Marjon Borgert

101 Potential risk factors in the delivery of enteral nutrition the patients received their daily protein targets of more than 80%. Our cohort, however, showed a wide variation in the delivery of EN intake as shown in Figure 1. Approximately 30% of the enterally fed patients still received inadequate EN, i.e. values below 80%. This indicates that there is room for further improvement in the delivery of EN. During the rst days of admission the actual calorie and protein EN intake was low. This was to be expected since in our EN protocol the rst ve days are used to build-up to patients’ ideal EN intake. Furthermore, our results show that EN was initiated within 24-48 hours following admission. The early initiation of EN within the rst 24-48 hours following ICU admission is strongly recommended in EN guidelines. 7 Observational studies have shown that patients who received an early start of EN had lower morbidity and lower mortality rates than patients who did not. 18 The results from our model explained 7% of the average EN intake per hour, and may be accounted for a large variance in EN intake. This study was, however, not performed as an attempt to identify factors that contribute to a success or failure of EN intake; rather to describe the daily EN intake in critically ill patients and to determine groups of patients or areas where the daily EN intake might be inadequate while controlling for covariates. The model showed, however, that nasoduodenal-, nasojejunal- and PEG feeding tubes were factors for improved success of feeding compared to the nasogastric feeding tube. This can be explained by the fact that patients who fail to be fed by nasogastric tubes are in most cases fed by post pyloric feeding tubes. 19 Furthermore the model shows that medical patients were associated with a 5.10-point increase in average EN intake per hour (beta 5.10, P -value < 0.001) and a decrease in planned admissions (beta -7.59, P -value < 0.001). It may suggest that medical patients are associated with better EN intake than surgical patients. Other studies showed similar ndings. 2,20 Dover et al. showed that surgical patients received less EN intake compared to medical patients. While patient undergoing cardiovascular and gastrointestinal surgery are even at higher risk of receiving inadequate nutrition. 20 It is suggested that there might be a delay in initiating EN due to the hemodynamic instability in these patients. Hemodynamic instability might be a barrier for some physicians to start the feeding protocol. 20 In our model we did not account for the di erent types of surgery. It is known that using nurse-driven EN protocols or advice from dieticians is associated with improved feeding practices on ICUs. 21 This may have contributed to the moderate to high levels of EN delivery in our ICU. However, we showed that a substantial part of the EN delivery is still suboptimal and needs to be improved. In our view this implies a strong argument for the development of a nutritional care bundle to support guideline uptake and thereby improve the delivery of EN. 7,22 Care bundles are designed