© Springer International Publishing AG 2018

Mette M. Berger (ed.)Critical Care Nutrition Therapy for Non-nutritionistshttps://doi.org/10.1007/978-3-319-58652-6_1

1. General ICU Patients

Mette M. Berger¹  

(1)

Service of Intensive Care Medicine and Burns, Lausanne University Hospital (CHUV), Lausanne, Switzerland

Mette M. Berger

Email: mette.berger@chuv.ch

This pocket book is dedicated to the intensive care physicians who take care of the critically ill patients on a daily basis. Nowadays, most doctors are bewildered by the controversies that increase the uncertainty as to the optimal metabolic management of their patients. Orientation becomes even more important with the appearance of a new category of intensive care (ICU) patients, the chronic critically ill (CCI). This book attempts to provide a rational, physiology-based way to deal with the most common questions while signalling areas of controversy [1].

This first chapter will address the generalities such as criteria to identify the patients in need of artificial nutrition, defining their basic needs, the timing of an intervention and the general monitoring tools. Specific organ failures, as well as related needs and the caveats, will be addressed in the following chapters.

1.1 Definition of the Critical Care Patient

What defines critical illness? The patients are admitted to an ICU because of organ failure due to an overwhelming infection, trauma or other types of tissue injury that render them dependent on complex mechanical and pharmacological therapies. They present an intense inflammatory response, which is a coordinated cytokine-, hormone- and nervous system-mediated series of events that alter temperature regulation and energy expenditure. This in turn invokes neuroendocrine and hematologic responses, reorients the synthesis and disposition of several proteins in the body and dramatically stimulates muscle protein catabolism [2]. The resulting catabolic critical illness is a life-threatening condition that complicates the admission condition and further compromises quality of life and outcome.

Which are the criteria enabling the identification of the patients with an indication to a nutritional intervention? The bed and breakfast patients, i.e. those staying up to 72 h in the ICU and resume oral feeding rather easily, are obviously not the target population. Figure 1.1 provides some criteria that can assist selecting the patients in need of metabolic therapy.

Figure 1.1

Categories of patients and potential nutritional management (GRV gastric residual volume, ONS oral nutrition supplement)

Scores will assist a more precise definition of patients at risk of nutritional problems. The European nutrition risk screening score (NRS) [3] and British MUST (Table 1.1) are the easiest to use although they have not been validated for ICU patients, being developed as screening tools for general hospital patients. Nevertheless, according to the upcoming ICU guidelines of the European Society for Clinical Nutrition (ESPEN) [4], the NRS remains the simplest and fastest tool. In the NRS, an ICU admission results in three risk points (out of seven maximum); therefore, a nutrition-related alteration is required to create a real metabolic risk, which is the reason why ESPEN recommends considering five points as the risk level prompting therapy (Fig. 1.2). The Canadian NUTRIC score was designed as a specific critical care score and is still not prospectively validated: its computing is dominated by the weight of two ICU severity scores (APACHE and SOFA scores), includes no nutrition criteria and takes more time to complete [5], reasons why ESPEN does not recommend it as screening tool.

Table 1.1

Scores assisting the identification of patients at metabolic risk

Figure 1.2

Types of nutritional intervention based on screening criteria

Figure 1.2 proposes a strategy to integrate the anamnestic information and the NRS score to decide about a nutrition intervention.

1.2 Timing

For metabolic reasons, tree periods should be considered: (1) the early phase, i.e. the first 48 h, (2) the stabilization phase and (3) in some patients, the chronic-acute phase that starts after 2–3 weeks and may last for months and implies important changes in body composition. The majority of patients will leave the ICU by the end of the stabilization phase. Should we start feeding at a full regimen immediately?

There are two main reasons not to full feed immediately:

(a)

The endogenous energy and glucose production, as per below

(b)

The risk of inappropriate refeeding syndrome

Endogenous energy production: During the early phase in the absence of external supply (i.e. starvation), the body is able to produce glucose on its own for the glucose-depending organs by glycolysis and endogenous glucose production already after 12 h of feeding interruption [6]. The endogenous production is maximal during the first 48 h then abates: the amounts produced can only be measured by tracer techniques that are not available in clinical settings. Therefore, whatever the route, the administration of feeds should follow a progressive pattern to respect this endogenous response (Fig. 1.3), and thereby prevent early overloading with extrinsic substrates at a period which is characterized by elevated insulin resistance.

Figure 1.3

Conceptual presentation of optimal feeding strategy to avoid both overfeeding and underfeeding in critical illness (Reproduced with permission from Oshima et al. 2017 [6])

Refeeding syndrome risk: During complete or partial starvation, evolutionary adaptation has conferred the organism the above protecting mechanisms. Some adaptations occur very rapidly (within hours) such as the reduction of the endogenous insulin secretion and the consequent changes in the fluxes of electrolytes. The next step of shut down is more complex and occurs around the third day of starvation with increased ketone body production in healthy subjects, but not in critically ill patients, who are facing an intense catabolism to deliver amino acids for continued endogenous glucose production. Any glucose supply will induce a nearly immediate reversal of this strategy and prompt insulin secretion as well as its consequences on electrolyte movements. The progressive reintroduction of feeds enables monitoring of this response and supplying the required phosphate, potassium and magnesium supplements, preventing devastating effects [7].

1.3 What Are the Needs?

How should we determine the individual patient’s needs? What are the factors to consider? A frequently unsolved question is what is the patient’s weight? The preadmission weight is often unknown, and the actual weight sometimes obtained in the ICU is frequently artificially increased by fluid resuscitation. The pragmatic solution is to use the preadmission weight if known and an observer’s estimation of it in absence of such information.

As previously mentioned, it is essential to distinguish the very early phase (first 48 h) from the stabilization phase and the subacute-chronic phase which starts at around the end of the second week and may last for months.

Energy: This topic has generated major controversy. Multiple equations exist which have all been shown to be inexact compared to an indirect calorimetry, which is the gold standard [8, 9], but the latter equipment is not yet widely available. The least inexact equations are the Penn State University for ICU patients and the Toronto equation for major burns, both being derived from multiple indirect calorimetric determinations (Table 1.2). The simplest appreciation is to use 20 (first days) and 25 (stabilization) kcal/kg*day as target. A target guided by repeated indirect calorimetry will become the standard when upcoming simpler devices become available.

Table 1.2

Most common energy target equations

REE-HB Harris and Benedict estimation of resting energy expenditure, CI caloric intake, T °C temperature in Celsius, V·E minute volume (in L/min), height in cm, weight in kg

Proteins: The requirements should be dissociated from energy intakes, a differentiation which is difficult in clinical practice due to the fixed combinations of energy and proteins proposed by the industry. The World Health Organisation recommends 0.8 g/kg/day for healthy subjects. During the last decade, several studies have shown this amount to be insufficient for critically ill patients. The recommendations have been progressively increased to 1.3–1.5 g/kg*day [10]. Some categories of patients such as major burns have requirements as high as 2.0 g/kg*day based on isotopic studies (see Chap. 5): the elevated requirements of obese patients are discussed in Chap. 8 and those of renal failure patients in Chap. 9.

Carbohydrates: Several organs are strictly glucose dependent (brain, leukocytes) during the first 24 h of starvation, while others can adapt to a combination of substrates (heart, kidney, muscle, liver, adipose tissue). Too much glucose results in de novo lipogenesis, i.e. triglyceride synthesis at the liver level. The maximal tolerable glucose intake has been determined by tracer studies to be 5.0 mg/kg*min (i.e. 7.2 g/kg/day). Clinically, while 2.0  g/kg*day is the strict minimum requirement, doses of up to 4.0  g/kg*day cover needs without exposing the patient to overload.

Fat: Lipids are a necessary component of nutrition. In parenteral nutrition, the debate has been about the optimal combination of different types of fatty acids (n-3, n-9 and n-6 fatty acids) to achieve anti-inflammatory effects [11] and the total amount of lipids provided enterally and intravenously. In case of lipid-free parenteral nutrition (PN), essential fatty acid deficiency can be detected already after 5 days. Therefore lipids should be delivered with other substrates, the minimum daily amount being 0.5 g/kg*day to cover essential fatty acid needs, while an exact maximum has not been determined: the ESPEN guidelines recommend a total amount of fat not exceeding 1.5 g/kg*day.

Micronutrients: The requirements will depend on the severity of disease and route of feeding. While a dose corresponding to the recommended daily intake (RDI) is usually included in enteral feeds, this is by definition not the case with parenteral nutrition for stability reasons: micronutrients must be provided separately on a daily basis. The ESPEN guidelines underline the necessity to provide one daily dose of trace elements and vitamins for each day on PN [12]. Reinforcement of antioxidant defences with doses of micronutrients up to ten times the recommended PN has been associated with reduction of complications [13]. On the other hand, high-dose selenium monotherapy does not improve outcome [14].

1.4 Enteral and Parenteral Routes

The enteral route is to be preferred, whenever not contraindicated, for many non-nutritional reasons such as stimulation of gut immunity and maintenance of intestinal function. But in case of absolute contraindication (Table 1.3), the parenteral route is a valuable and safe alternative as shown by the most recent trials and meta-analysis [15]. Feeding by the enteral route should be initiated as early as possible to prevent losing it, i.e. within 24–48 h. The 2017 guidelines of the ESICM expert group recommend early enteral feeding with only five exceptions where delay is recommended (relative contraindication): active gastric bleeding, overt bowel ischaemia, gastric residuals >500 mL, abdominal compartment syndrome and high output intestinal fistulae [16].

Table 1.3

Absolute contraindications to enteral feeding

Using the gut is most successful when attempted within 24 h of admission and before the oedema from resuscitation affects the intestines and reduces their motility. This does however not mean that full feeding is to be achieved immediately (see Sect. 1.2).

Parenteral nutrition timing is still a matter of debate because of the negative results from studies carried in the 1980s and 1990s period, during which energy targets were much higher and glucose control nonexisting. Since the 2000, several large-size studies have been published showing that the outcome after PN compares to EN provided overfeeding is avoided [17]. Two recent large RCTs have shown equipoise between EN and PN when using early rapid progression to energy targets set by equations [18, 19]. Nevertheless, considering