Criteria abstracted from The Users' Guide to Medical Literature, from the Health Information Research Unit and Clinical Epidemiology and Biostatistics, McMaster University

Highlighted lines and questions below provide links to the pertinent description of criteria in The EBM User's Guide, now available at the Canadian Centres for Health Evidence

Article Reviewed:

Please visit the new Evidence Based Journal Club Reviews

Gastric vs Small-Bowel Feeding in Critically Ill Children Receiving Mechanical Ventilation: A Randomized Controlled Trial.

Meert KL, Daphtary KM, Metheny NA.

Chest 2004;126 872-878. [abstract]

Reviewed by Deyin D. Hsing MD and Andranik Madikians MD, Mattel Children's Hospital, University of California, Los Angeles, California

Review posted January 21, 2005

---

I. What is being studied?:

The study objective:

To evaluate the effect of feeding tube position, gastric vs. transpyloric (small-bowel), on: a) nutrient delivery; b) feeding complications, such as microaspiration , in critically ill children receiving mechanical ventilation

The study design:

Randomized unblinded controlled clinical trial in a pediatric ICU (PICU) with the possibility of cross over (but NOT a cross over study).

The patients included:

All PICU patients of the children's Hospital of Michigan, who are > 18 years of age, receiving mechanical ventilation and who were to begin enteral tube feeing. Informed consent was obtained from the parent or legal guardian of all participants.

The patients excluded:

Those PICU patients, with 1) overt GI bleeding; 2) anatomic obstruction of the GI tract; 3) recent esophageal or GI surgery, and 4) presence of a percutaneous feeding tube.

The interventions compared:

Enteral tube feeding in mechanically ventilated patients, either via gastric or small-bowel route.

The outcomes evaluated:

The primary outcome measure was nutrient delivery in the form of percentage of daily caloric goal achieved during tube feeding. A secondary outcome measure was the incidence of feeding complications, especially microaspiration, along with other complications such as vomiting, diarrhea, and abdominal distension.

II. Are the results of the study valid?

Primary questions:

1. Was the assignment of patients to treatments randomized?

Yes. The assignment of patients to treatments were randomized via blocked randomization in blocks of 12 with a ratio of five gastric to seven small-bowel patients, stratified initially for the presence of either diagnosed or treated gastroesophageal reflux disease (GERD). The randomization was unbalanced, and the failure rate, which they used in the study as their justification for this unbalance, was based on the 70% success rate used in one of the authors' previous study (1). However, I do not think that this unbalance is necessarily justified. Overall, the success rate of bedside placement of transpyloric feeding tube ranges between 66-93% according to available literature (2). The article indicates that the feeding tubes were placed by one of the investigators using "a previously described standard technique" in the study protocol. The original article describing this referenced technique published by Zaloga et al. in 1991, and the original study demonstrated success rate up to 92% (3). It is clear that the wide range of success rate for bedside placement of transpyloric tube is operator-dependent.

2. Were all patients who entered the trial properly accounted for and attributed at its conclusion?

Was followup complete?

Yes. Follow up was complete in the sixty-two patients, who received the clinical intervention to be studied; no specifics were given for the 12 patients who exited the study prior to clinical intervention due to difficulties in placement of small-bowel feeding tube.

Were patients analyzed in the groups to which they were randomized?

Yes, for those patients (n=32), who were randomized to the gastric feeding tube group, were analyzed in the group to which they were randomized. 15 of the NG patients had tubes clogged or dislodged and these were all replaced.

All patients who entered the trial, (n=74), where accounted for at its conclusion. Data from only sixty-two of those patients were analyzed because placement of a small-bowel feeding tube at the bedside was unsuccessful and these twelve patients exited the study. Intention to treat analysis was used to account for cross over patients (n=5), who were initially in the small-bowel group, but was later continued on gastric feeding till the end of the clinical trial due to inability to replace a dislodged small bowel tube (n=12, patients had dislodged feeding NJ tube; 15 ng patients had tubes clogged or dislodged and these were all replaced).

Secondary questions:

3. Were patients, health workers, and study personnel "blind" to treatment?

No. This clinical trial was not blinded, the patients, and/or their families, and study clinicians were aware of the group assignment; the only blinded personnel were: 1) The pharmacist who was responsible for stratified blocked randomization for the presence of GERD prior to initiation of the trial; 2) the laboratory personnel, who were responsible for assess tracheal aspirate pepsin and serum prealbumin concentration.

4. Were the groups similar at the start of the trial?

The groups were similar in their baseline characteristics at the start of the trial. Those characteristics include: age, gender, GERD, cuffed endotracheal tube, number of patients with infiltrates on CXR, PRISM score at PICU admission, weight, prealbumin, intensive care days at randomization, and days fasted prior to the start of the trial.

5. Aside from the experimental intervention, were the groups treated equally?

Aside from the experimental intervention, the groups were treated comparably in terms of total number of days (n=121 days for both groups) when tube feedings where temporarily interrupted for at least 1h.

However, the feeding regimen was not standardized and equal for the two groups. First, non-blinded bedside staffs controlled feeding advancement rates in this study. Second, the duration for which tube feeds were held once interrupted might have been longer in the gastric group than the small-bowel group A longer time holding feeds prior to a procedure in the gastric group could easily have made a significant contribution to their study findings. Further more, the reasons for interruption of the tube feedings in the two groups seemed to demonstrate more noticeable differences especially in numbers:

1. Medication administration and increased residual volume appeared to be more prevalent reasons for interruption of the feedings in the gastric group (n=13 respectively) vs. the small bowel group (n=0);
2. More patients in the small-bowel group (n=18) seemed to have received pharmacological paralysis during the trial as a reason of interruption of tube feedings than those in the gastric group (n=7); (This difference, however, may be due to the patients' underlying illness since the authors indicated that paralysis was used either due to patient-ventilator dyssynchrony or increased intracranial pressure.)
3. More patients in the small-bowel group (n=13) had interrupted tube feeding because of worsening underlying illness vs. the gastric group (n=7)

The statistical significance between these possibly confounding variables was not analyzed.

III. What were the results?

1. How large was the treatment effect?

The primary outcome measure showed that the average percentage of daily caloric goal achieved was less in the gastric group compared to the small-bowel group (mean ± s.d.: 30 ± 23% vs. 47 ± 22%, p < 0.01). A difference of 17% between the two groups was demonstrated based on these results. A confidence interval was not provided in this study. However, using standard calculation of confidence interval (CI) for differences (10), a 95% confidence interval for the point estimate difference of 17% of 5.8 to 28.2 can be derived.

The secondary outcome measures showed:

1. No differences were found in the proportion of tracheal aspirates positive for pepsin between the gastric and small-bowel group (50 of 146 aspirates vs. 50 of 172 aspirates, respectively; p=0.3) using intent to treat analysis.

2. In other words, 34% of the tracheal aspirates were positive for pepsin in the gastric group compared to 29% in the small-bowel group (intent-to-treat analysis), for a relative risk of having a positive tracheal aspirate 0.85 in the small-bowel group, and an absolute risk reduction (ARR) of 0.05 in the small-bowel group vs. gastric group (4). The calculated Odds Ratio (OR) of positive tracheal aspirate with small-bowel group compared with gastric group is 0.8.
3. No differences were found in the frequency of feeding tube displacement, abdominal distension, vomiting or diarrhea between groups.
4. Overall, patients in the small-bowel group seemed to have longer total ventilator days (15.3 ± 19.1 vs. 8 ± 4.5, p < 0.2), longer total intensive care days (17.8 ± 18.4 vs. 10.7 ± 4.7, p < 0.2), and longer total hospital days (30.4 ± 34 vs. 18.2 ± 9.7, p < 0.1) than the gastric group. However, none of these differences were statistically significant.

2. How precise was the estimate of the treatment effect?

The study demonstrates an increased average daily caloric delivery of 17% in the small-bowel feeding group with a calculated 95% CI range of 5.8 to 28.2. This wide confidence interval indicates that either the study has not enrolled sufficient number of patients to demonstrate a more precise estimate of the difference of daily caloric delivery between the two groups, or that there really is a great amount of inherent variability in this measurement.

For the secondary outcome measures such as microaspiration, an absolute risk reduction (ARR) of 0.05 (95% CI: -0.05 - +0.15) is not considered clinically significant. Furthermore, the 95% confidence intervals around the relative risk reduction (RRR) of 15% are Ð18% and 39%. These are broad intervals and this interval crosses zero, which represents no treatment (small-bowel feeding) effect on risk reduction of microaspiration.

This study was designed to look at caloric intake, not clinically significant aspiration risk. The tracheal aspirates are only proxy measures for the risk of microaspiration. This study does not have enough patients enrolled to demonstrate a clinically significant difference in aspiration rates. The authors did not specify the amount of difference in aspiration rates, which they may consider clinically significant. The study's sample size was based on the primary outcome of interest, difference in percentage of daily caloric delivery.

IV. Will the results help me in caring for my patients?

1. Can the results be applied to my patient care?

The patients involved in this study were similar in certain characteristics to our patient population, which also consists of critically ill pediatric patients in a university teaching hospital. The most common primary diagnosis of our patient population is also comparable to the study population. However, we also have a substantial transplant and congenital heart disease population, which were not specifically mentioned in this study. One should therefore still exercise caution in applying the study results to our own patient population.

2. Were all clinically important outcomes considered?

In addition to the average percentage of daily caloric goal achieved and the proportion of tracheal aspiration positive for pepsin, the total number of tube feeding days, the change in body weight per day, the change in prealbumin per day, daily net fluid balances/BSA, frequency of vomiting/diarrhea/abdominal distension, as well as total number of ventilator days/intensive care days/hospital care days were also evaluated. Some of the following clinical outcomes, which have been examined in similar adult studies (5-8), may also be worth further investigation:

1. Time until feeding started from initiation of placement attempt.

   Neumann el al (8), found that patients fed in the stomach received nutrition sooner and with fewer tube placement attempts than patients fed in the small bowel, in their 2002 prospective, randomized study of adult ICU patients.

2. Overall nutrient delivery assessment, for the entire duration of tube feeding, in terms of average total calories/ kg/day and total protein delivered/kg/day.

   Kearns et al (6), in a prospective, randomized, controlled trial of intubated adult medical ICU patients, with blinded bedside nurses, demonstrated that small-bowel feeding achieved higher total protein and caloric intake in addition to the increased percentage of calories delivered.

3. The effect (if any) of high gastric residual on total nutrient delivery in the two treatment groups.

   Montejo et al (7), found high gastric residual as one of the statistically important gastrointestinal complications of gastric feeding, in their 2002 multicenter, prospective, randomized, single-blind study of adult ICU patients receiving early (within the first 36 hours of admission) using a double lumen nasogastrojejunal tube.

3. Are the likely treatment benefits worth the potential harms and costs?

This study was able to demonstrate statistically significant increased average percentage of daily caloric delivery via the small-bowel feeding route as one of the treatment benefits. However, placement of a gastric tube at the bedside is much easier than placing a transpyloric feeding tube, and this was confirmed in the present study. Indeed, their results might have been different if they used a true intention-to-treat analysis, considering all those initially randomized to the transpyloric route as the small bowel group, instead of just those with successful initial placement. In addition, tube-related complication such as occlusion is more likely to occur in patients receiving small-bowel feeding (7).

In adult studies, fluoroscopic guidance is often used for difficult transpyloric tube insertions. This procedure not only is more invasive but also entails radiation exposure, which is a potential harm perhaps not justified in the pediatric population. Therefore, it may be in our patients' best interest to receive gastric feeding, especially when most of the studies to date fail to demonstrate increased risk of aspiration with gastric tube feeding. Furthermore, early enteral nutrition administration have been shown to decrease the number of infectious complications, length of stay and mortality in critically ill patients (7). Neumann et al (8) found that patients in the gastric group received nutrition sooner with fewer tube placement attempts. The potential harms and costs of utilizing small-bowel feeding vs. gastric feeding were not closely examined in this study.

In conclusion, we have concerns regarding the validity of this study because of the following reasons:

1. We do not feel that the unbalance in patient randomization was entirely justified;
2. The study did not use a true "Intent-to-treat" analysis considering the 5 patients from the small bowel group and the 12 patients who initially exited the study because of failure to place NJ tube;
3. The study was not a blinded clinical trial. As a result, aside from experimental interventions, there were other confounding variables not well controlled, i.e. the interventions such as feeding regimen, feeding advancement as well as duration of withholding of feeds were not standardized in the study, and were instead carried out by non-blinded bedside staff.

However, we do agree with the author's conclusion that even though small-bowel feeds enables more daily nutrient delivery, gastric feeding offers the advantage of easy placement and earlier feeding, without the added risk of aspiration, with the caveat noted above about the relative weakness of this safety claim.

References:

1. Gharpure V, Meert KL, Sarnaik AP. Efficacy of erythromycin for postpyloric placement of feeding tubes in critically ill children: a randomized, double-blind, placebo controlled study. JPEN J Parenter Enteral Nutr. May-Jun 2001;25(3):160-165. [abstract]
2. Da Silva PS, Paulo CS, de Oliveira Iglesias SB, de Carvalho WB, Santana e Meneses F. Bedside transpyloric tube placement in the pediatric intensive care unit: a modified insufflation air technique. Intensive Care Med. Jul 2002;28(7):943-946. [abstract]
3. Zaloga GP. Bedside method for placing small bowel feeding tubes in critically ill patients. A prospective study. Chest. Dec 1991;100(6):1643-1646. [abstract]
4. Guyatt GH, Sackett DL, Cook DJ. Users' guides to the medical literature. II. How to use an article about therapy or prevention. B. What were the results and will they help me in caring for my patients? Evidence-Based Medicine Working Group. JAMA. Jan 5 1994;271(1):59-63. [full-text]
5. Heyland DK, Drover JW, MacDonald S, Novak F, Lam M. Effect of postpyloric feeding on gastroesophageal regurgitation and pulmonary microaspiration: results of a randomized controlled trial. Crit Care Med. Aug 2001;29(8):1495-1501. [abstract]
6. Kearns PJ, Chin D, Mueller L, Wallace K, Jensen WA, Kirsch CM. The incidence of ventilator-associated pneumonia and success in nutrient delivery with gastric versus small intestinal feeding: a randomized clinical trial. Crit Care Med. Jun 2000;28(6):1742-1746. [abstract]
7. Montejo JC, Grau T, Acosta J, et al. Multicenter, prospective, randomized, single-blind study comparing the efficacy and gastrointestinal complications of early jejunal feeding with early gastric feeding in critically ill patients. Crit Care Med. Apr 2002;30(4):796-800. [abstract]
8. Neumann DA, DeLegge MH. Gastric versus small-bowel tube feeding in the intensive care unit: a prospective comparison of efficacy. Crit Care Med. Jul 2002;30(7):1436-1438. [abstract]
9. Guyatt GH, Sackett DL, Cook DJ. Users' guides to the medical literature. II. How to use an article about therapy or prevention. A. Are the results of the study valid? Evidence-Based Medicine Working Group. JAMA. Dec 1 1993;270(21):2598-2601. [full-text]
10. Confidence interval for differences in means at http://graphpad.com/quickcalcs/CatMenu.cfm

Comments

Back to the EB Journal Club Index