Early versus Late Parenteral Nutrition in Critically Ill Adults

Casaer. New England Journal of Medicine 2011;365(6):506-17.

Clinical Question

  • In critically ill adult patients, does late supplemental parenteral nutrition (PN), compared with early PN reduce the duration of dependency on intensive care?


  • This is an important nutrition paper published in 2011. At this time European guidelines recommended initiating PN within 2 days of admission to the ICU for patients who could not be adequately enterally fed. Whereas the American guideline recommended withholding PN and tolerating hypocaloric nutrition during the first week in patients who were not malnourished at baseline. This study aimed to compare these two practices.


  • Prospective, parallel-group, multicentre randomised controlled trial
  • Unblinded
  • Consecutive patients
  • Stratified allocation according to 16 disease categories
  • Allocation concealment maintained by sequentially numbered, sealed, opaque envelopes. Replaced by a digital system during the study
  • Permitted blocks of 10 per stratum (treating team unaware of block size)
  • Power calculation: 4640 patients provided a power of 80% to detect a 1 day change in ICU stay and 70% power to detect a change of 3% in the rate of ICU death, at an alpha level of 0.05.
  • Intention to treat analysis


  • 7 ICUs in Belgium
  • Data collected: August 2007- November 2010


  • Inclusion criteria:
    • Adult patient admitted to the ICU
    • Nutritionally Risk Screening (NRS) score ≥3 (indicative that the patient was nutritionally at risk)
  • Exclusion:
    • Ketoacidotic or hyperosmolar coma on admission
    • BMI < 17 (kg/m2)
    • Short Bowel Syndrome
    • Pregnancy or lactating
    • Taking oral nutrition
    • Moribund or coded DNR
    • Home mechanical ventilation
    • No central catheter
  • 8703 screened for eligibility, of whom 4640 underwent randomisation
  • Groups were comparable at baseline, comparing late vs. early groups
    • Male: 63.8% vs. 64.3%
    • Age: 64 vs. 64
    • Weight: 75kg vs. 76kg
    • Diabetes: 17.9% vs. 16.9%
    • Cardiac surgery:60.5% vs. 61%
    • APACHE II score: 23 vs. 23
    • Nutritional risk screening score
      • 3: 45.1% vs. 43.9%
      • 4/5: 45.9% vs. 46.8%
      • 6/7: 8.9% vs. 9.3%


  • Late parenteral nutrition
    • 5% dextrose in volume equal to that of PN in the early group
    • If oral diet could not be started on day 2, enteral nutrition (EN) was initiated
    • If EN was insufficient after 7 days, PN was initiated on day 8 to meet caloric goal
    • Received a median of 31 units of insulin to reach target blood glucose level


  • Early parenteral nutrition
    • 20% dextrose targeted to provided 400kcal on day 1 and 800kcal on day 2
    • If oral diet could not be started on day 2, EN was initiated
    • PN initiated on day 3 (OlioClinomel or Clinimix, Baxter) to provide 100% of the caloric goal when combined with EN
    • PN was not started if patient expected to tolerate adequate EN
    • When EN covered 80% of the calorie goal or oral intake was commenced the PN was stopped
    • When EN or oral intake fell to less than 50% of the goal, PN was restarted.
    • Received a median of 58 units of insulin to achieve target blood glucose level

Management common to both groups

  • Standing orders specified twice daily review for increase on EN, prokinetic agents, duodenal feeding tubes
  • Received early parenteral trace elements, minerals and vitamins
  • Patient data management system (MetaVision) was used to calculate daily EN and PN targets
  • Ideal body weight was used for energy calculations
  • Calories provided was capped at 2880. Kcal/day
  • Intensive insulin therapy (IIT) was provided to maintain blood glucose levels of 4.4-6.1 mmol/L
  • Management of nutrition after ICU stay was as per the treating medical team

Comparison of nutrition delivery between the two groups

  • No statement in the paper, but from the figures provided, patients in the early initiation group received approximately 25-30 kcal/kg ideal body weight/day from study day 3 and patients in the late group received 25-30 kcal/kg ideal body weight/day after day 8.


  • Primary outcome: Duration of ICU stay (defined as time from admission until met pre-specified criteria for discharge, regardless of whether they were actually discharged that day) – significantly reduced in late group
    • 3 days (IQR 2-7) vs. 4 days (2-9), p=0.02
  • Safety outcomes
    • Discharged live from ICU within 8 days – significantly increased in late group
      • 75.2% vs. 71.7%, p=0.007
    • Mortality rates – no significant difference
      • In ICU: 6.1% vs. 6.3%
      • In hospital: 10.4% vs. 10.9%
      • At 90 days: 11.2% vs. 11.2%
    • Hypoglycaemia during intervention – significantly increased in late group
      • 3.5% vs. 1.9%, p=0.001
  • Secondary outcomes, comparing late vs. early groups:
    • Late group had significantly lower:
      • New infections
        • Any: 22.8% vs. 26.2%, p=0.008
        • Airway or lung: 16.4% vs. 19.3%, p=0.009
        • Bloodstream: 6.1% vs. 7.5%, p=0.05
      • Duration of mechanical ventilation >2 days
        • 36.3% vs. 40.2%, p=0.006
      • Duration of renal replacement therapy
        • 7 days vs. 10 days, p=0.008
      • Duration of hospital stay
        • 14 days vs. 16 days, p=0.004
      • Health Care Cost
        • €16,863 vs. €17,973, p=0.04
    • Late group had significantly higher
      • Median C-reactive protein level
        • 190.6 vs. 159.7, p<0.001
    • No significant differences in functional status at hospital discharge
      • 6 minute walk: 227m vs. 283m (1227 completed assessment)
  • Pre-defined subgroup analysis – no difference in primary outcome or rate of discharged alive from ICU within 8 days for:
    • Patients at increased risk as indicated by the body-mass index <25 or >40 and NRS score ≥5
    • Cardiac surgery patients
    • Sepsis on admission
  • Post-hoc subgroup analysis – patients for whom early enteral nutrition was surgically contraindicated (n=517)
    • Rates of infection – significantly lower in late group
      • 29.9% vs. 40.2%, p=0.01
    • Likelihood of earlier discharge from ICU – significantly greater in late group
      • Hazard ratio 1.2 (95% C.I. 1-1.44, p=0.05)

Authors’ Conclusions

  • Early parenteral nutrition to supplement insufficient enteral nutrition appears inferior to late parenteral nutrition


  • Large scale, robustly conducted RCT (largest in ICU nutrition to date)
  • Specified objective criteria for when patient was ready for the ward
  • Registered on clinicaltrials.gov
  • Intention to treat analysis
  • Allocation concealment maintained


  • Non-blinded
  • Use of intensive insulin therapy makes the results difficult to generalise to ICUs who do not use intensive insulin therapy
  • Short ICU length of stay (3 vs 4 days), raises a question whether the patients included needed artificial nutrition therapy at all
  • Majority of patients had left the trial by day 8 when the late intervention was commenced (655 in late group and 736 in early group at day 7
  • Significant amounts of IV dextrose used in the intervention group- is this the signal the study is picking up?
  • Primary outcome was not patient centred

Competing Interests

  • Baxter Healthcare provided research grant but were not involved in design of the study, analysis of data, or decision to submit for publication.

The Bottom Line

  • Critically ill patients receiving late supplemental parenteral nutrition were more likely to be discharged alive and earlier from the ICU compared to standard care. There was no difference in the mortality rates in hospital or at 90 days.
  • Intensive insulin therapy and significant amounts of IV dextrose were used and thus make the results difficult to generalise to populations where this is not standard practice.
  • There is a question if the patients in this trial required artificial nutrition therapy at all given the short length of stay and high rate of cardiac surgery patients.
  • Practice change can not be recommended, particularly for ICUs who see more unwell patients than those included in this trial and where intensive insulin therapy is not standard care.

External Links


Summary author: Emma Ridley @icunutrition
Summary date: 21/02/2018
Peer-review editor: @davidslessor

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.