Background

The prevalence of malnutrition in hospitalized patients was around 40% and is even higher among critically ill patients [1]. And this is accompanied by poor outcomes such as higher mortality rate, increasing length of mechanical ventilation (MV), and pediatric intensive care unit (PICU) stay length among malnourished critically ill one [2]. So, appropriate nutrition is one of the most important components in the management of those patients [3].

The main goal of nutrition in PICU patients is to give a good nutrition support to those who need it, according to their clinical and nutritional status, metabolic capability, and which route of administration is available [4]. Follow-up of the patients as regard the energy and protein imbalance is a must to avoid under or overfeeding [5].

There are two types of nutrition delivery in PICU: enteral nutrition (EN) and parenteral nutrition (PN). If the critically ill patient has a functioning gastrointestinal (GI) tract, EN should be the preferred mode of nutrition to maintain gut mucosa integrity [6]. This indicated when oral intake cannot be maintained to meet the metabolic demands of the patients. Otherwise, when there is impaired gastrointestinal function and contraindications to enteral nutrition, PN which involves the infusion of mixture of amino acids, carbohydrates, and lipids, as well as electrolytes and micronutrients, can be given instead [7].

Recent guidelines for nutrition support in children who need PICU admission recommend early EN within 24–48 h of admission; this greatly improves their clinical outcomes [8].

In addition, delaying PN for 1 week in the PICU will result in lowering the sepsis rate, lowering the length of PICU stay, and overall hospital stay [9].

Based on the conflicting findings of the different reviews, it is fair to say that the question as to the best form and the timing of nutrition start is still largely unanswered. Therefore, the aim of our study is to compare the outcomes of early EN with early PN in critically ill children.

Methods

Study design

A case–control study was conducted at PICU of our University Children’s Hospital, Faculty of Medicine.

Objective

It is to compare the outcomes of early EN with early PN in critically ill children with respiratory distress, as regards the clinical and biochemical parameters in children.

Patient selection

We included 180 ill children, randomly selected, aged 2 months–5 years who were admitted to the PICU with respiratory distress on the following: mechanical ventilator, noninvasive CPAP, or high-flow nasal oxygen and then subdivided into two groups: Group (1): 90 cases receive early EN had no contraindication for beginning enteral nutrition &Group (2): 90 cases receive early PN.

We excluded patients expected to die within 12 h, have risk of aspiration, transferred from another PICU after a stay of more than 7 days, suffering from severe gut ischemia, any surgical problems, uncontrolled shock, severe respiratory distress with uncontrolled hypoxemia, and acidosis on admission and patient’s suspicious or established inborn metabolic diseases requiring specific diet.

Methodology

Patients will be subdivided to start enteral or parenteral feeding within the first 24 h of admission according to their clinical condition. Firstly, energy requirements in every patient should be individualized. The initial energy calculation is based on the formulas of Schofield and the World Health Organization (WHO) which provide estimates of caloric needs (BMR) [10]. Then, reported BMRs based on ideal weight or estimations of basal caloric expenditure can be multiplied by a stress factor. Then, calculated total energy requirements by the following equation are as follows: TEE = BMR + SDA + energy for activity + energy for growth + energy for thermoregulation [11].

*TEE, total energy expenditure; BMR, basal metabolic rate; SDA, specific dynamic action of food

Patients who assigned to start enteral feeding (artificial formula) started on (1 cc/kg/h) and then increased each feeding by 25%volume until goal is reached.

Patients on parenteral nutrition will start total parenteral nutrition in general by initial proteins and lipids with a dose of 1 g/kg/day with daily increase 1 g/kg/day, with a maximum of 3 g/kg/day. Calculations of macronutrients were done according to Joosten et al., and GIR will be adjusted according to the blood glucose level [10].

All data were recorded from patients’ files at days 1, 3, and 7.

Anthropometric measurements are weight (weight gain will be assessed on days 1, 3, 7, and weekly until discharge) and height, clinical assessment, and physical examination.

Type of oxygen support, mechanically ventilated or not, pulmonary indices, i.e. oxygen saturation index (OSI), and So2/Fio2 ratio will be calculated and recorded which are used to assessing severity of lung injury [12, 13], length of stay on mechanical ventilation, length of stay on PICU, mortality rate, and complications, e.g. sepsis, VAP, intolerance, aspiration, and diarrhea were recorded.

And the following laboratory investigations were taken from the sheet of patient as follows:complete blood count (CBC): platelets count, hemoglobin (Hb) level and total leucocyte count (TLC), C-reactive protein (CRP), blood urea nitrogen (BUN) and serum creatinine liver enzyme analysis (ALT and AST), and serum Na and K.

Statistical analysis

All data will be collected, tabulated, and analyzed using the Statistical Package for Social Science (SPSS) version 16. The following methods will be employed: mean and standard deviation (SD) or median and interquartile range (IQR) will be estimates of qualitative data. Differences in clinical and biochemical characteristics will be tested by student’s paired and unpaired t-test, by Mann–Whitney U-test for quantitative data, and by chi-square test for qualitative data. P-values equal or less than 0.05 will be considered statistically significant.

Results

In total, 180 patients were included in the study: 113 (62.8%) were males, and 67 (37.2%) were females. The mean ± SD for age was 11 ± 8.5 with 10 median and range 6–38/months. The median (IQR) baseline weight was 7.27 (5–11.25)/kg. The commonest underlying diagnosis was pneumonia in 94 (52.2%) and bronchiolitis in 54 (30%). Of them, 79 (43.9%) were put on a mechanical ventilation. The descriptive statistics of the sample are displayed in Table 1.

Table 1 Baseline demographics and clinical characteristics of the studied population
Full size table

No significant differences were observed between the PN and EN groups in terms of age and baseline weight; only significant difference was observed as regards gender (P = 0.03) and admission diagnosis (bronchiolitis & pneumonia) (P = 0.01, 0.001).

Comparing both groups, we observed that there was highly statistically significant weight gain for patients on EN compared with those on PN (P = 0.001). The median (IQR) for weight for the EN group was 8.6 (5.2–12.3)/kg, 10 (6–14.5)/kg after 1st and 2nd weeks, while it was 6.1 (4.5–8.5)/kg and 6 (3.8–8)/kg for the PN group as shown in Fig. 1.

Fig. 1
figure 1

Weight gain in different times of the studied patients regarding nutrition type

Full size image

As regards Z score, there was no statistically significant difference between both groups at admission times, but after 1 week, there was slightly statistically significant difference between both of them with P = 0.05; also, at the second week, there was statistically significant difference between both groups with P = 0.01 (Table 2).

Table 2 Comparison between both groups as regard Z-score grade
Full size table

The mean ± SD for time to reach caloric target for those receiving early EN was 4.0 ± 1.9 days which is earlier than that of PN group that was mean ± SD (6.2 ± 1.7 days) (P = 0.001) as shown in Fig. 2.

Fig. 2
figure 2

Comparison between both groups as regard time to reach caloric target

Full size image

As regards laboratory assessment, we found that TLC, CRP, ALT, and urea were elevated in PN group. The mean ± SD for TLC was 15.1 ± 7.4 and 35.4 ± 7.3 × 103/µl) at the 3rd and 7th days of admission, respectively, and median (IQR) of CRP was 34.7 (3.1–251) mg/l at the 7th day of admission in PN group. But, in EN group, the mean ± SD for TLC was 11.7 ± 4.2, 10.8 ± 3.5 × 103/Ul at the 3rd and 7th days of admission, respectively, and median (IQR) for CRP was 12 (5.1–25) mg/l at the 7th day (P = 0.001).

The median (IQR) for ALT was 49 (18.3–121) µ/l, and the mean ± SD for serum urea was 16.8 ± 8.8 mg/dl which was elevated on PN group than EN group; the median (IQR) of ALT was 29 (16–49) µ/l, and the mean ± SD of serum urea was 12 ± 6.9 mg/dl (P = 0.001) as shown in Table 3.

Table 3 Comparison between both groups as regards laboratory investigations
Full size table

Regarding complications, it was presented in both groups, but we observed high percentage of sepsis 37 (41.1%) and VAP 23 (25.6%) on those who received PN (P = 0.001) as shown in Table 4.

Table 4 Comparison between both groups as regard complications
Full size table

The mean ± SD for survival time for EN group and PN were 28.0 ± 1.16 days and 20.8 ± 0.72 days, respectively (P = 0.001). In the group who received early EN, recovered patients were 88.9%, and the mortality rate was 11.1%, while in the PN group, recovered patient’s percentage was 72.2%, and mortality rate was 27.8% (P = 0.01) (Table 5).

Table 5 Comparison between both groups as regard outcome
Full size table

No significant difference was observed between both groups as regard PICU stay length, MV stay length, and O2 support stay length as shown in Table 6.

Table 6 Comparison between both groups as regard MV, PICU, and oxygen support length of stay
Full size table

The results from a multivariate logistic regression model for mortality are displayed in Table 7 showed that sepsis was related to the mortality (odd ratio = 0.551) (P < 0.001).

Table 7 Multivariate logistic regression model for mortality
Full size table

Discussion

Patients need admission to PICU should be correctly evaluated, and nutritional plane should be put according to their caloric and energy needs [5]. Comparison between the two methods of nutrition (parenteral and enteral nutrition) in PICU is so difficult which is due to physiologic differences between both of them. However, the most preferred route was the EN route as initiation of early PN was associated with a lot of complication, e.g., sepsis, prolonged hospital stay, and mechanical ventilation stay [9].

The aim of our study is to compare the outcomes of early enteral nutrition with early parenteral nutrition in critically ill children with respiratory distress. And we concluded that early EN remains the preferred route for nutrient delivery as the PN route was accompanied by a lot of complication such as sepsis, VAP, and high mortality rate.

Importantly, it is essential to prevent malnutrition in critically ill patients, as malnutrition is associated with impaired immune function, as a result of a cascade of metabolic and hormonal derangements especially in critically ill patients [14], which in turn leads to prolonged mechanical ventilation stay and increase rate of morbidity and mortality. EN is an active therapy that can modulate the immune system and prevent intestinal villi atrophy, enterocyte apoptosis, and impairment of gut immune functions especially if it is given at the first 48–72 h of admission.

Our study concluded that the EN route is preferred over the PN route in preventing the malnutrition. As we could detect after the first and second week from admission, significant weight gain for patients on EN compared with those on PN and the mean time for reaching caloric target for those receiving early EN was earlier than that for early PN. In contrary, other study concluded that the weight gain for the adult patients is not significant with a nutrition [15].

On the other hand, we observed high percentage of sepsis and VAP on those who received PN compared with EN group. This is in accordance with studies which revealed that patients who received early EN have a low sepsis and VAP frequencies [16, 17]. This was not in agreement with another study in critically ill adults that revealed no difference in sepsis rate between both routes of nutrition [18].

So again, EN seems to have the upper hand over PN in lowering number of infectious and noninfectious complications. EN has been shown to stimulate the growth and function of the intestine, both directly intra-luminally, as it supplies substrates for enterocyte oxidation, and indirectly, through stimulation of hormone secretion, which could result in reduction of bacterial translocation and the complications accompanied with it [19]. In addition, it maintains the normal pH value of gastric juice, and this in turn lowers the incidence of VAP [17].

On the other hand, the high incidence of sepsis in PN may be related to catheter insertion, uncontrolled glycemic state of the patient, and high lipid content. All of these factors are good media for microorganism growth.

However, PN is recommended within 24–48 h by ESPEN if EN is contraindicated. The following strategies which is: (1) a good control for the blood glucose level, (2) the use of olive oil-based lipid emulsions instead of soybean oil-based ones, (3) the adoption of insertion and care bundles for central venous access devices, and (4) application of a policy of targeting “near-zero” catheter-related blood stream infections, which can be minimized by the following: (i) strict aseptic technique with the line, (ii) the use of a dedicated lumen on a CVC or a PICC line, and (iii) use of antimicrobial-coated lines [20].

PN can be safely provided without a higher incidence of complications. Moreover, the presence of nutrition support teams may add additional benefits.

In addition, our study revealed that there were elevations in the ALT and urea in patients who received PN than EN groups. These pathologies are grouped together in the acronym parenteral nutrition-associated liver disease (PNALD) [21], which is a multifactorial complication. Sepsis and PN duration are modifiable contributing risk factors [22]. In contrary, a study concluded no significant difference between both groups as regard hepatic and kidney function [23]. So, strict measures should be applied on patients who were receiving PN to decrease the incidence of sepsis and beginning of EN as early as the patients stabilize to minimize the PN duration.

As regards survival analysis, our study concluded that mortality rate was lower in the EN group than the PN group. This is in agreement with Wong et al. study [24]. In contrary, other studies show no difference in mortality rate between both groups of nutritional support [16, 18]. This may be explained by high incidence of underweight together with complication such as sepsis and VAP in group who received PN in comparison with those on EN groups. So, this could be avoided by decreasing the duration of PN and beginning of EN as early as the patients stabilize to avoid the high rate of complication and consequently the high mortality rate.

Strict measures should be taken to manage sepsis occurrence properly as according to the logistic regression model in our study, sepsis is the most leading factor to mortality; this comes with Mathias et al., who concluded that the most common cause of pediatric death worldwide is sepsis [25].

Our study recommends nutrition should be personalized for each patient taking into consideration the state of disease and the present nutritional status of the patient. EN route should be started whenever possible. If EN is contraindicated, PN can be used with application (near-zero) PN complication strategies.

As study limitation, we did not study the outcomes of the patients shifted from PN to EN route.

Conclusion

Early EN was associated with low frequency of sepsis, VAP, and mortality than early PN, and early EN was associated with preservations of body weight with weight gain compared to PN. Early EN remains the preferred route for nutrient delivery.