3.1 Study identification and selection
A total of 93 relevant articles were identified through a database search and a review of reference lists of related articles. We excluded 78 articles due to the following reasons: 21 were conference abstracts, or the full text was not available, 17 were not RCTs, 13 did not exclusively enroll ICU patients, 6 were not in English, 8 did not report relevant outcomes, 5 had multiple interventions, 2 did not include exclusive EN, 1 did not have a placebo receiving control group, 5 due to other reasons e.g. irrelevant intervention or inconsistency in results. Finally, 15 RCTs, with a total of 1139 patients, were included (Figure 1). As the articles included in this systematic review were heterogeneous in population, inclusion and exclusion criteria, intervention, duration, and reported outcomes, the further statistical comparison was not possible. The mean JADAD score of all trials was 4.2. The minimum JADED score was 1 point for randomization. A summary of the findings of the studies is presented in Table 2.
Table 2 Randomized controlled trials evaluating the effect of pre, pro or synbiotics on feeding tolerance of enterally-fed critically ill patients
Abbreviations: NR, not reported; NGT, nasogastric tube; EN, enteral nutrition; ICU, intensive care unit; OGT, orogastric tube; GRV, gastric residual volume; VSL#3, a single daily high dose probiotic preparation.
3.2 Effect on energy intake and feed volume
Six trials examined the effect of pre, pro, or synbiotics on energy intake or feed volume in critically ill patients.
3.2.1 prebiotics
In one trial in 2000, 44 critically ill patients receiving EN and antibiotics were randomized to receive fiber-containing or fiber-free formula and pectin or placebo for six days. Mean energy intake ranged from 1200 kcal on day 1 to 1563 kcal on day 5. Mean energy or protein intake was not significantly different in the four study groups (19). Rushdi et al. also evaluated the effect of guar gum enriched formula in 20 critically ill tube-fed patients with persistent diarrhea for four days. They showed that patients in the intervention group tolerated significantly higher formula volumes on days 1, 2, and 4. On the fourth day, the feed volume was 1775±450 ml in the intervention group compared to 1070±604 ml in the controls (p<0.01) (21). In 2018, Fazilaty et al. evaluated the effect of EN containing β-glucan on inflammatory markers and clinical outcomes. They reported no significant difference in the mean tolerated calories between study groups (1710.5±117.03 kcal vs. 1718.2±182.4 kcal, p=0.6) (30). Tuncay et al. compared the effect of an enteral formula enriched with prebiotic versus standard EN on nutritional parameters among 46 neurocritically ill patients. Results showed that feed volume and mean energy intake significantly increased from baseline to day 21 in both groups. Patients in the intervention group tolerated a significantly higher amount of energy and feed volume on day 1 and 21 (32).
3.2.1 probiotics
In a trial conducted in 2014, 40 critically ill patients were randomly assigned to receive a multi-strain probiotic or placebo for seven days. Results showed no significant difference between groups in terms of the mean energy intake (1503.75±231.6 kcal vs. 1617.5±185.51 kcal, p= 0.09). The percentage of patients who met energy requirements in the synbiotic and placebo groups was 84.98±3.6 and 87.24±3.92, respectively (p=0.06) (27).
3.2.2 synbiotics
Knight et al. investigated the effect of enteral synbiotic on ventilator-associated pneumonia in critically ill patients. They reported an increase in the daily tolerated feed volume from day 1 to 7 in both groups. The feed volume ranged from 488.9± 622.8 ml on day 1 to 1055.6±722.6 ml on day 7 in the synbiotic group and from 360±431.7 ml to 1243.9±810.3 ml in the placebo group. There was no significant difference between the two groups regarding the mean tolerated enteral feed volume (22).
3.3 Effect on target calorie achievement
Four trials assessed the effect of prebiotics or probiotics on the prevalence of target calorie achievement or time to receive the target calorie.
3.3.1 prebiotics
In a trial conducted in 2001, severe sepsis or septic shock patients were randomly assigned to receive EN supplemented with partially hydrolyzed guar or fiber-free EN. All patients were on mechanical ventilation, antibiotics, and catecholamine therapy. The time to reach the preconceived protein/calorie goals was 5±3 days in the prebiotic and 6±3 days in the control group. The difference was not statistically significant (20). In another trial conducted by Tuncay et al., the prevalence of target dose achievement in 21 days intervention was 95.7% in prebiotic supplemented and 78.3% in standard EN groups. The difference was not statistically significant (p=0.19) (32).
3.3.2 probiotics
Malik et al. investigated the effect of 7-days microbial cell preparation administration on the return of gut function. Time to return to normal gut function was defined as the time taken to receive a minimum of 80% of the estimated calorie for a consecutive 48-hour period. They reported that patients in the treatment group achieved a faster return of the gut function (3±1.75 days vs. 7±1.7 days, p<0.001) (29). Ferrie et al. also investigated the effect of Lactobacillus rhamnosus GG on feeding intolerance in critically ill patients with established diarrhea. The frequency of patients with feeding intolerance (tolerate less than 80% of calorie goal for two consecutive days) was 11.1% in the probiotic group and 16.6% in the control group (p= 0.63) (26).
3.4 Effect on diarrhea
3.4.1 Prebiotics
Schultz et al. investigated the effect of pectin on the prevalence of diarrhea in a critical care setting. Diarrhea was more prevalent in the fiber-free/placebo group compared to the fiber-free/pectin group (36% vs. 9%, p=0.31). Diarrhea was also more prevalent in the fiber/ placebo than fiber/ pectin group, however, the differences were not statistically significant (55% vs. 9%, p= 0.06) (19). Majid et al. also demonstrated that fiber-enriched EN with additional prebiotic had no significant effect on the prevalence of diarrhea. The prevalence of having at least one day of diarrhea was 92% in the prebiotic and 90% in the placebo group (p=0.99). The number of days of diarrhea was 3.8±3.5 in the placebo and 3.9±4.1 in the prebiotic group (p= 0.94) (28). Tuncay et al. reported that administration of prebiotic-enriched EN was associated with a significant tendency toward lower prevalence (8.7% vs. 56.5%) and faster amelioration of diarrhea (none vs. 52.2% diarrhea prevalence on day 7) (32).
In another trial, the prevalence of having at least one day of diarrhea was 46.1% in the prebiotic and 91.7% in the placebo group (p=0.03). Further, the mean frequency of days of having diarrhea was significantly lower in the fiber group. Besides, in the fiber group, diarrhea occurred in 10.8% of feeding days, compared to 31.5% in the controls (p<0.001) (20).
Rushdi et al. also investigated the effect of soluble guar gum on the number of liquid stools during the four days of intervention. The number of liquid stools was significantly lower at day 4 compared to day 1 in the intervention group, while it was significantly higher in the control group. The number of liquid stools on the fourth day was 1.2±0.7 in the intervention group, compared to 2.1±0.8 in the control group (p<0.01) (21).
3.4.2 probiotics
In the study by Bleichner et al., 128 critically ill patients were randomized to receive Saccharomyces Boulardii or placebo capsules. The prevalence of diarrhea was not significantly different between the two groups. However, treatment with S. boulardii reduced the mean frequency of diarrhea days per feeding days from 18.9 to 14.2% (p= 0.006). The number of days with diarrhea was also significantly lower in the probiotic group (p<0.001) (18). Barraud et al. also investigated the effect of probiotic administration on the prevalence of diarrhea. They demonstrated no significant effect of probiotic therapy on diarrhea prevalence (55.2% vs. 52.5%, p=0.72) (24). In another RCT, Morrow et al. demonstrated that probiotic administration had no significant effect on the incidence of ICU associated diarrhea. However, the number of days with ICU-associated diarrhea was significantly higher in the placebo group compared to the probiotic group (5.9±3.8 vs. 4.1±3.7, p=0.03) (25). Another trial in 2010 examined the effect of probiotic VSL#3 on diarrhea among 45 critically ill patients. The mean frequency of liquid stool in the probiotic and placebo groups was 0.53±0.54 and 1.05±1.08 episodes per patient per day, respectively (p=0.03) (23). Ferrie et al. also reported that critically ill patients who received probiotic had more diarrhea episodes compared to the control group, although the difference was not statistically significant. Diarrhea days in the 14-day study period was 7.22±3.63 in the probiotic and 5.72±2.88 in the synbiotic group (p=0.17) (26).
3.4.3 synbiotics
Knight et al. reported the overall prevalence of diarrhea to be 5% in the synbiotic group, and 7% in the controls (p= 0.59) (22). Shimizu et al. also investigated the effect of daily synbiotic therapy on infectious complications, including enteritis in the intensive care unit. Enteritis was defined as acute onset of continuous liquid stool for more than 12 hours. The results showed that the incidence of enteritis was significantly lower in the synbiotic group (6.3% vs. 27.0%; p < 0.05) (31).
3.5 Effect on length of stay
The effect of pre, pro or synbiotics on ICU and hospital LOS was reported in 10 and 5 trials, respectively (Table 3). Malik et al. demonstrated that probiotic administration was associated with significantly lower ICU LOS (29). Other studies found no significant difference between groups, regarding ICU or hospital LOS (19, 22, 23, 25, 26, 30-33).
Table 3 Reported fedding tolerance-related outcomes in RCTs evaluating the effect of pre, pro or synbiotics on feeding tolerance of enterally-fed critically ill patients
Abbreviations: NR, not reported; ICU, intensive care unit.
ǂ Mean energy intake was reported for the entire intervention duration