Enteral nutrition (EN) is the preferred route for nutritional support of critically ill patients in intensive care units (ICUs), in whom maintaining adequate nutrition is challenging. This is usually caused by increased nutrient and energy needs resulting from catabolic stress (1). Negative energy balances are frequent and generally associated with a high occurrence of complications and poor clinical outcomes (2). To meet the nutritional requirements of patients in the ICU, it is necessary to implement early initiation of an EN schedule using enteral formulae that are well tolerated.
Complications associated with EN by tube feeding are not uncommon and can reduce the delivery of nutritional requirements to patients in the ICU; therefore, algorithms used in EN protocols to begin and monitor the progress to the caloric target need to be adapted to the pathophysiological status of the gastrointestinal (GI) tract. Alterations in GI motility as well as digestive and absorptive functions can also limit tolerance to tube feeding (3).
EN was reported to help maintain the mucosal barrier integrity (4), thereby preventing the dissemination of bacteria from the intestinal lumen into the internal milieu (5,6) and reducing septic complications (7). EN has been associated with alterations in the intestinal luminal environment, disturbing the resident microbiota and favouring overgrowth of pathogens, for example, Clostridium difficile(8). These alterations may influence the colonic secretory/absorptive functions, reducing the absorption of water and electrolytes in the colonic lumen. Supplementation of EN formula with fructooligosaccharides (FOS) may prevent such adverse changes (9). The metabolic activity of the luminal microbiota can also be upset during tube feeding, affecting colonisation resistance and contributing to the development of diarrhoea considered to be a major marker of intolerance to EN. Current clinical practice indicates that nutritional support must be adapted or reduced upon diarrhoeal episodes, and this results in delays in the progression to the caloric goal and energy deficit during hospitalisation. Consequently, EN formulations that have a positive effect on gut ecology and intestinal function and also provide appropriate nutritional support of patients in the ICU are of major interest.
A potential solution to manage intestinal microbial disturbances is to provide healthy live microorganisms, known as probiotics, to modulate the gut ecosystem (10,11), in combination with functional carbohydrates, called prebiotics, to selectively promote the growth and metabolic activity of beneficial bacteria, for example, bifidobacteria (12).
To meet the nutritional needs of hospitalised children, a new formula 50:50 casein:whey supplemented with a combination of defined prebiotics and probiotics—also called synbiotics (13)—and docosahexaenoic acid (DHA) was developed. The blend of probiotics contained 2 probiotic strains (Lactobacillus paracasei NCC 2461 and Bifidobacterium longum NCC 3001), and the blend of prebiotics was a mix of FOS, inulin, and Acacia gum. Such a combination of pre- and probiotics is more likely to be effective in regulating the luminal microbial environment all along the gut and providing complementary benefits on the gut physiology.
The aim of the present study was to demonstrate the tolerance and safety of an enteral formula containing a synbiotic blend and to investigate its effect on the intestinal microbiota in critically ill children receiving antibiotics.
SUBJECTS AND METHODS
The study was a controlled, double-blind, randomised, clinical trial of 2 parallel groups conducted in 2 medical centres. The study was performed in the Department of Paediatrics at the Maharat Nakhon Ratchasima Hospital and in the Department of Paediatrics at the Khon Kaen Hospital in Thailand. The protocol was approved by the institutional review board, Maharat Nakhon Ratchasima Hospital (December 28, 2005), and by the Ethical Review Committee for Research in Human Subjects, Ministry of Public Health, Thailand (number 38/2006, May 4, 2006), and conducted according to the principles stated in the Declaration of Helsinki and its amendments.
Critically ill patients between 1 and 3 years old who have obtained their parents’/carers’ written consent were eligible for recruitment. Patients were recruited between August 2006 and May 2009 from the paediatric intensive care unit (PICU) of the 2 clinical centres. Inclusion criteria were the following: patients experiencing pneumonia, neurological, or cardiovascular disease, in need of mechanical ventilation, and EN. Patients presenting 1 or more of the following situations were not eligible for the study: overt GI bleeding, anatomic obstructions of the GI, recent oesophageal or GI surgery, immunodeficiency, or other GI disorders that would affect enteral feeding (eg, malabsorption, intestinal dysmotility). Patients participating in another clinical trial were not included. Any medication required for the normal therapeutic management of the subjects was allowed during the trial and recorded in case report form, which is part of the study documents.
Products were provided in metallic tins and distinguished by 2 different colours and letter labels: A, yellow and B, blue. The identity of the products was blind to the subjects, support staff, and investigators.
Products, Dosage, and Administration
The patients were assigned to 1 of the 2 treatment groups: (1) test: enteral formula supplemented with 2 probiotic strains (L paracasei NCC 2461 [5 × 106 CFU/g], B longum BB536 from Morinaga coded at the Nestlé Culture Collection to be NCC 3001 [2 × 106 CFU/g]), prebiotics (oligofructose/inulin [2.6 g/L], Acacia gum [2.8 g/L]), and DHA [43 mg/L]; or (2) control: same enteral formula without added pre- and probiotics or DHA. Randomisation was performed in blocks for each centre to avoid bias for clinical practice. Paediatric Risk of Mortality (PRISM) scores (14) were used for randomisation with a stratification cutoff set at <8 and ≥8. Sealed envelopes with the codes were sent to each centre.
Both products were isocaloric and isoproteic and consisted of proteins, carbohydrates, fats, vitamins, and minerals in amounts intended for the full nutritional support of critically ill patients. Each product delivered 100 kcal/100 mL. The composition of the products is provided in Table 1. Both formulae were prepared in a dedicated area to avoid contamination. Each tin containing 153 g of powder was reconstituted in 600 mL of preboiled filtered water, making a total of 710 mL of solution. Then, reconstituted formulae were divided in aliquots corresponding to the volumes of bolus to be administered during the day and kept in the refrigerator until required. The daily energy intake for a child 1 to 3 years of age with weight between 8 and 15 kg was set at 70 kcal · kg−1 · day−1, and the daily intake of probiotics and prebiotics was estimated to be 109 CFU and 3.8 g, respectively.
Patients were exclusively fed with the formulae for the first 7 days either through a nasogastric tube while they were on the ventilator or orally. Subsequently, formulae were administered orally for up to 14 days, during which period they covered 50% of the patients’ caloric intake. To advance to the full caloric goal of 70 kcal−1 · kg−1 · day, the following algorithm was respected: day 1 = 20 mL (or kcal)−1 · kg−1 · day, day 2 = 40 mL (or kcal)−1 · kg−1 · day, day 3 = 60 mL (or kcal)−1 · kg−1 · day, day 4 = 70 mL (or kcal)−1 · kg−1 · day.
Clinical Assessment of Tolerance and Adverse Effects
Gastric residual volume (GRV) was recorded before the administration of each of the following boluses. Twelve feedings per day were administered to the patients. Tolerance markers such as abdominal distension, episodes of vomiting and diarrhoea, and GRV were used as checkpoints for decisions on the advancement and continuation of tube feeding. For example, if no distention, vomiting, diarrhoea, or GRV < 50% of the previously administered volume, an increment of feeding was implemented. The average abdominal distension was defined to be the daily sum of the abdominal perimeter divided by the number of days at the PICU, and the maximum value was the largest perimeter recorded during the PICU stay. Adverse events were recorded according to the WHO Adverse Reaction Terminology (15).
Stool Bacteria Analysis
Faecal Sample Collection
Faecal samples were collected at admission in the ICU (day 1), discharge from the ICU (day 7), and 14 days postadmission, and processed within 30 minutes after emission. The samples were collected in a sterile tube, placed in an aluminium bag under anaerobic conditions (AneroGen, Oxoid, Basingstoke, UK), and transported to the microbiology laboratory at 4°C. One gram of faeces from each volunteer was homogenised in sterile Ringer solution supplemented with 10% glycerol as cryoprotectant and stored at −80°C until analysis.
Quantification of Selected Bacterial Groups of the Faecal Microbiota
Classical culture methods (16) were used to quantify viable bacterial members of the microbial community, for instance, lactobacilli, bifidobacteria, Enterobacteriae, enterococci, Bacteroides, and Clostridium perfringens. Serial dilutions of the homogenised faecal samples were cultured on appropriate media, and results were reported to be colony-forming units per gram of faeces (log10 CFU/g).
Determination of Faecal Pathogens
Vancomycin-resistant enterococci (VRE), pathogenic members of the Enterobacteriaceae family, and Pseudomonas spp were quantified using plate count methods. For VRE, 10 random colonies from the enterococcosel agar were streaked onto the BBL enterococcosel vancomycin agar (Difco, Franklin Lakes, NJ). Streaks with positive growth corresponded to VRE. Results were reported to be the percentage of VRE among the enterococci population.
Antibiotic Resistance of Enterobacteriaceae and Pseudomonas spp
The disk diffusion method was used for antimicrobial susceptibility tests, and the interpretation depended on zone diameter to be mentioned in the Performance Standards for Antimicrobial Susceptibility Testing (Clinical and Laboratory Standards Institute).
Detection of Probiotic Strains in Faecal Samples
B longum NCC 3001 was detected according to Rouge et al (17). Primers for L paracasei NCC 2461 were designed using a similar approach on its partial genome sequence (oNCC2461_A: TGGACTTAGCGCAGTTTGAA; oNCC2461_B: ACTGAACCATCGTCCCAGAC; B. Berger, unpublished data). The amplification conditions were the same as those for B longum NCC 3001. Results for both strains were reported to be presence/absence values.
Analysis of the Faecal Microbiota Using Denaturing Gradient Gel Electrophoresis
Bacterial genomic DNAs were extracted from faecal samples using the BioRobot EZ1 apparatus (Qiagen, Hilden, Germany), according to the manufacturer's instructions. Polymerase chain reaction (PCR) amplifications were performed using the universal bacterial primers HDA1-GC and HDA2 and subjected to denaturing gradient gel electrophoresis (DGGE), described previously (18). DGGE profiles were compared by determining the Dice's similarity coefficient using the Bionumerics software package (Applied Maths, Austin, TX) at a sensitivity of 1% to 2%. Bacterial diversity was calculated to be number of bands according to Seksik et al (19).
Sample Size Calculation and Statistical Analysis
The primary outcome of the study was cumulative daily energy intake for 7 days, and was expressed as percentage of caloric goal: 70 kcal−1 · kg−1 · day. The primary statistical hypothesis was to demonstrate noninferiority of the test formula compared with the control formula. A boundary of Δ = −15% was defined to be the smallest difference that was clinically acceptable on the overall percentage of caloric intake so that a difference smaller than −15% would matter in practice (noninferiority boundary). The standard deviation (SD) was expected to be 23% [background knowledge was taken from Meert et al (20)]. A noninferiority test would need a sample size of n = 41 per group to demonstrate significant noninferiority on an α-level of 5% with a power of 90% (PASS software, Kaysville, UT). Taking into account 15% dropouts, a total of 96 subjects had to be enrolled for the 2 groups.
The noninferiority margin was set at −15%; thus, noninferiority would be demonstrated if the lower limit of the 95% confidence interval (CI) of the difference (test formula minus control formula) exceeds −15%. The difference was estimated by the pseudo median and the CI were calculated according to Hodges-Lehman implemented in the Wilcox test in the statistical environment R (2.6.1).
Time to reach the target daily caloric intake was analysed by a log-rank test. Abdominal distension was analysed by analysis of covariance, and vomiting and diarrhoea were analysed by Cochran-Mantel-Haenszel tests.
Description of the Groups: Baseline Demographic Characteristics
During the study period, 99 patients were considered eligible, and a total of 94 were enrolled and randomly assigned to receive either the test or the control treatment. The per protocol (PP) population was defined to be all of the randomised subjects without a major protocol deviation; that is, without <4 days of enteral feeding. At the end of the study, there were 80 subjects in the PP population: 41 and 39 in the test and control groups, respectively (Fig. 1). The mean age of the patients was 1.98 ± 0.95 years; the mean weight was 10.29 ± 2.55 kg. The median PRISM score was 0.0 ± 2.4, and 3 subjects had a score ≥8. Eighty-two percent of the subjects were under antibiotic treatment at inclusion before randomisation. Afterwards, 100% and 95.7% of the patients from the test and control groups, respectively, received antibiotics during hospitalisation. The reasons for entering the PICU were pneumonia in 71.3% of the cases, neurological diseases in 17%, combined pneumonia and neurological diseases in 7.4%, and combined pneumonia and cardiovascular diseases in 3.2%.
Primary Endpoint of the Study: Tolerance
The primary endpoint was the overall percentage of caloric intake during the hospitalisation period in the PICU. Caloric intake in percentage of full caloric goal (70 kcal−1 · kg−1 · day) during PICU stay was similar between the 2 formulae. Medians were 76% and 75% for the test and control, respectively (Table 2). In the PP population, the difference between the test and control formulae was −0.02%, and the 95% CI range was −7.8% to +7.4%. The lower limit of the 2-sided 95% CI is well above −15% (noninferiority margin), indicating that noninferiority of the test formula was statistically significant. In the present study, a 3-day progression-feeding plan was applied, and the 100% caloric target was expected to be attained after the third day. The median time to achieve the targeted caloric goal was similar between the groups, 4.13 days and 4.36 days for the test and the control, respectively (P = 0.999, P = 0.748 for PP and intention to treat, respectively).
Abdominal distension was measured daily during PICU hospitalisation. Average distention (sum of the abdominal distension divided by the number of days at ICU) was 42 cm for the test group and 43 cm for the control. The change from baseline was similar in the 2 groups (P = 0.83).
During PICU hospitalisation, 21.3% (10) and 25.5% (12) subjects had at least 1 vomiting episode in the test or control formula, respectively (P = 0.586). At least 1 diarrhoeal episode was recorded in 42.6% (20) and 34% (16) of the subjects of the test and control formula, respectively, which was not significant (P = 0.387).
Faecal Bacteria Analysis
The composition of the faecal microbial population as proxy of the intestinal bacteria was assessed with both classical culture and molecular-based methods. We aimed to evaluate the level of viable bacteria of selected groups and the variation on the predominant bacteria along the study (Fig. 2).
Total bifidobacterial counts decreased in both groups during PICU hospitalisation (7 days). Although the tendency to diminish was noticed in the control group, a progressive increase was observed in the test group. This tendency reached statistical significance between the 2 groups (14 days, P = 0.046). Viable lactobacilli increased progressively during the study for both treatment groups. Subjects receiving EN supplemented with synbiotics presented a trend for a larger population of lactobacilli than the group receiving the nonsupplemented formula (P = 0.085). No statistically significant changes were observed in the levels of faecal bacteria belonging to the Bacteroides group during the study for both treatment groups. At entry, both groups exhibited a similar level of Enterobacteriaceae, which was reduced by 1 log unit in the group receiving the test formula after 7 days at the PICU. These levels remained lower than those at entry in both treatment groups at the end of the study. With regard to enterococci, both treatment groups showed an increase along the study, and the differences between the groups were not significant.
Clostridium perfringens, a species frequently reported to be involved in antibiotic-associated diarrhoea (21,22), was detected in 32% and 31% of the individuals at baseline (control and test group, respectively). The values dropped to 17% and 9%, respectively, after the administration of the products (exclusive feeding period) to increase to about 30% in both groups at the end of the study.
The bacterial strains B longum NCC 3001 and L paracasei NCC 2461 were monitored in this trial. B longum NCC 3001 was detected only in the group receiving the test formula in 17.4% (8) and 13% (6) of the subjects at the end of the PICU and the end of study, respectively. L paracasei NCC 2461 was more frequently detected. It was present in 80.4% and 73.9% of the subjects at the end of the ICU and the end of study, respectively, in the test group (Table 3).
VRE, considered important agents for nosocomial infections, were evaluated in this study. At baseline, VRE represented a mean ± SD of 42% ± 39.9% and 47% ± 36.5% in the test and control groups, respectively. At the end of the study, the values were 38% ± 40% and 51% ± 37% in the test and control groups, respectively, suggesting a trend (not statistically significant P = 0.12; difference −16.4%; 95% CI −37.0 to 4.2) in favour of the test formula.
Antibiotic resistance of some members of the family Enterobacteriaceae and P aeruginosa were tested in all of the patients. The percentage of resistance to at least 1 antibiotic in Enterobacteriaceae and P aeruginosa was similar between the groups (57.4%; P = 0.20).
DGGE-PCR was used to compare population diversity and similarity in predominant bacteria. Similarity indices against baseline were on average low for both groups (∼60%) after 7 and 14 days, suggesting a relatively unstable microbiota. No differences were observed with regard to the bacterial diversity (number of bands) in both groups during the whole study (Fig. 3).
The most frequently reported category was GI disorders, particularly diarrhoea. No significant differences between groups were detected. There were no reported secondary infections during the ICU stay associated with any of the 2 administered probiotics.
We report on a study conducted in 2 ICUs to assess the tolerance and safety of an enteral formula containing a synbiotic blend administered to critically ill children. We found that this formula was as well tolerated as the currently used formula and that it was safe. A secondary objective was to measure changes in the gut microbial composition using faecal samples as proxy for intestinal contents. The numbers of lactobacilli and bifidobacteria were boosted relative to the control formula.
EN is the preferred mode of nutrient intake in critically ill patients because of its lower cost and complication rates when compared with parenteral nutrition (23). Both early institution of EN and adequate caloric delivery have proven to be important factors to promote recovery in patients in the ICU (2,4). Unfortunately, adequate energy intake is sometimes not attained because of low tolerance to the formulae (24), interruptions in enteral feeding (25), or poor performance of nutrition guidelines and procedures (26). This last factor is less important in cases in which EN is administered by bolus, as in the present study, and not by continuous tube feeding.
Enteral diets can affect gut physiology because of modifications in the transit time, alterations of the intestinal secretory/absorptive capacity, and modification of the microbial ecology (9,27). We hypothesised that the enteral test formula used in this study was nutritionally adapted for the paediatric population in the ICU, and can improve the intestinal microbial ecology of paediatric patients in the ICU. Considering that fibre (oligosaccharides) and live probiotics were used in the formulation, we examined its safety and tolerance compared with a nutritionally identical recipe devoid of the pre- and the probiotics, which have a history of good tolerance and safe use. All of the tolerance markers studied in this trial, for example, percentage of caloric intake, time to achieve the caloric goal, abdominal distension, vomiting, and diarrhoea/stool frequency, indicated that the new formulation was as well tolerated as the current formula. The weight change during ICU stay and until the end of follow-up was also similar between groups.
Enteral tube feeding and antibiotic use can profoundly disturb the ecological homeostasis in patients in the ICU (28). Other factors such as the global stress reaction, the neutralisation of gastric secretion, sedation and analgesia that impair intestinal motility, gut ischemia, and immune dysfunction can further contribute to gut ecology modifications (27,29). The disruption in the composition of the gut microbiota has a potential relevance to the final outcome in critically ill patients (30). Therefore, efforts to preserve homeostasis could provide a clinically relevant advantage. Because the new formulation containing synbiotics was designed to support the ecological intestinal homeostasis of patients in the ICU, we evaluated the effect of the administration of the formulae on the gut microbiota using classical culture methods and molecular-based tools. This comprehensive approach allowed us to assess changes in viable bacteria and those in not-yet-cultured bacteria.
The selection of the 2 probiotics present in the product was based on the previous demonstration that L paracasei NCC 2461 and B longum NCC 3001 can resist passage through the GI tract. Moreover, L paracasei NCC 2461 has a proven efficacy for the management of bacterial infectious diarrhoea in children (31), and B longum NCC 3001 has been associated with improvement of GI discomfort induced by antibiotics in healthy subjects (32). Additional reports support the use of these 2 strains in clinical settings (33,34,37). Of interest, Chouraqui et al (35) administered the same L paracasei strain in combination with prebiotics to healthy full-term infants with successful results regarding safety and tolerance. L paracasei was also used in a study performed in Bangladeshi children with acute diarrhoea. The administration of the probiotic was not only safe but also resulted in a clinically significant benefit in the management of diarrhoea (31). A series of clinical trials were performed with B longum BB536 (coded at the Nestle Culture Collection as NCC 3001) on healthy, but “at-risk” subjects because of their extremely low weight, preterm neonates. No adverse events were recorded (36,37).
It has been reported in both animal and human trials that L paracasei NCC 2461 is recovered in high numbers in the faeces of the subjects who had consumed the probiotic (38–40). It was not surprising, therefore, that this strain was present in a high number of patients who received the supplemented formula (80% of the faecal samples at 7 days and 74% at 14 days) despite the majority of patients being under antibiotic treatment. In contrast, B longum NCC 3001 was detected in a small number of patients (<17%), suggesting that the strain persisted only transiently. Bennet et al (41) failed to recover this strain 10 days after the consumption of the product containing the probiotic ceased.
Most of the antibiotics administered to the children during the study belonged to the β-lactam group, to which both probiotic strains are sensitive (data not shown). We speculate that the frequent intake of formula containing probiotics is likely to overcome dynamic elimination because of the antibiotics resulting in sustained persistence, more evident for L paracasei NCC2461 than for B longum NCC3001.
A few patients from the control group tested positive for the L paracasei strain, probably because of cross-reaction of the primers with normal resident lactobacilli. Although the specificity of our primers was validated against a diverse collection of strains, it is still restricted compared with the diversity of strains found in nature. New isolates to test the primers against will help improve the specificity of the method.
A clear drop in total bifidobacterial counts was observed during the first week of ICU hospitalisation followed by recovery in the second week only in the test group. We suspect that the increase in the total bifidobacterial population was related to the fibres included in the formula rather than to the contribution of the probiotic counts. FOS and insulin have been shown to influence colonic microbiota composition (42), and Acacia gum is a well-tolerated fibre with bifidogenic properties believed to benefit intestinal health (43,44).
Pre- and probiotics can antagonise reported global changes of the microbiota in patients in the ICU. It has been observed that during the ICU stay there is a reduction in the repertoire of dominant microbiota members (29) and that faecal samples collected at various time points during hospitalisation were 62% similar to baseline on DGGE profiles. In our study, a comparable reduction in the number of DGGE bands from the beginning and during the intervention, including follow-up, was observed in both groups. The harsh luminal environment in this clinical condition has dramatic effects on the microbiota stability.
Secondary nosocomial infections are frequently caused by organisms acquired in the ICU, and they are preceded by oropharyngeal or GI colonisation (45–47). For instance, the presence of P aeruginosa, Acinetobacter spp, methicillin-resistant Staphylococcus aureus, and vancomycin-resistant Enterococcus faecium has been shown to correlate with the development of secondary infections (45,47,48). An interesting observation of our study was the increase in the colonisation levels of global enterococci similar to previous reports (29). In addition, a high proportion of children were colonised at baseline, with VRE representing around 45% of the total enterococci population regardless of the group, with a reduction (16.4%) of VRE in the test group at the end of the study. Infection levels with VRE were not recorded in this study, and further studies are needed to consolidate the relevance of this preliminary observation.
Constituents of the commensal intestinal microbiota and exogenous agents, for example, probiotics and prebiotics, could contribute to the restoration of gut homeostasis. These agents are increasingly receiving attention as a novel approach to prevent intercurrent infections and dysbiosis in the microbial makeup of the colon. In conclusion, our study showed that the administration of an enteral formula supplemented with synbiotcs was safe and well tolerated in patients from PICUs. Changes in numbers of bifidobacteria and lactobacilli—2 bacterial groups of previously reported beneficial effects, which were observed in this study—should encourage the design of future clinical trials powered enough to unequivocally determine their clinical benefits.
The authors thank Mr Ludovic Cottet and Miss Isabelle Rochat for technical assistance, Mrs Florilène Bouisset for comments on the statistical section, and Dr Pascal Stucki from the Paediatric Intensive Care Unit at Lausanne University Hospital for help in the design of the feeding algorithm. The participation of the children in this study is gratefully acknowledged.
1. Petros S, Engelmann L. Enteral nutrition delivery and energy expenditure in medical intensive care patients. Clin Nutr 2006; 25:51–59.
2. Villet S, Chiolero RL, Bollmann MD, et al. Negative impact of hypocaloric feeding and energy balance on clinical outcome in ICU patients. Clin Nutr 2005; 24:502–509.
3. Reintam A, Parm P, Kitus R, et al. Gastrointestinal symptoms in intensive care patients. Acta Anaesthesiol Scand 2009; 53:318–324.
4. Kreymann KG, Berger MM, Deutz NEP, et al. ESPEN guidelines on enteral nutrition: intensive care. Clin Nutr 2006; 25:210–223.
5. Moore FA, Feliciano DV, Andrassy RJ, et al. Early enteral feeding, compared with parenteral, reduces postoperative septic complications: the results of a meta-analysis. Ann Surg 1992; 21:172–183.
6. Kudsk KA, Croce MA, Fabian TC, et al. Enteral versus parenteral feeding: effects on septic morbidity after blunt and penetrating trauma. Ann Surg 1992; 21:503–511.
7. Jiang XH, Li N, Li JS. Intestinal permeability in patients after surgical trauma and effect of enteral nutrition versus parenteral nutrition. World J Gastroenterol 2003; 9:1878–1880.
8. Zimmaro Bliss D, Johnson S, Savik K, et al. Acquisition of Clostridium difficile and Clostridium difficile-associated diarrhea in hospitalized patients receiving tube feeding. Ann Intern Med 1998; 129:1012–1019.
9. Whelan K, Judd P, Preedy VR, et al. Fructooligosaccharides and fiber partially prevent the alterations in fecal microbiota and short chain-fatty acid concentrations caused by standard enteral formula in healthy humans. J Nutr 2005; 135:1896–1902.
10. Doron SI, Hibberd PL, Gorbach SL. Probiotics for prevention of antibiotic-associated diarrhea. J Clin Gastroenterol 2008; 42:S58–S63.
11. Ruszczyński M, Radzikowski A, Szajewska H. Clinical trial: effectiveness of Lactobacillus rhamnosus (strains E/N, Oxy and Pen) in the prevention of antibiotic-associated diarrhoea in children. Aliment Pharmacol Ther 2008; 28:154–161.
12. Gibson GR, Roberfroid MB. Dietary modulation of the human colonic microbiota: introducing the concept of prebiotics. J Nutr 1995; 125:1401–1412.
13. Schrezenmeir J, de Vrese M. Probiotics, prebiotics, and synbiotics—approaching a definition. Am J Clin Nutr 2001; 73:361S–364S.
14. Pollack MM, Patel KM, Ruttimann UE. The Pediatric Risk of Mortality III-Acute Physiology Score (PRISM III-APS): a method of assessing physiologic instability for pediatric intensive care unit patients. J Pediatr 1997; 131:575–581.
15. World Health Organization, Adverse Reaction Terminology (WHO-ART). The Uppsala Monitoring Centre, December 2005. http://www.umc-products.com
. Accessed November 1, 2010.
16. Tannock GW, Munro K, Harmsen HJM, et al. Analysis of the fecal microflora of human subjects consuming a probiotic containing lactobacillus rhamnosus DR 20. Appl Environ Microbiol 2000; 66:2578–2588.
17. Rouge C, Goldenberg O, Ferraris L, et al. Investigation of the intestinal microbiota in preterm infants using different methods. Anaerobe 2010; 16:362–370.
18. Tannock GW, Munro K, Bibiloni R, et al. Impact of consumption of oligosaccharide-containing biscuits on the fecal microbiota of humans. Appl Environ Microbiol 2004; 70:2129–2136.
19. Seksik P, Rigottier-Gois L, Gramet G, et al. Alterations of the dominant faecal bacterial groups in patients with Crohn's disease of the colon. Gut 2003; 52:237–242.
20. Meert KL, Daphtary KM, Metheny NA. Gastric vs. small-bowel feeding in critically ill children receiving mechanical ventilation: a randomized controlled trial. Chest 2004; 126:872–878.
21. Sparks SG, Carman RJ, Sarker MR, et al. Genotyping of enterotoxigenic Clostridium perfringens fecal isolates associated with antibiotic-associated diarrhoea and food poisoning in North America. J Clin Microbiol 2001; 39:883–888.
22. Abrahao C, Carman RJ, Hahn H, et al. Similar frequency of detection of Clostridium perfringens enterotoxin and Clostridium difficile toxins in patients with antibiotic-associated diarrhoea. Eur J Clin Microbiol Infect Dis 2001; 20:676–677.
23. Heyland DK, Dhaliwal R, Drover JW, et al. Canadian clinical practice guidelines for nutrition support in mechanically ventilated, critically ill adult patients. JPEN J Parenter Enteral Nutr 2003; 27:355–373.
24. Rogers EJ, Gilbertson HR, Heine RG, et al. Barriers to adequate nutrition in critically ill children. Nutrition 2003; 19:865–868.
25. Adam S, Batson S. A study of problems associated with the delivery of enteral feed in critically ill patients in five ICUs in the UK. Intensive Care Med 1997; 23:261–266.
26. Elpern EH, Stutz L, Peterson S, et al. Outcomes associated with enteral tube feeding in a medical intensive care unit. Am J Crit Care 2004; 113:221–227.
27. Whelan K, Judd PA, Tuohy KM, et al. Fecal microbiota in patients receiving enteral feeding are highly variable and may be altered in those who develop diarrhea. Am J Clin Nutr 2009; 89:240–247.
28. Volakli E, Speis C, Michalopoulus A, et al. Infections of the respiratory or abdominal origin in ICU patients: what are the differences? Crit Care 2010; 14:R32.
29. Iapichino G, Callegari ML, Marzorati S, et al. Impact of antibiotics on the gut microbiota of critically ill patients. J Med Microbiol 2008; 57:1007–1014.
30. Shimizu K, Ogura H, Goto M, et al. Altered gut flora and environment in patients with severe SIRS. J Trauma 2006; 60:126–133.
31. Sarker SA, Sultana S, Fuchs GJ, et al. Lactobacillus paracasei strain ST11 has no effect on rotavirus but ameliorates the outcome of non-rotavirus diarrhea in children from Bangladesh. Pediatrics 2005; 116:e221–e228.
32. Colombel J F, Cortot A, C. Neut C, et al. Yoghurt with Bifidobacterium longum reduces erythromycin-induced gastrointestinal effects. Lancet 1987; 330:43–47.
33. Bunout D, Barrera G, Hirsch S, et al. Effects of a nutritional supplement on the immune response and cytokine production in free-living Chilean elderly. JPEN J Parenter Enteral Nutr 2004; 28:348–354.
34. Puccio G, Cajozzo C, Meli F, et al. Clinical evaluation of a new starter formula for infants containing live Bifidobacterium longum BL999 and probiotics. Nutrition 2007; 23:1–8.
35. Chouraqui JP, Grathwohl D, Labaune JM, et al. Assessment of the safety, tolerance and protective effect against diarrhea of infant formulas containing mixtures of probiotics or probiotics and prebiotics in a randomized controlled trial. Am J Clin Nutr 2008; 87:1365–1373.
36. Akiyama K, Hosono S, Takahashi E, et al. Effects of oral administration of Bifidobacterium breve on development of intestinal microflora in extremely premature infants. Acta Neonat Jap 1993; 30:130–137.
37. Akiyama K, Shimada M, Ishizeki S, et al. Effects of administration of Bifidobacterium in extremely premature infants: development of intestinal microflora by orally administered Bifidobacterium longum in comparison to Bifidobacterium breve. Acta Neonat Jap 1993; 30:257–263.
38. Gueniche A, Benyacoub J, Bureau-Franz I, et al. Probiotic strain Lactobacillus paracasei CNCM I-2116 decreases skin reactivity. J Invest Dermatol 2007; 127:S17.
39. Ibnou-Zekri N, Blum S, Schiffrin E, et al. Divergent patterns of colonization and immune response elicited from two intestinal Lactobacillus strains that display similar properties in vitro. Infect Immun 2003; 71:428–436.
40. Solomons NW, Bulux J, Parreno F, et al. Incidence of diarrhea cases and diarrheal episodes with or without daily feeding of Lactobacillus paracasei NCC2461. J Pediatr Gastroenterol Nutr 2000; 31 (Suppl 2):S254.
41. Bennet R, Nord CE, Zetterstrom R. Transient colonization of the gut of newborn infants by orally administered bifidobacteria and lactobacilli. Acta Paediatr 1992; 81:784–787.
42. Schley PD, Field CJ. The immune-enhancing effects of dietary fibres and prebiotics. Br J Nutr 2002; 87:S221–S230.
43. Reid G, Sanders ME, Gaskins HR, et al. New scientific paradigms for probiotics and prebiotics. J Clin Gastroenterol 2003; 37:105–118.
44. Cherbut C, Michel C, Raison V, et al. Acacia gum is a bifidogenic dietary fibre with high digestive tolerance in healthy humans. Microb Ecol Health Dis 2003; 15:43–50.
45. Vincent J-L, Rello J, Marshall J, et al. Internal study of the prevalence and outcomes of infection in intensive care units. JAMA 2009; 302:2323–2329.
46. Rogers CJ, van Saene HK, Suter PM, et al. Infection control in critically ill patients: effects of selective decontamination of the digestive tract. Am J Hosp Pharm 1994; 51:631–648.
47. de Jonge E, Schultz MJ, Spanjaard L, et al. Effects of selective decontamination of digestive tract on mortality and acquisition of resistant bacteria in intensive care: a randomized controlled trial. Lancet 2003; 362:1011–1016.
48. Marshall JC, Christou NV, Meakins JL. The gastrointestinal tract. The “undrained abscess” of multiple organ failure. Ann Surg 1993; 218:111–119.