This study indicates that, within our population, 72% of the nurses’ malnutrition risk assessments agree with those of the dieticians, with a weak inter-rater reliability (Cohen’s Kappa = 0.44, p < 0.001). Falsely labeling patients ‘not at risk’ for malnutrition, is correlated with a delayed involvement of the dietician (1.06 days; [95% CI 0.92–1.20], p < .001). Importantly, however, a later involvement of the dietician, represented by the STD, is correlated with a prolonged LoS for patients ‘at risk’ for malnutrition with 31% (HR = 0.69, [95%CI 0.63–0.75]), when this STD is between three and six days as compared to patients with a STD less or equal than 3 days. This even increases up to 49% (HR = 0.51, [95%CI 0.43–0.61]) for the patients falsely labeled ‘not at risk’ in the same timeframe.
The nurse plays a key role in the malnutrition risk screening, documenting the patient’s status in the electronic health record, and ensuring that (future) nutrition prescriptions are initiated. With timely and correct judgment, they can give patients a jump start on improving their health status as this will ensure a comprehensive assessment and nutritional care plan from the dietician12. However, nutrition care is one of the elements that is sometimes implicitly rationed within the already large nurse workload16, 17. Nevertheless, screening in general remains of upmost importance to deliver excellent patient care.
As this is the first study evaluating the use of NRS2002 performed by two different health care workers within a real-world data set, a comparison with other studies is not yet possible. Although the problem of poor inter-rater reliability is not unique for malnutrition risk screening, results of inter-rater reliability in malnutrition screening need to be interpreted carefully as the assessment at admission by the nurse and the follow-up assessment by the dietician are not exactly the same. Nevertheless, the level of agreement between both assessors in our study seems largely similar to results obtained within pain research9, pressure injury10, and triage at the emergency department18. In the assessment of pain, several barriers from the nurses’ perspective are raised19, such as the nurse work load (patient-to-nurse ratio), the lack of training and guidance, but also the possible indifference of the nurse, and the difficulties in obtaining relevant answers to the screening questions from patients. Meehan et al. revealed similar barriers regarding nutritional screening 12. Strategies that might help to overcome these barriers can among others be: deploying multidisciplinary nutrition teams, educating all health professionals emphasizing more awareness to the impact of malnutrition on patient outcomes and healthcare costs12, and above all convincing them to consider nutrition as a true medicine20. Last but not least, actively involving patients in their nutritional care might also be a promising strategy: as on the one hand active participation is extremely empowering21, and on the other hand it might simultaneously optimize screening results, since patients are best placed to answer nutrition related questions themselves. Nevertheless, despite all flaws screening tools may possess, it remains of upmost importance to keep screening patients for their potential malnutrition risk upon admission.
The longer LoS for patients ‘at risk’ for malnutrition was already shown in previous studies1, 4, 22. We add timely involvement of the dietician, in the form of the STD, as an extra factor in relation with the LoS. We observe a significant longer STD in the group of patients classified as FN compared with the group of patients classified as TP. In our population, we found that the larger the STD, and thus the later the dietician gets involved, the longer the hospital stay is extended, independently of an accurate assessment by the nurse. However, focusing only on patients with an ‘at risk’ assessment by the dietician, we demonstrate that the LoS is even longer in the group of patients classified as FN than in the group of patients classified as TP (Table 4). These findings emphasize the importance of optimizing accurate and timely malnutrition screening of all inpatients to improve their outcome, i.e. LoS, and reducing health care costs. Timely screening and initiation of individualized (nutritional) care is also confirmed to be effective in other studies11–13.
A first novelty of this study is the analysis of real-world data, reflecting genuine practice in tertiary care. A second novelty is that it reveals a positive correlation between the STD (the time until dietician involvement) and the LoS, taking into account the pathology and severity of illness. This study may act as a catalyst for further research. A pre-post study may be set up, for example after training nurses in nutritional status assessment. Also empowering patients in their nutritional care should be further explored and evaluated21. Using for instance a web-based questionnaire in which patients can enter their height and weight and/or answering questions regarding their current nutritional status before a planned admission can be very promising, since it might optimize screening results and give patients an active role in improving their general health status during and after hospitalization.