ESPEN diagnostic criteria for malnutrition - A validation study in hospitalized patients.

BACKGROUND & AIMS: The European Society for Clinical Nutrition and Metabolism (ESPEN) released a consensus statement for undernutrition diagnosis: ESPEN diagnostic criteria for malnutrition (EDC). The EDC lacks validation and therefore, the present study aims to assess the concurrent and predictive validity of this tool in a cohort of inpatients.

METHODS: A prospective observational study took place in a university hospital. Concurrent validity of EDC was evaluated using the Patient Generated Subjective Global Assessment (PG-SGA) nutrition status classification as the reference method. Sensitivity, specificity, positive and negative predictive values were determined. The EDC predictive validity was assessed by its independent association with length of hospital stay (LOS), applying Cox proportional hazards ratio method.

RESULTS: Of the 632 included patients, 455 participants (72%) were nutritionally-at-risk (Nutritional Risk Screening initial screening). For those that had screened positive, 260 (57.1%) and 55 participants (12.1%) were undernourished according to PG-SGA and to EDC, respectively. Compared to PG-SGA, the EDC revealed a sensitivity of 17.1% and a specificity of 98.3%. Positive and negative predictive values were respectively 89.1% and 58.9%. Undernutrition evaluated by EDC was independently associated with lower hazard ratio for being discharged home over time, 0.695 (95% confidence interval: 0.509; 0.950).

CONCLUSIONS: The EDC could be used in clinical settings to confirm undernutrition suggested by other methods. The independent association of undernutrition by EDC with LOS shows this method is of clinical relevance.