Quick Links

Recommendations Summary

CI: Gastric versus Small Bowel Feeding Tube Placement 2006

Click here to see the explanation of recommendation ratings (Strong, Fair, Weak, Consensus, Insufficient Evidence) and labels (Imperative or Conditional). To see more detail on the evidence from which the following recommendations were drawn, use the hyperlinks in the Supporting Evidence Section below.

  • Recommendation(s)

    CI: Feeding Tube Placement

    Enteral Nutrition (EN) administered into the stomach is acceptable for most critically ill patients. Consider placing feeding tube in the small bowel when patient is in supine position or under heavy sedation. If your institution's policy is to measure gastric residual volume (GRV), then consider small bowel tube feeding placement in patients who have more than 250ml GRV or formula reflux in two consecutive measures. Small bowel tube placement is associated with reduced GRV. Adequately-powered studies have not been conducted to evaluate the impact of GRV on aspiration pneumonia. There may be specific disease states or conditions that may warrant small bowel tube placement (e.g., fistulas, pancreatitis, gastroporesis), however they were not evaluated at this phase of the analysis.

    Rating: Fair

    • Risks/Harms of Implementing This Recommendation

      • Underfeeding may result if the small bowel tube placement is delayed
      • Excess GRV or reflux of EN may be associated with feeding intolerance.

    • Conditions of Application

      • Varying interpretations of the impact and importance of GRV may be a barrier to implementing this recommendation
      • The placement of feeding tubes in the post-pyloric position requires skilled practitioners (hospitals vary in who places tubes from medical or nursing to dietetics personnel)
      • Some but not all protocols use fluoroscopy to aid in tube placement, while others use promotility agents or magnets and radiographic confirmation of tube tip position.

    • Potential Costs Associated with Application

      The impact of feeding tube placement on cost of medical care has not been adequately evaluated.

    • Recommendation Narrative


      • The impact of feeding tube placement on mortality has not been adequately evaluated in large trials.
      • Mortality was not a primary outcome variable for any of these four negative-quality RCTs (Kearns et al, 2000; Boivin et al, 2001; Esparza et al, 2001; Davies et al, 2002) and none did a power analysis on mortality
        • The mortality reported was not different by whether the feeding tube was in the gastric or postpyloric position and mortality was generally low, but ranged from 11% (Davies et al, 2002) to 41% (Esparza et al, 2001)
        • These studies may have been underpowered to detect a significant difference in mortality, if a difference existed.
      • A negative-quality RCT with only 80 patients did not detect a significant difference in mortality (Kortbeek et al, 1999)
      • A positive-quality meta-analysis (Heyland et al, 2002) reported no difference in mortality by tube tip placement
      • While power analysis of number of subjects needed to detect a difference in mortality were not found in the literature, aspiration pneumonia occurs more frequently than mortality, thus mortality outcomes will require larger groups of patients than pneumonia
      • Since no study provided a power analysis on mortality, it is not clear how large a sample size is needed.

      Infectious Complications (Aspiration Pneumonia)

      • One positive quality RCT (Heyland et al, 2001), one neutral-quality RCT (Montejo et al, 2002), one negative quality RCT (Davies et al, 2002) and one positive quality meta-analysis (Marik and Zaloga, 2003) used 150ml GRV as a threshold measure to hold enteral feedings
      • One positive quality RCT (Heyland et al, 2001) showed no difference in aspiration pneumonia, but was underpowered
        • An interim statistical analysis that was conducted after 39 patients suggested a need for more than 200 patients to detect a significant difference if it existed.
      • One neutral quality RCT (Montejo et al,  2002) suggested that 152 patients were needed to detect a 50% reduction in aspiration pneumonia and one positive quality RCT (Drakulevic et al,  1999) calculated that 184 patients were needed to detect a 50% reduction in pneumonia with 80% power and P<0.05
        • Most of the studies reviewed included far less than 100 patients. 
      • Two positive quality meta-analyses attempted to combine data from a series of small trials to evaluate the risk of aspiration pneumonia 
        • Heyland et al, 2002, detected a reduction in ventilator-associated pneumonia (odds=0.76, CI=0.59-0.99)
        • In a subsequent meta-analysis by the same author (Heyland et al, 2003), where one study was removed, the RR of infections (pneumonia) with small bowel feeding tube placement was only 0.83, P=0.30, and no longer statistically significant.
      • A negative quality RCT (Esparza et al, 2001) reported no difference in aspiration of isotope provided as a component of EN, regardless of tube tip position, however only 54 patients were studied and not enough to detect a 20% reduction in aspiration from the low 10% frequency.
      • A positive quality meta-analysis (Marik and Zaloga, 2003) included nine trials and found no difference in pneumonia
      • A positive quality consensus statement (McClave et al, 2002) in the North American Summit on Aspiration in the Critically Ill Patient, gave expert opinion that small bowel feeding would reduce gastroesophageal reflux and possibly aspiration risk
      • Six other studies found no significant difference in aspiration pneumonia, though all were likely underpowered (Montecalvo et al, 1992; Kortbeek et al, 1999; Kearns et al, 2000; Davies et al, 2002; Montejo et al, 2002; Neumann and DeLegge, 2002).


      • In the Canadian Clinical Practice Guidelines, a positive quality evidence analysis (Heyland et al, 2003), they suggest small bowel feeding tube placement for all ICU patients when feasible and particularly for high-risk patients (those with high GRV, sedation, supine positioning)
        • They recognize that small bowel feeding tube placement may not be feasible for all patients
        • Clearly, a large RCT is needed to answer the question of whether pneumonia is reduced by small bowel feeding tube placement.
      • A positive quality consensus statement (McClave et al, 2002) also includes the following as risks of aspiration pneumonia that should be considered: 
        • Major risk factors for aspiration include documented previous episode of aspiration, decreased level of consciousness, neuromuscular disease and structural abnormalities of the aerodigestive tract, endotracheal intubation, vomiting, persistently high GRVs and need for prolonged supine position
        • Additional risk factors include presence of nasoenteric tube, non-continuous or intermittent feeding, abdominal/thoracic surgery or trauma, delayed gastric emptying, poor oral care, age, inadequate nursing staff, large size or diameter of feeding tube, malpositioned feeding tube, transporting patient.

      Length of Hospital Stay

      • The impact of feeding tube placement on LOS has not been adequately evaluated in large trials: None looked at LOS as the primary outcome and none calculated a power analysis
        • ICU and LOS were not significantly different in two negative quality RCTs (Kearns et al, 2000; Boivin et al, 2001) by feeding tube placement
          • The ICU LOS was 16 days for NG and 17 days for ND tube placement and LOS was 43 days (NG) vs. 39 days (ND) (Kearns et al, 2000) and not statistically different.
        • A negative quality RCT (Kortbeek et al, 1999) found no difference in ICU LOS by tube position (seven vs. 10 days) and neither did the positive quality meta-analysis by Marik and Zaloga, 2003, where LOS differed by 1.4 days.
      • These studies may have been underpowered to detect a significant difference in LOS related to gastric vs. small bowel feeding tube placement, if a difference existed.


      • The impact of feeding tube placement on cost of medical care has not been adequately evaluated.

    • Recommendation Strength Rationale

      • Mortality is based on lack of a power calculation and use of mortality only as a secondary variable. There is insufficient evidence
        • Conclusion statement is Grade V.
      • Usefulness is based on two review articles 
        • Conclusion statement is Grade IV.
      • Infectious complications (aspiration pneumonia) is based on inadequately-powered trials and insufficient evidence
        • Conclusion statement is Grade III. 
      • LOS is based on limited and possibly underpowered studies 
        • Conclusion statement is Grade V.
      • No studies were found evaluating the impact of cost of medical care 
        • Conclusion statement is Grade V.