Impact of Intravenous Fluids and Enteral Nutrition on the Severity of Gastrointestinal Dysfunction: A Systematic Review and Meta-analysis

Gut dysfunction (GDF) is a common problem in critically ill patients. It is the leading cause of multiple organ dysfunction syndrome/failure (MODS/MOF) and a significant cause of mortality and morbidity in critically ill patients [1]. In addition to this, the treatment of acute and critical illness can exacerbate GDF. Commonly used ICU interventions such as intravenous fluid resuscitation, early aggressive enteral feeding and vasopressor therapy are key factors leading to a secondary gut injury. In critical illness, intravenous fluid is the mainstay of early management for hemo-dynamic instability. It is vital to resuscitate a patient before commencing vasopressor therapy, particularly to delay the onset of an ischemic insult commonly occurring in hemodynamically unstable patients [2]. On the flip side, over-resuscitation can lead to bowel oedema leading to an ileus, while under-resuscitation with persistent splanchnic and peripheral vasoconstriction can trigger intestinal mucosal ischemia [3]. Although, enteral nutrition is the preferred approach to meet nutritional and modest fluid requirements in these patients, the delivery of early but aggressive enteral nutrition (EN) in hemodynamically unstable patients can precipitate the development of severe GDF, potentially leading to non-occlusive mesenteric ischemia which increases the chance of sepsis, multi-organ failure and mortality [4]. Intravenous fluid and enteral nutrition are two sides of the same coin and play a crucial role in determining the outcome of GDF if used wisely. However, very few studies have evaluated the role of these two modalities, thus making it difficult to understand their relationship with relevance to the severity of GDF. The aim was to review the evidence of the impact of intravenous fluid resuscitation and enteral nutrition individually on determinants of gut function and the implications in clinical practice.

The results of the meta-analysis demonstrate that restricted/goal-directed fluid therapy regardless of the type of fluid reduces mild GDF (vomiting) but not other complications associated with GDF. Enteral feeding, on the other hand, significantly increased vomiting episodes compared to parenteral nutrition but ‘early’ enteral nutrition significantly reduced the incidence of vomiting compared to delayed feeding. Enteral feeding was likely to reduce severe gut complications such as anastomotic leak and peritonitis compared with parenteral nutrition or an NBM status. Other mild to moderate variables of GDF (i.e. nausea, abdominal distension, ileus or diarrhoea) and moderate to severe complications (i.e. GI bleed, perforation, intestinal obstruction or intestinal ischaemia) were not associated with significant changes in outcomes. The results suggest that although the beneficial effects of restricted/goal-directed intravenous fluids and enteral feeding are essential to reduce some form of GDF, the impact is not prevalent for other variables of GDF (e.g. ileus and intestinal ischemia) associated with poor clinical outcomes. This may reflect the paucity of high-quality literature on the interaction between intravenous fluid (resuscitation) and enteral feeding as a combined therapy on the impact of GDF. The role of these two modalities in combination should be regarded as an important aspect in identifying the impact on the severity of GDF in acute surgical and critically ill patients.

Intravenous fluid therapy is frequently the first line of treatment in acute surgical and critically ill patients but hypervolemia and hypovolemia, both, are deemed detrimental. A revival of interest emerged almost two decades ago when hypovolemia in the form of restrictive fluid therapy was associated with improved post-operative clinical outcomes [9,16,18,31]. These studies suggested that a preferred approach of ‘zero’ or ‘neutral’ fluid balance not only improves outcomes related to gut motility but also may prevent adverse long-term outcomes. The current study demonstrated that mild GDF, i.e. vomiting, was significantly lower in patients on a restrictive/ targeted intravenous fluid regime. Studies have also reported similar results when colloids have been administered postoperatively [14,17]. The benefit of this outcome may be explained by cumulative administration of smaller volumes (of colloids) compared to crystalloids. Hypervolemia from excessive or liberal fluid administration, particularly crystalloids is associated with poor outcomes in postoperative [9,18] and in critically ill patients [25,47]. It can precipitate intestinal oedema leading to an ileus, delayed gastric emptying, feeding intolerance and hence sub-optimal nutrition delivery. Another school of thought indicates that complex surgical patients with high-risk surgeries possibly require judicious amounts of fluids to avoid complications associated with circulatory failure and gut mucosal ischemia [41,54,116,119]. This may be particularly relevant when liberal intravenous fluids are necessary to resuscitate patients after massive haemorrhagic losses for haemodynamic stability. In recent decades, goal-directed fluids have been advocated to prevent tissue hypovolemia [20] but maintain euvolemia by using targeted fluid approach raising the possibility of improved clinical outcomes in high-risk patients [22, 42, 84,121]. Hence, it is expected that a modest amount of fluids might be necessary to prevent anastomotic hypoperfusion, gut mucosal ischemia and reduce postoperative complications. Although the benefit of goal-directed fluid therapy is projected at improving organ perfusion without the onset of tissue oedema [54,117,122]; a paucity of studies exists warranting more research in this area [15, 41, 52, 118].

Enteral Nutrition forms an integral part of overall fluid administration in addition to intravenous fluids. Enteral nutrition and intravenous fluids combined play a crucial role in GDF outcomes, but due to a paucity of studies, this area has not received due attention. Enteral nutrition is invariably the first choice of nutrition compared to parenteral nutrition over decades [112]. The current study demonstrated that mild GDF, i.e. vomiting significantly increased in patients receiving enteral nutrition but reduced significantly when enteral feeding was commenced earlier. This is possible because ‘early’ enteral nutrition has multiple advantages over parenteral nutrition [75,82,114,121], and these benefits are evident in high-risk surgical and critically ill patients [70,124]. The initiation of enteral feeding is known to stimulate gut motility which reduces the incidence of GDF symptoms such as nausea and vomiting postoperatively. However, a significant difference for ileus between groups was not observed, although the number of events were lower in the enterally fed group. In cases of gut failure, when enteral feeding is contraindicated, parenteral nutrition becomes the sole choice of feeding and may be commenced within 24 hrs of ICU admission or post-surgery [120]. Administering parenteral nutrition appears to be a logical clinical decision, especially if enteral feeding raises the suspicion of non-occlusive mesenteric ischemia in the critically ill, with haemodynamic compromise. Our review showed no differences for intestinal 'ischaemia' between groups, although the events were half in the control group compared to the intervention (enteral nutrition) arm. Considering that the current review included a heterogeneous mix of patients, it is evident that in a sub-set of patients, i.e. post-cardiac surgery, severe acute pancreatitis or septic shock, administration of early enteral nutrition may potentially pose more risk than benefit by increasing the risk of bowel ischemia.

Nevertheless, the use of trophic enteral feeding has been suggested in haemodynamically unstable patients to maintain gut integrity [4]. Authors have argued that enteral nutrition comes with its risks such as aspiration, pneumonia, intestinal obstruction, necrosis and pneumonitis intestinalis. However, the present study demonstrated no such differences for any of these complications. For gastrointestinal complications, a significant reduction in anastomotic leaks in the enteral nutrition group suggesting its benefits irrespective of the feeding route was observed. It is common practice in some areas, particularly intensive care, to commence patients on parenteral nutrition with anastomotic leaks before a trial of enteral nutrition. However, it should be acknowledged that a correct assessment for an enteral nutrition challenge can be countered in patients on parenteral nutrition with significant complications (e.g. anastomotic leaks), hence lowering the threshold of initiating enteral nutrition. Barlow et al. [93] found a lower incidence (2 vs 7) of anastomotic leaks in the early enteral nutrition group. They attributed a three-day shorter length of stay and reduced postoperative complications from installing early enteral nutrition. A similar effect was confirmed by a Cochrane review [115] in which enteral nutrition reduced the risk of anastomotic leaks from 27% in the standard group to 13% in early enteral group. These results affirmed with the present findings. It is hypothesised that enteral nutrition may improve perfusion at the anastomosis site, which promotes mucosal wound healing and prevents further leaks.

In comparison, Lewis et al. (2009) did not support this finding and observed mortality of 50% in the intervention group (enteral group) with anastomotic leaks [114]. However, it is likely that a smaller sample size may result in a false positive rate for mortality, thus exaggerating the magnitude of the negative result. The benefit of enteral feeding in complications such as perforation and peritonitis has been confirmed by several reports, which resonated with our findings. Early enteral feeding seems to maintain gut integrity by improving mucosal circulation and oxygen delivery that may reduce the risk of peritonitis [74, 87,113].

The present study is not without limitations:

Further, inconclusive results from large-scale fluid and enteral feeding trials raise the suspicion that GDF may be the missing link, which perhaps may be associated with long-term outcomes. This dimension is often ignored when evaluating endpoints. To observe a difference in the key outcome, we first need to understand the combined effects of intravenous fluids and enteral nutrition in influencing clinical outcomes, including GDF. It is expected that as a result of the potential interaction between these two modalities, patients receiving liberal fluid resuscitation and early aggressive feeding are more likely to be at risk of severe GDF. More work is required to understand the implications of intravenous fluids and enteral nutrition on GDF and how this may impact overall patient outcomes. Future studies should evaluate this potential interaction and assess the combined impact of these two modalities on GDF in surgical and critically ill patients.

A restricted/goal-directed fluid regime and early enteral feeding compared to parenteral or a nil-by-mouth regime may reduce the risk on mild GDF in some, but not all complications of severe GDF. Because of a preventive strategy, we need to first understand the interaction between both (intravenous fluids and enteral feeding) and their impact on the gut so its implications can be translated into clinical practice eventually. Hence, it can be hypothesised that conservative fluid resuscitation and aggressive enteral feeding may potentially be the fundamental cause of developing severe life-threatening GDF (i.e. intestinal ischemia) and complications that can delay recovery and affect clinical outcomes in acute surgical and critically ill patients. Future research should evaluate and focus on an extended conceptual framework on the cross-interaction of conservative and aggressive modes across these two interventions and its impact on various levels of severity of GDF.