Aktuelle Ernährungsmedizin 2012; 37 - P3_1
DOI: 10.1055/s-0032-1312511

Grading the severity of disease – a key component for nutritional screening in the hospital?

M Jeske 1, 2, KM Ebner 3, O Galvan 4, R Beer 5, H Ulmer 6, M Joannidis 3
  • 1Innsbruck University Hospital, Department of Pharmacy, Innsbruck, Austria
  • 2Institute of Pharmacy, Department of Pharmacology and Toxicology, Innsbruck, Austria
  • 3Innsbruck Medical University, Department of General Internal Medicine, Innsbruck, Austria
  • 4Innsbruck University Hospital, Department of Otorhinolaryngology, Innsbruck, Austria
  • 5Innsbruck University Hospital, Department for Neurology, Innsbruck, Austria
  • 6Innsbruck Medical University, Department of Medical Statistics, Informatics and Health Economics, Innsbruck, Austria

Introduction: The grading of severity of disease is a central element of many nutritional screening tools. However, evaluation of the severity of disease can be difficult for the nursing staff or the dietitian. The aim of this study was to investigate if this assessment is essential for a valid and significant nutritional risk screening in inpatients at admission.

Material and methods: 1000 randomly selected, consecutive adult patients admitted to the departments of internal medicine, neurology or otorhinolaryngology were included in the study. All patients were assessed by Malnutrition Universal Screening Tool (MUST) and the Innsbruck Nutrition Screening Tool (INST), which do not include severity of disease. The results were compared to the validated Nutrition Risk Screening NRS 2002 performed in the same group of patients. Classification by SGA was used as a reference. Sensitivity, specificity of MUST and INST as well kappa-coefficients with SGA were calculated.

Results: 554 of the patients did not show diseases influencing the nutritional status. 95,5% of those 554 patients were classified well nourished by SGA. NRS 2002 compared to SGA successfully detected patients “at risk”: sensitivity 0,91 95%CI (0,86202–0,94757), specificity 0,88 95%CI (0,85–0,90) and kappa-coefficient 0,67. Although sensitivity (0,85) indicated that MUST correctly identified most patients “at risk for undernutrition”, 169 patients where falsely classified as “at risk for malnutrition” (specificity 0,79, cappa-coefficient 0,50). This revealed a moderate agreement between the two methods. 115 of the 169 patients did not suffer from diseases influencing their nutritional status. INST in contrast classified only 25 patients falsely as “at risk” (specificity 0,96535), but failed to identify 77 patients “at risk” (sensitivity 0,60 95%CI (0,53–0,67)), kappa-coefficient was even lower (0,47). 67 of those 77 patients had a disease influencing their nutritional status.

Conclusion: In our study population MUST or INST was inferior to NRS 2002. Correct grading of “severity of disease” seems to be a key component for an overall routine nutritional screening on admission to hospital.