To assess agreement between the CONUT and Mini Nutritional Assessment (MNA) instruments in detecting malnutrition in hospitalized elderly patients and to determine their prognostic value compared to that of serum albumin alone in relation to in-hospital and 1-month and 6-month post-discharge mortality rates.

Prospective observational study.

A Catalan regional hospital.

2155 patients admitted to an acute geriatric unit were assessed using MNA and CONUT and were followed up for 6 months after discharge.

On admission, data were collected on age, sex, referral, geriatric syndromes, cognitive status, functional status and nutritional status according to MNA (as the gold standard). Plasma albumin, total cholesterol and lymphocyte levels were recorded to implement CONUT. Mortality was recorded until 6 months after discharge.

Sample characteristics: 61.3% females, mean age 84.9 years, mean Charlson index 2.2. CONUT sensitivity and specificity for malnutrition were 43% and 71.6%, respectively, with positive and negative predictive values of 88.9% and 19.2%. MNA, CONUT and albumin alone were good predictors of mortality but showed similar sensitivity and specificity results.

CONUT agreement with MNA in nutritional risk assessments for elderly people is poor. Although CONUT is a good predictor of short-and medium-term mortality, it adds little to information provided by albumin alone.