159 / 208
General discussion
157
8
does not provide recommendations on macronutrient intake, whereas the ESPEN/ESPGHAN
only has a guideline on adult EN, not on EN in children. Moreover, the PN guidelines of the
ESPEN/ESPGHAN are targeted at children in general, and information on critically ill children is
limited to specific and generally small sections. Finally, the phase of critical illness (Chapter 1
and 4) is not taken into account in any of the recommendations for critically ill children, while
a large proportion of nutritional studies are limited to the (semi)acute phase.
Even after a highly needed update of current guidelines, recommendations guided by high-
level evidence remain scarce. However, as alternatives are lacking, they reflect the best
available evidence. Furthermore, since availability of a nutritional protocol, even if based on
low-grade evidence
8
, is associated with improved outcome, clinical implementation of these
recommendations is useful. The judgment of the clinician based on individual circumstances
of the patient must always take precedence over these guidelines
10
.
ENERGY EXPENDITURE
Measurement of REE
Current guidelines advise to match REE by caloric intake to limit caloric deficit
1
. In a subgroup
of patients with suspected metabolic alterations or malnutrition, accurate measurement of
energy expenditure using IC is desirable
1
. This practice is challenging however, due to the
limited availability of IC (Chapter 2) and the inaccuracy of predictive equations (Chapter 4)
to calculate REE. Ventilator-derived VCO
2
measurements are a promising alternative method
to determine REE in mechanically ventilated children weighing 15 kg or more because of
its increased precision compared to the current predictive equations (Chapter 3). However,
these results should mainly be regarded as a proof of concept, mainly due to the small study
population. Further validation of this method to improve accuracy of the measurements
and to detect sources of error in a larger cohort of critically ill children is needed before
implementation in clinical practice is possible. This could be done by collecting and comparing
REE values derived from IC and the ventilator whenever IC is performed (Chapter 3), and
ultimately by investigating whether adjustment of nutrition based on ventilator-derived VCO
2
measurements improves clinical outcome of critically ill children.
Since this method appears to be valid only in children weighing ≥15 kg, around 60% of PICU
patients at risk will still need to rely on inadequate predictive equations to determine REE in
absence of IC. The inadequacy of the ventilator-derivedmethod in children <15 kg, presumably
caused by sampling specifications, is not likely to be resolved in the near future. However, this
method should again be investigated in smaller children, when the technical performance
of the CO
2
sensor is improved (by an increased accuracy and sampling frequency), and an
improved version of the airway adapter causes less increase of dead space and turbulence.
Moreover, even in children weighing ≥15 kg, this method cannot be validated using regular IC.