INTRODUCTION
In critically ill burns patients in the intensive care unit (ICU), the nutrient needs are higher than patients who are recovering from trauma or major surgery. This is due to a post-burn hypercatabolic and hypermetabolic state and the additional need for nutrients required for healing burn wounds and skin grafts.(1) Although the benefits of early and ongoing enteral nutrition (EN) in seriously ill patients are well documented, there are certain risks and complications associated with enteral feeding.(1) Administering nutrients via the enteral route increases the oxygen requirement of the gastrointestinal mucosa and the intestinal villa. If the increased demand for oxygen is not met, mucosal necrosis may occur.(2) Additional challenges exist in haemodynamically unstable burns patients on intravenous vasopressor support (e.g. adrenalin) to increase the cardiac output and blood pressure. In these cases, the vasoactive substance can decrease the splanchnic blood supply to the gastrointestinal tract,(3) resulting in hypo-perfusion, thereby increasing the chances of tissue hypoxia and intestinal ischemia.(4) In addition, Wells stated that vulnerability of the intestine to tissue death/necrosis is increased when patients who receive vasopressors, have experienced shock/trauma, or have undergone surgery or are severely burnt.(4) Moreover, this report emphasized that when EN is administered via jejunal or naso-jejunal routes, the patients are especially vulnerable to tissue death/necrosis in the intestines.(4)
Although the aetiology of bowel necrosis (intestinal ischemia) in the presence of enteral feeding is not clear, associations with patients who had undergone surgery or those who had undergone laparotomy exist.(5) Another suggestion was that the impact of EN on the gastrointestinal tract (GIT) is paradoxical, because the increased supply of nutrients increases the splanchnic blood flow. However, the absorption of these nutrients also increases mucosal oxygen requirements and when this increased demand for oxygen is not met at intestinal mucosal level, it can result in bowel necrosis(2,3). The above views on the aetiology of bowel necrosis justify further research on EN in burn patients who are haemodynamically unstable and on vasopressors. Although research has confirmed the association of non-occlusive bowel necrosis (NOBN) with early enteral nutrition, the causal relationship between these has not been verified, contributing to the lack of consensus regarding the concomitant use of EN and vasopressors in haemodynamically compromised burns patients. The decision to commence EN in such patients thus remains controversial.(1) This places clinicians in a predicament as to when to start EN in the hemodynamically compromised and unstable burns patient receiving vasopressor support.(1)
However, Scaife et al. reviewed the complications associated with tube feedings and found bowel obstruction and ischaemic necrosis to be an uncommon complication after early and intensive enteral feeding.(6) Similarly, another report found that when EN is started in the first 2 to 3 days of admission to the ICU, it was not associated with intestinal ischemia. The latter studies also noted that in most of the NOBN cases described in the literature, the EN was not given early and moreover, specific details pertaining to the hemodynamic condition of the patients at the start of EN, the mode of EN administration, and details pertaining to the daily monitoring of the clinical condition of the patients, were not provided.(3)
In contrast to the risk of intestinal mucosa ischaemia, intolerance to enteral feeding is not an uncommon problem. This is characterised by delayed gastric emptying, vomiting, diarrhoea and abdominal distention. Intolerance to EN is frequently encountered in patients who are seriously ill and the introduction of EN in these patients must be gradual and carefully monitored.(4) Despite these concerns, Wells suggests that providing EN together with the minimal effective dose of vasopressors and careful monitoring for feeding intolerance will present ‘very little risk for bowel necrosis’.(4) Indeed, only a small number of reported cases of bowel necrosis linked to the simultaneous use of EN and vasopressors in the haemodynamically unstable patient exists.(4)
The aim of this study was to determine whether EN should be continued in the haemodynamically unstable burns patient on high-dose intravenous vasopressor support, which is the current protocol in the Chris Hani Baragwanath Academic Hospital (CHBAH), adult burns unit (ABU).
PATIENTS AND METHODS
Patient data collection
A retrospective medical record review was conducted at the ABU of the CHBAH, Soweto, South Africa. Adult burn ICU patients who received concomitant EN and vasopressor support for at least two hours or more were included in this study. A minimum of 2 h was chosen since there is no consensus in the literature on the time of simultaneous inotropic support and EN.
The ABU patient register that indicates when a burns patient was admitted and discharged and whether the patient was in the ward or the ICU was used to locate the ICU patients admitted from 1 September 2010 to 1 May 2016. Only 44 patients (9%) satisfied the inclusion criteria for entry into this study, namely age (10 years and older), total burnt surface area (TBSA) >20%, inhalation injury requiring mechanical ventilation and patients who required both ventilation and intravenous vasopressor support. The exclusion criteria were patients on oral feeds and/or parenteral nutrition.
Feeding
The ABU ICU feeding protocol was commenced on admission to this unit. If there were no contraindications to EN, nasogastric or orogastric tube feeds were commenced. A nutritionally complete, peptide-based liquid (Fresenius Kabi, Suvimed OPD HN) with a caloric content of 665 kJ, 33.5 g protein, 91.5 g CHO and 18.5 g lipids (0.3 eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA)/500 ml) in 500 ml plus the average adult recommended daily requirements for vitamins, minerals and trace elements was initiated. The patient's caloric and nutritional requirements were assessed, calculated, recorded and adjusted as necessary by a registered dietician as per the standard operation procedures.
Feeding intolerance was identified by high gastric aspirates (≥500 ml/6 h), vomiting, diarrhoea and abdominal distension. In the absence of these signs in the first 24 h of receiving the peptide-based feed, this was changed to a non-peptide-based liquid feed (Fresenius Kabi high protein energy) with 750 kJ, 37.5 g protein in 500 ml and the vitamin, mineral, trace element supplement.
In patients with early signs of feeding intolerance, the volume of the feed was immediately reduced to 20 ml/h increasing slowly every 6 h until the prescribed volume for the individual patient was achieved. Monitoring with repeated gastric aspirates and abdominal examination remained mandatory for these critically ill patients according to the unit's standard operating procedures.
This study was approved by the Human Research Ethics Committee of the University of the Witwatersrand.
RESULTS
Data analysis
Statistical analysis was done using the STATA Version 14.2 (College Station, Texas, 77845, USA) statistical programme. The means and standard deviations of the continuous variables, age and TBSA were determined. Frequencies and proportions were calculated to describe the distribution of patient gender, age category, TBSA burns category and feeding intolerance.
For the categorical variables, a two-sample test of proportions was used to assess the equality of proportions. For numerical variables such as age and TBSA%, a Wilcoxon signed-rank test was used to determine the P-values. A P-value of <0.05 with a 95% confidence interval was considered statistically significant.
Demographics and burns
The number of burns patients during this period totalled 1109, of which 475 patients (43%) were admitted to the ICU. Of the 475 critically ill patients admitted to the ICU, only 44 met the study inclusion criteria. There were 18 female and 26 male patients in the study cohort with a mean age of 36.5 years (10–83), with female patients at 37.5 years (range 33–53) being somewhat older than males 33.5 years (range 26–43) (Table 1).
Summary of measures of patient characteristics.
| Characteristics | Sex | N | Median (IQR)* | Min. | Max. |
|---|---|---|---|---|---|
| Age (years) | Female | 18 | 37.5 (33–53) | 25 | 83 |
| Male | 26 | 33.5 (26–43) | 14 | 65 | |
| Total | 44 | 36.5 (28–52.5) | 14 | 83 | |
| TBSA Burn (%) | Female | 18 | 26.5 (20–36) | 10 | 63 |
| Male | 26 | 39 (22–50) | 11 | 70 | |
| Total | 44 | 33 (22–50) | 10 | 70 |
IQR: Interquartile Range
Enteral feeding
The patients were on inotropes for as long as they were haemodynamically unstable, after which they were weaned off inotropic support as they became haemodynamically stable (Table 2).
Clinical signs of feeding intolerance.
| Feeding intolerance | Number of patients |
|---|---|
| Moderately high gastric aspirates (>200 ml/6 h) | 11 |
| Vomiting | 4 |
| Diarrhoea | 9 |
| Acute abdomen | 0 |
| Distended abdomen | 0 |
A total of 24 patients exhibited intolerance to EN but the intolerance was temporary in 21 patients and persistent in 3. Thus, of the 44 patients in the study group, 41 (93%) tolerated full enteral feeding after a temporary intolerance while receiving intravenous vasopressor support (Table 3). There was no significant demographic or clinical variable that correlated with feeding intolerance.
Comparison of participants proportions by feeding tolerance.
| Feeds stopped n(%) | |||||
|---|---|---|---|---|---|
| Characteristics | Level | No | Yes | Total | P-value |
| 41 (93.18) | 3 (6.82) | 44 (100) | |||
| Sex n(%) | Female | 17 (94.44) | 1 (5.56) | 18 (40.91) | 0.7822 |
| Male | 24 (92.44) | 2 (7.69) | 26 (59.09) | 0.7822 | |
| Median age (years) (IQR) | 34 (28–52) | 56 (42–78) | 0.056 | ||
| Median TBSA burn % (IQR) | 34 (22–50) | 28 (25–55) | 0.8339 | ||
| Gastric aspiration n(%) | <50 | 10 (100) | 0 | 10 (22.73) | 0.3305 |
| 50–100 | 6 (85.71) | 1 (14.29) | 7 (15.91) | 0.3927 | |
| 100–200 | 15 (100) | 0 | 15 (34.09) | 0.1969 | |
| >200 | 10 (83.33) | 2 (16.67) | 12 (27.27) | 0.1125 | |
| GIT upset n(%) | No | 32 (96.97) | 1 (3.03) | 33 (75.00) | 0.0842 |
| Vomiting | 1 (100) | 0 | 1 (2.27) | 0.7844 | |
| Diarrhoea | 6 (85.71) | 1 (14.29) | 7 (15.91) | 0.3927 | |
| Vomiting & Diarrhoea | 2 (66.67) | 1 (33.33) | 3 (6.82) | 0.0591 | |
DISCUSSION
There are no specific, clear-cut guidelines in the literature on providing EN to the haemodynamically unstable patient who is on vasopressors. Due to the differing opinions on initiating and withholding enteral feeds, the aim of this study was to determine whether severely burnt patients in ICU, on vasopressor agents, were negatively affected by administering enteral feeding.
The studies on enteral feeding that have been reviewed in the literature have reported on postoperative cardiac patients, general ICU patients or patients provided with post-pyloric tube feeding.(7) Thus, the current study is unique in that the study included only haemodynamically unstable burn patients, who all received high doses (0.5 µg/kg/min) of vasopressor support to reduce the risk of septic shock.
Khalid et al. assessed the outcome of EN on the seriously ill and haemodynamically unstable patients requiring mechanical ventilation and inotropic support, but all patients in that study were non-surgical.(8) The American Society for Parenteral and Enteral Nutrition and the Society of Critical Care Medicine recommend abstinence from enteral feeding in the unstable patient on high-dose catechol-amines.(9) These organisations have warned that enteral feeding will increase the oxygen requirement of the intestinal mucosa, while the presence of splanchnic hypoperfusion secondary to vasopressor agents, will add to the risk of intestinal mucosal necrosis.
Despite these potential risks and guidelines, the literature contains accounts of clinicians initiating feeds in an unstable patient.(2,4,8,10) Mancl et al. found that <1% of patients who received EN and simultaneous vasopressors developed intestinal ischemia and bowel perforation.(10) In support of Mancl et al., Turza et al. theorised that EN is a relatively safe intervention in the haemodynamically unstable patient on vasopressor therapy, provided that a comprehensive four-phase process is followed, which includes the following: (i) evaluating the dangers and benefits of providing EN, (ii) analysis of the patient's physiologic state while on vasopressors, (iii) once the benefits are determined to outweigh the risks of EN, the type of nutrition formula should be carefully chosen and (iv) after the initiation of EN, constant monitoring for clinical signs of distress is essential.(2)
A significant finding of the current study with regard to safety and tolerability is that the administration of vasopressors in critically ill, haemodynamically compromised burn patients is not contraindicated if early and ongoing EN is carefully monitored. From evaluation of the results of this study, it can be concluded that the practice of EN can be continued in the haemodynamically unstable burns patient on intravenous vasopressor support. Despite the small study sample, the finding that 93% (41 of 44 patients) were able to continue enteral feeds supports the conclusion of other studies to continue feeding the haemodynamically unstable patient who simultaneously requires vasopressor support.(2,8,10) No patients in this pilot study suffered thrombosis, ischemia or bowel necrosis. This suggests that the carefully managed administration of EN in the seriously ill, haemodynamically compromised burn patients who are on vasopressor support is a safe treatment regime.
CONCLUSION
As the extent of the burn increases, so does the nutritional requirement of the patient. This study of South African patients shows that carefully administered enteral feeding is a reasonably safe method of meeting the nutritional needs of severely burnt and haemodynamically unstable patients in the ICU.