Proefschrift Kerklaan

Chapter 2

Variation inpracticewas not onlyobservedbetweenPICUs inour current study; we also received conflicting statements within single institutions. We corrected for this issue, by weighting by the inverse of the number of completed questionnaires per center. The conflicting statements underline the observed variation of nutritional practices, which occurs not only between but also within individual institutions. A similar discordance in practice within institutions was reported in a U.K. survey of glycemic control in PICUs 22 . Globally, guidelines for nutritional support have been released by nutritional organizations.The American (A.S.P.E.N). and European (ESPEN/ESPGHAN) societies provide specific guidelines for nutrition in critically ill children 7,8 . However, they do not advise on every aspect of nutritional support. Agreements and differences between these guidelines and current practice, as shown by our survey, are summarized in Supplementary Table 2. Overall, the most striking similarity between guidelines and local implementation is the preference for EN as the preferred route of nutrient delivery and its early initiation in critically ill children. A specialized NST and feeding protocol are recommended by the A.S.P.E.N guidelines for critically ill children 7 . Availability of an EN protocol is associated with a lower prevalence of hospital-acquired infections 3 , implementation of an NST with an increase in EN use, and decreased reliance on PN 23 . Our survey showed that a nutritional protocol and/or NST were available in approximately half the PICUs. In our point prevalence, we found no significant difference in caloric intake and use of EN between patients from centers with and without a protocol. However, since this was a secondary analysis, it cannot prove or disprove the utility of NST/protocols in general. In single centers, a stepwise EN algorithm has been shown to significantly improve the timing of EN initiation and the ability to reach nutrient delivery goals 24,25 . The role of protocols and NSTs in optimizing clinical outcomes in the PICU population needs to be further examined in well-designed trials. The ESPEN/ESPGHAN guidelines prefer the measurement of resting energy expenditure (REE) to the use of equations. The A.S.P.E.N. guidelines recommend targeted use of IC in a select group of patients with suspected metabolic alterations or malnutrition. Both state that in the absence of IC, reasonable values can also be derived from formulas, for example, Schofield 17 , but only when applied without the use of universal correction factors 7,8 . Several other sources state that nutritional therapy should be targeted at REE throughout the course of illness 26,27 . However, due to the limited availability and practice of IC 11 , and also to inaccurate predictive equations 26-28 , it is difficult to assess REE in critically ill children. Use of the WHO and Schofield equations, most commonly used to determine requirements, may lead to underfeeding and overfeeding and potentially impacts morbidity and mortality 3,4 . We confirmed the finding of previous studies 11 that IC to measure REE is used in a small minority of European (20%) and

42