A large (2400) pragmatic (real-life!), randomised, unblinded trial in non-elective admissions to 33 UK ICUs, CALORIES looked at the difference in early feeding enterally ‘v’ parenterally, powered for a 20% difference in mortality at 30 days. Feeding was started within 36 hours of admission, aiming to achieve 25 kcal/kg/day by 72hr, and continued for 5 days. A different focus to EPaNIC and Doig 2013 in that the patients were able to receive either EN or PN.
Most patients in each group didn’t achieve their target, and caloric intake was about the same in each. The parenteral group had less hypoglycaemia (4 ‘v’ 6%), smaller gastric residuals (~100 v ~1000ml) and less vomiting.
But 90 day mortality was no different (around 33%), nor was the length of stay or rate of infective complications, liver dysfunction, GI ischaemia, abdominal distension, aspiration or electrolyte disturbance.
About 7% of the parenteral group (ie to enteral) and 1.5% in the other direction – of uncertain significance.
PN might be expected to be more likely to deliver target calories but in the real world there are logistical barriers to this. Based on this study route doesn’t matter, so long as the patient group is similar to yours (15% post-op, APACHE 20, mean age 65, mortality 30% etc)!
Early ‘v’ late PN’s not beneficial but early EN and PN have the same outcomes? Depends on how you feel about EpANIC‘s shortcomings perhaps. Better feed formulation and less CRBSI may have changed the risk/benefit balance.
Is this a myth busted? Cost, technical aspects and patient specific factors will continue to inform our choice, but maybe we can we be freer with our early PN. And will the guideline compilers act on this?