Nutritional risk screening in a Danish university hospital is insufficient and may underestimate nutritional risk: A...

Disease-related-malnutrition predicts poor clinical outcomes in elderly patients, and screening is pivotal for identifying patients at nutritional risk. The present study aimed to investigate nutrition screening rates in electronic patient records and validate the scores given. A secondary aim was to investigate whether the proportion of patients at risk differed between patients where screening was documented and those where no screening was documented.

This cross-sectional observational study was conducted in a Danish university hospital during November 2020. Patients aged 65 years or more admitted to a medical department were included. The Nutrition Risk Screening 2002 (NRS-2002) tool was used to identify patients at nutritional risk, both in routine clinical care, where data were collected retrospectively, and during a validation process in a random patient sample, where data were collected prospectively.

In total, 817 patients were admitted for more than 24 h. Of these, an NRS-2002 score was documented in 294 (36%), among whom 177 (60%) were at nutritional risk. In 146 patients where no score was documented, 88 (60%) were at risk. Validation was possible in 91 patients where a record-based score and a validated score were documented. The specificity of the record-based score was 100%, whereas the sensitivity was 75%, indicating that routine screening underestimated nutritional risk (p < 0.001, proportion difference 19%; 95% confidence interval = 10%-28%).

Electronic documentation does not solve issues about compliance with nutritional risk screening. In patients where screening was not documented, the occurrence of nutritional risk was similar, indicating that omission of screening is not related to the score.