Scores of nutritional risk and parameters of nutritional status assessment as predictors of in-hospital mortality and readmissions in the general hospital population.

Background & Aims: We have no "gold standard" for the diagnosis of malnutrition. The aim of this study was to determine the importance of many of the parameters used in nutritional status screening and assessment among inpatients for the prediction of in-hospital mortality, readmission and length of hospitalization.

Methods: On the base of the medical documentation a retrospective analysis was performed of nutritional status screening and assessment parameters for all 20,237 non-selected, consecutive hospitalizations in 15,013 patients over 18 years of age treated in one hospital during the course of one year.

Results: The risk of malnutrition expressed as a Nutritional Risk Screening (NRS)-2002 score ≥ 3 concerned 6.4% hospitalizations. The greater risk of in-hospital death, as well as readmission within 14 days and 30 days, was related to an NRS-2002 score ≥3, age >65 years, male gender, urgent admission, body mass deficit calculated as the difference between actual body mass and ideal weight determined according to the Lorentz formula, higher degree of Instant Nutritional Assessment (INA), greater value of a C-reactive protein (CRP)/albumin ratio, and plasma glucose concentration. Whereas, greater blood concentration of albumin, hemoglobin, cholesterol and triglycerides, as well as a greater blood lymphocyte count, were associated with reduced risk of the measured outcomes.

Conclusions: NRS-2002 score, blood albumin, CRP/albumin ratio, and INA seem to be good predictors of in-hospital mortality, readmission rate and length of hospital stay.