Are traditional screening tools adequate for monitoring the nutrition risk of in‐hospital patients? An analysis of the nutritionDay database

Abstract Background Monitoring of adequate food intake is not a priority in hospital patients' care. The present study aimed to examine selective data from the nutritionDay survey to determine the impact of food intake during hospitalization on outcomes according to the nutrition risk status. Methods We conducted a descriptive analysis of selected data from 7 consecutive, annual, and cross‐sectional nutritionDay samples from 2009 to 2015. The impact of food intake on outcomes was assessed by univariate and multivariate Cox models controlling for PANDORA scores. Results A total of 7994 adult patients from Colombia, 7243 patients from 9 Latin American countries, and 155,524 patients worldwid were included. Less than half of the patients worldwide consumed their entire meal on nutritionDay (41%). The number of reduced eaters is larger in the “no nutrition risk group” than in the “nutrition risk group” (30% vs 25%). Reduced eating is associated with higher mortality and delayed discharge in patients, regardless of the nutrition risk status. Patients without nutrition risk at the screening who ate “nothing, but were allowed to eat” had 6 times more risk of mortality (hazard ratio, 6.48; 95% CI, 3.5311.87). Conclusions This is the first large‐scale study evaluating the relationship of food intake on clinical outcomes showing an increase of in‐hospital mortality rates and a reduction in the probability of being discharged home regardless of the nutrition risk status. Traditional screening tools may not identify a group of patients who will become at risk because of reduced intake while in the hospital.