross-sectional analysis of electronic medical records from an HSCT outpatient clinic over 2 months. NRI was classified as severe risk when <83.5 and Nutriscore as at risk when >5. Spearman’s correlation evaluated continuous variables. Concordance between categorical classifications was assessed using Cohen’s Kappa coefficient, sensitivity, specificity, predictive values, and accuracy. Significance was set at p<0.05.

Results: The sample included 46 patients. Continuous variables showed a significant negative correlation between NRI and Nutriscore (r=-0.640; p<0.001). However, categorical analysis revealed drastic discordance: 100% (n=46) were “severe risk” by NRI (<83.5), while only 2.2% (n=1) were “at risk” by Nutriscore (>5). The Kappa coefficient was 0.000, indicating null concordance. NRI showed 100% sensitivity (1/1), 0% specificity (0/45), 2.17% positive predictive value, and 2.17% accuracy, demonstrating total incompatibility between the instruments’ classifications.

Conclusion: Despite continuous score correlation, the analyzed cut-offs result in a total lack of diagnostic concordance in post-allogeneic HSCT patients. These thresholds lead to extremely divergent risk classifications, making the instruments non-interchangeable. Findings reinforce the critical need to validate specific, clinically relevant cut-off points for each tool to ensure accurate risk assessment.

Disclosure of Interest: None declared