Comparison of different scoring methods for assessing the nutritional status of hospitalised patients.

OBJECTIVE: Despite intense clinical research, no commonly accepted diagnostic tool for assessment of nutritional status is yet available. In this study a comparison of four different methods for diagnosis of the nutritional status of patients admitted to a university hospital in Austria is presented. PATIENTS AND METHODS: Clinical data of 640 hospitalised patients were analysed in a prospective-descriptive study design. Four recommended methods, the Innsbruck nutrition score (INS), the Prideaux nutritional risk assessment (PNRA), the well established nutrition risk index (NRI), and the body mass index (BMI) were used to analyse nutritional status. RESULTS: The BMI showed 90.2% of the patients evaluated to have normal nutritional status, whereas the PNRA identified 48.9%, the NRI 40% and the INS 58.6% as well nourished. Patients were variously diagnosed with moderate malnutrition: 9% (BMI), 42% (PNRA), 54.8% (NRI) and 30% (INS). Severe malnutrition was detected in 0.5% (BMI), 9.1% (PNRA), 5.2% (NRI) and 11.4% (INS) of the patients evaluated. Cancer patients had the worst nutritional status. CONCLUSION: Malnutrition seems to be a common diagnosis among hospitalised patients in Austria. Screening and assessment of nutritional status should be integrated into clinical routine. The methods tested scored malnutrition at different frequencies. BMI seemed to underestimate the prevalence of malnutrition. The PNRA provided some information on clinical outcome, whereas the NRI had the best relationship between the degree of malnutrition and length of stay. Calculation of the INS may give correct diagnosis of severe malnutrition. Further prospective clinical studies are needed to validate the scoring systems used in this study and to provide accurate clinical diagnosis of malnutrition.