Skip to main content

Foundational Trials

NUTRIREA-1

Image: Shutterstock / luchschenF

Publication

  • Title: Effect of Not Monitoring Residual Gastric Volume on Risk of Ventilator-Associated Pneumonia in Adults Receiving Mechanical Ventilation and Early Enteral Feeding: A Randomized Controlled Trial
  • Acronym: NUTRIREA-1
  • Year: 2013
  • Journal published in: JAMA
  • Citation: Reignier J, Mercier E, Le Gouge A, et al; for the Clinical Research in Intensive Care and Sepsis (CRICS) Group. Effect of not monitoring residual gastric volume on risk of ventilator-associated pneumonia in adults receiving mechanical ventilation and early enteral feeding: a randomized controlled trial. JAMA. 2013;309(3):249-256.

Context & Rationale

  • Background
    • Routine gastric residual volume (GRV) monitoring was widely used during gastric enteral nutrition (EN) to identify “feeding intolerance”, reduce regurgitation/aspiration risk, and prevent ventilator-associated pneumonia (VAP).
    • GRV measurement is technique-dependent (tube size, aspiration technique, patient position) and may not reflect true gastric volume or aspiration risk.
    • GRV-triggered interruptions commonly reduce EN delivery, creating calorie deficits during prolonged critical illness.
  • Research Question/Hypothesis
    • In invasively ventilated adults receiving early gastric EN, is a strategy of not monitoring GRV non-inferior to routine GRV monitoring for the incidence of VAP?
    • Secondary hypothesis: omitting GRV monitoring would improve nutrition delivery (fewer interruptions), without excess harm.
  • Why This Matters
    • GRV checks are labour-intensive and drive feeding pauses; a safe “no-GRV” strategy could simplify care and improve delivered calories.
    • Conversely, if GRV monitoring prevented aspiration/VAP, abandoning it could cause harm; a pragmatic multicentre trial with robust VAP ascertainment was required.

Design & Methods

  • Research Question: Among mechanically ventilated ICU adults started on early EN, does omitting routine GRV monitoring result in a non-inferior incidence of VAP (noninferiority margin 10%) compared with routine GRV monitoring?
  • Study Type: Randomised, multicentre (9 French ICUs), open-label, parallel-group, noninferiority trial (investigator-initiated); web-based randomisation stratified by centre with permuted blocks (size 4); independent adjudication of VAP and ICU-acquired infections by a blinded committee.
  • Population:
    • Inclusion: Adults (≥18 years) intubated <36 hours; expected mechanical ventilation >2 days; enteral nutrition started within 36 hours after intubation.
    • Key exclusions: Enteral nutrition before ICU admission; allergy to enteral formula; pregnancy/breastfeeding; legal protection; prior trial enrolment; expected death within 48 hours; long-term enteral nutrition; contraindication to enteral nutrition; pneumonia after intubation; expected extubation; tracheotomy performed before day 2.
    • Feeding route/setting details: Gastric feeding via 14–16 Fr nasogastric tube; continuous infusion; semi-recumbent position targeted at 30–45°.
  • Intervention:
    • No GRV monitoring strategy: GRV was not measured.
    • Definition of intolerance: Vomiting or regurgitation (macroscopic gastric contents in the mouth or on the endotracheal tube cuff).
    • Response to intolerance: Stop EN for 6 hours; aspirate gastric contents; restart EN at half the previous rate and advance by 25 mL/h every 6 hours if tolerated.
  • Comparison:
    • Routine GRV monitoring strategy: GRV measured every 6 hours using a 60 mL syringe aspiration (patient supine); aspirated content reinjected.
    • Definition of intolerance: GRV >250 mL at any measurement and/or vomiting/regurgitation.
    • Response to intolerance: Stop EN for 6 hours; aspirate gastric contents; restart EN at half the previous rate and advance by 25 mL/h every 6 hours if tolerated.
  • Blinding: Open-label at bedside (nursing/clinicians aware of allocation); outcome adjudication committee for VAP and ICU-acquired infections was blinded to group assignment.
  • Statistics: Power calculation assumed 19% VAP incidence in the control group; noninferiority margin 10%; one-sided α=0.05; 80% power → 382 patients (191/group) required; inflated to 420 to allow ~10% with ventilation <48 hours. Primary analysis used a modified intention-to-treat population (randomised and consent retained) plus a per-protocol analysis for noninferiority.
  • Follow-Up Period: VAP surveillance to 2 days after extubation (up to 90 days); ICU-acquired infections to ICU discharge; mortality assessed at day 28 and day 90.

Key Results

This trial was not stopped early. Follow-up for VAP was planned to 90 days; 452 patients were randomised and 449 were included in the modified intention-to-treat analysis.

Outcome No GRV monitoring (n=227) Routine GRV monitoring (n=222) Effect p value / 95% CI Notes
Ventilator-associated pneumonia within 90 days (primary; modified ITT) 38 (16.7%) 35 (15.8%) Risk difference 0.9% 90% CI −4.8% to 6.7%; one-sided P (noninferiority)=0.007 Noninferiority margin 10%; upper CI bound <10% supported noninferiority
Ventilator-associated pneumonia within 90 days (per protocol) 35/208 (16.8%) 31/215 (14.4%) Risk difference 2.4% 90% CI −3.7% to 8.6%; one-sided P (noninferiority)=0.02 Per-protocol population n=423
Vomiting (any) 90 (39.7%) 60 (27.0%) Risk difference 12.7% 90% CI 5.4% to 20.0%; P=0.01 Higher vomiting despite similar prokinetic use
Patients receiving 100% of calorie target during first week 211 (92.9%) 171 (77.2%) Risk difference 15.7% 90% CI 10.1% to 21.4%; P<0.001 Process/separation measure; target 20–25 kcal/kg/day during week 1
Cumulative calorie deficit during first week Median 94 kcal (IQR 44–176) Median 191 kcal (IQR 66–299) Difference −78 kcal 90% CI −144 to −9; P=0.04 Reported as between-group difference in median
ICU-acquired infection (any) 60 (26.4%) 60 (27.0%) Risk difference −0.5% 90% CI −7.7% to 6.8%; P=0.92 Adjudicated by blinded committee
Duration of mechanical ventilation Median 7.7 days (IQR 3.9–13.8) Median 7.1 days (IQR 3.9–14.3) Difference 0.3 days 90% CI −0.8 to 1.5; P=0.66 Not different
ICU length of stay Median 13.1 days (IQR 7.5–22.6) Median 13.8 days (IQR 8.4–23.0) Difference 0.0 days 90% CI −1.6 to 1.6; P=0.97 Not different
Mortality (day 28) 59 (25.7%) 56 (25.3%) Risk difference −0.4% 90% CI −6.9% to 6.1%; P=0.92 Not powered for small mortality differences
Mortality (day 90) 78 (34.4%) 72 (32.4%) Risk difference 2.0% 90% CI −5.3% to 9.3%; P=0.65 Not different
  • No-GRV monitoring met the pre-specified noninferiority criterion for VAP (risk difference 0.9%; 90% CI −4.8% to 6.7% vs a 10% margin), with concordant per-protocol findings.
  • Omitting GRV monitoring improved nutrition delivery (100% calorie target in week 1: 92.9% vs 77.2%) and reduced calorie deficit (median 94 vs 191 kcal).
  • Vomiting was more frequent without GRV monitoring (39.7% vs 27.0%) despite similar prokinetic use (51.1% vs 50.9%).

Internal Validity

  • Randomisation and allocation concealment: Centralised web-based randomisation; stratified by centre; permuted blocks (size 4); allocation concealed until assignment.
  • Post-randomisation exclusions / drop-out:
    • Randomised: 452.
    • Modified intention-to-treat: 449 (3 withdrew initial consent post-randomisation and were not analysed).
    • Per-protocol: 423 (excluded 7/222 in routine-monitoring group and 19/227 in no-monitoring group for protocol deviations, including ventilation <48 hours and other violations).
  • Performance and detection bias:
    • Open-label design could influence bedside behaviours (e.g., feeding interruptions, thresholds for antiemetics/prokinetics, microbiological sampling).
    • VAP and ICU-acquired infections were adjudicated by an independent committee blinded to allocation, mitigating detection bias for key outcomes.
  • Protocol adherence and separation of the variable of interest:
    • Calorie delivery separation: 100% target in first week 92.9% (no GRV) vs 77.2% (routine GRV).
    • Calorie deficit separation: median 94 vs 191 kcal in week 1.
    • Mechanism-consistent harm signal: vomiting 39.7% vs 27.0%.
  • Baseline comparability: Broadly similar severity and case-mix (age 61.2 vs 62.4 years; SAPS II 49.5 vs 50.8; SOFA 8.0 vs 7.9; vasoactive drug use 69.8% vs 71.4%).
  • Timing and dose of intervention:
    • Early EN mandated (within 36 hours of intubation), and the monitoring strategy applied from initiation.
    • Energy targets (20–25 kcal/kg/day week 1; 25–30 thereafter) and protein targets (1.0–1.2 g/kg/day) were protocolised and identical across groups.
  • Outcome assessment: VAP definition required radiographic criteria plus clinical signs and quantitative microbiology thresholds; although adjudicated, VAP remains sensitive to sampling and imaging practices.
  • Statistical rigour: Noninferiority margin and analytic approach pre-specified; both modified ITT and per-protocol analyses reported; noninferiority assessed using 90% CIs consistent with one-sided α=0.05.

Conclusion on Internal Validity: Overall, internal validity appears moderate-to-strong given concealed randomisation, blinded adjudication for primary outcomes, and clear separation in the exposure (GRV monitoring), tempered by open-label care and asymmetric per-protocol exclusions inherent to noninferiority designs.

External Validity

  • Population representativeness: Adult ICU patients receiving early gastric EN and expected prolonged ventilation; exclusions removed patients with contraindications to EN or very early death/extubation, under-representing some high GI-risk and peri-operative cohorts.
  • Care context: Nine French ICUs with protocolised EN delivery and semi-recumbent positioning; centres did not use specialised endotracheal tubes with subglottic secretion drainage.
  • Applicability: Findings most applicable to ICUs delivering early gastric EN via continuous infusion with standard aspiration precautions; generalisability is less certain in settings with different VAP bundles, in patients with severe gastrointestinal dysmotility, or where vomiting carries disproportionate risk (e.g., prone ventilation) unless mitigation strategies are in place.

Conclusion on External Validity: Generalisability is good for typical mechanically ventilated adults receiving early gastric EN in resourced ICUs, but is limited for populations intentionally excluded and for settings with materially different VAP prevention and feeding practices.

Strengths & Limitations

  • Strengths:
    • Pragmatic multicentre randomised design in real-world ICUs.
    • Clinically meaningful primary endpoint (VAP) with blinded independent adjudication.
    • Pre-specified noninferiority framework with both modified ITT and per-protocol analyses.
    • Clear process separation in delivered calories, supporting mechanistic plausibility.
  • Limitations:
    • Open-label bedside care introduces potential performance bias (feeding decisions, sampling intensity, aspiration precautions).
    • Noninferiority margin (10% absolute risk) may be viewed as clinically large for a safety endpoint.
    • VAP diagnosis, while adjudicated, remains vulnerable to local sampling and radiographic interpretation variability.
    • Higher vomiting without GRV monitoring highlights potential trade-offs; rare catastrophic aspiration events may be underpowered.
    • French-only ICUs and key exclusions constrain generalisability to some surgical/high GI-risk populations.

Interpretation & Why It Matters

  • Practice implication
    • Routine GRV monitoring can be omitted in many ventilated ICU patients receiving early gastric EN without an observed increase in VAP, while improving calorie delivery.
    • Adopting a no-GRV strategy requires proactive vomiting/regurgitation mitigation (positioning, airway protection, early response protocols).
  • Conceptual shift
    • Supports a tolerance paradigm based on clinical intolerance rather than GRV numbers, aligning EN delivery with avoidance of unnecessary interruptions.

Controversies & Subsequent Evidence

  • Editorial framing: The accompanying editorial interpreted the findings as supporting discontinuation of routine GRV monitoring in many ventilated adults, while emphasising that regurgitation/vomiting and aspiration prevention strategies remain central when adopting a no-GRV protocol.1
  • Correspondence on generalisability and safety trade-offs: A subsequent letter questioned generalisability (exclusions and limited reporting of haemodynamic support details) and highlighted the increased vomiting signal; the authors replied that GRV lacks validation as an aspiration/VAP surrogate and that VAP pathogenesis is not well captured by GRV thresholds.23
  • Subsequent syntheses: Systematic reviews/meta-analyses (including Cochrane) generally report no clear increase in VAP with omission of GRV monitoring and improved nutrition delivery, while noting possible increases in vomiting/regurgitation and low certainty for uncommon aspiration-related harms.456
  • Guideline uptake: Post-2013 critical care nutrition guidelines increasingly recommend avoiding routine GRV monitoring (or, if measured, using higher thresholds alongside clinical intolerance assessment) to reduce unnecessary feed interruption and underfeeding.78910

Summary

  • In 449 mechanically ventilated adults receiving early gastric EN, omitting routine GRV monitoring was non-inferior to routine monitoring for VAP (16.7% vs 15.8%; risk difference 0.9%; 90% CI −4.8% to 6.7%).
  • No-GRV monitoring improved nutrition delivery (100% calorie target in week 1: 92.9% vs 77.2%) and reduced calorie deficit (median 94 vs 191 kcal).
  • Vomiting was more frequent without GRV monitoring (39.7% vs 27.0%), highlighting an important trade-off requiring mitigation.
  • Other outcomes (ICU infections, ventilation duration, ICU/hospital length of stay, day 28 and day 90 mortality) were not different between groups.
  • The trial contributed to a paradigm shift away from routine GRV-based stopping rules in ventilated ICU patients receiving gastric EN.

Overall Takeaway

NUTRIREA-1 reframed “feeding intolerance” in ventilated adults by showing that routine GRV monitoring is not required to prevent VAP, while improving delivered calories. The benefit is accompanied by increased vomiting, so implementation should emphasise aspiration precautions and active management of regurgitation. The trial helped drive a guideline-supported move towards clinically driven EN tolerance assessment rather than GRV-driven interruption.

Overall Summary

  • Noninferior VAP with no GRV monitoring (16.7% vs 15.8%; risk difference 0.9%; 90% CI −4.8% to 6.7%).
  • Improved nutrition delivery but more vomiting (39.7% vs 27.0%).
  • Supports abandoning routine GRV checks in many ventilated adults, with attention to aspiration mitigation.

Bibliography