Introduction

Because of physiological and pathological differences, physicians face specific challenges in the treatment of the critically ill child. By providing insight into the respiratory and cardiovascular performance, hemodynamic monitoring plays a key role in the treatment of the critically ill child. To date, hemodynamic monitoring is generally limited to macrocirculatory parameters such as cardiac output and blood pressure. However, a well-functioning macrocirculation does not always guarantee a well-functioning microcirculation . Normal functioning of the microcirculation is pivotal for effective organ and tissue perfusion and oxygen delivery to prevent organ dysfunction . Research in children with respiratory failure and cardiovascular instability found microcirculatory alterations despite well-maintained systemic variables . Moreover, persistent microcirculatory alterations at admission could independently predict poor outcome in meningococcal disease and postresuscitation cases. Non-invasive monitoring of the microcirculation in children is possible with hand-held video-microscopes. However, monitoring of the microcirculation cannot yet be used routinely in the clinical settings.

The concept of hemodynamic coherence (HC) describes the relationship between the microcirculation and the macrocirculation as follows: When HC is present, improvements in the macrocirculation will lead to improvement of the microcirculation. The concept of HC provides a framework for bedside decision-making, aiding clinicians when HC is lost and improvements of macrocirculatory parameters are not reflected by an improvement in the microcirculation . The concept of HC has been retrospectively shown and several interventions improving the microcirculation in children have been described. However, several issues need to be addressed before the concept of HC can be translated to the bedside. Until then, the concept of HC does provide a framework for future research.

The critically ill developing child

The cardiovascular system in neonates and children is a dynamic changing system in function of age, weight, body proportions, metabolism, and compensatory mechanisms compared to that of adults. These differences in physiology and pathophysiology make the treatment of developing children particularly challenging.

The postnatal cardiovascular system changes dramatically after birth when the pulmonary vascular resistance decreases and the blood flow through the pulmonary vessels increases . The increase in return of pulmonary blood in the left atrium leads to a higher pressure and the mechanical closure of the foramen ovale. The removal of the low resistance placental circulation leads to an increased systemic vascular resistance (SVR). With closing of the ductus arteriosus and the ductus venosus and shrinkage of the umbilical vessels, the cardiovascular physiology starts to mimic that of adults. In neonates, the ventricular diastolic functioning is still suboptimal due to less compliant ventricles. This condition has negative influence on the effect of inotropes and fluid challenges . The heart needs several years to adapt to the new preload and afterload conditions, which increases the inotropic reserve capacity. Important cardiovascular parameters change during the first years of life. After a sharp increase in SVR, a more gradual decrease occurs, which ends around the age of 5 years. Similarly, the stroke volume index increases to adult values, until around 13 years of age. The cardiac index reaches adult levels after the child reaches the age of 10 years . Like the macrovascular development, changes also occur in the microcirculation. Top et al. found that the functional capillary density (FCD) of the buccal microcirculation decreases 1 week after birth in term neonates . The FCD represents the density of vessels containing red blood cells (RBC). The same seems to happen in preterms, judging from a decrease of skin FCD over the first month of life .

Hemodynamic instability is frequently observed in children with sepsis, following surgery, especially cardiac surgery, trauma and respiratory insufficiency. In these circumstances, when normal physiology is under pressure, differences in compensatory mechanisms compared to adults become more pronounced. The energy consumption of critically ill children, for example, does not resemble the hypermetabolic phases found in adults . Cardiac function in neonates and infants is characterized by an increased contractile state, increased sensitivity to afterload, and increased oxygen demands at an increased heart rate or preload state, which is less clear for older children . While adults in shock usually present with a high cardiac index and low SVR, children can present similarly with warm shock, with high cardiac output (CO) and SVR. Children may also present with cold shock, with low CO and high SVR. They can lose up to 35% of their circulating volume before blood pressure drops . Hypotension is a late sign of circulatory failure in children and is however rapidly followed by deterioration of the child’s condition.

The critically ill developing child

The cardiovascular system in neonates and children is a dynamic changing system in function of age, weight, body proportions, metabolism, and compensatory mechanisms compared to that of adults. These differences in physiology and pathophysiology make the treatment of developing children particularly challenging.

The postnatal cardiovascular system changes dramatically after birth when the pulmonary vascular resistance decreases and the blood flow through the pulmonary vessels increases . The increase in return of pulmonary blood in the left atrium leads to a higher pressure and the mechanical closure of the foramen ovale. The removal of the low resistance placental circulation leads to an increased systemic vascular resistance (SVR). With closing of the ductus arteriosus and the ductus venosus and shrinkage of the umbilical vessels, the cardiovascular physiology starts to mimic that of adults. In neonates, the ventricular diastolic functioning is still suboptimal due to less compliant ventricles. This condition has negative influence on the effect of inotropes and fluid challenges . The heart needs several years to adapt to the new preload and afterload conditions, which increases the inotropic reserve capacity. Important cardiovascular parameters change during the first years of life. After a sharp increase in SVR, a more gradual decrease occurs, which ends around the age of 5 years. Similarly, the stroke volume index increases to adult values, until around 13 years of age. The cardiac index reaches adult levels after the child reaches the age of 10 years . Like the macrovascular development, changes also occur in the microcirculation. Top et al. found that the functional capillary density (FCD) of the buccal microcirculation decreases 1 week after birth in term neonates . The FCD represents the density of vessels containing red blood cells (RBC). The same seems to happen in preterms, judging from a decrease of skin FCD over the first month of life .

Hemodynamic instability is frequently observed in children with sepsis, following surgery, especially cardiac surgery, trauma and respiratory insufficiency. In these circumstances, when normal physiology is under pressure, differences in compensatory mechanisms compared to adults become more pronounced. The energy consumption of critically ill children, for example, does not resemble the hypermetabolic phases found in adults . Cardiac function in neonates and infants is characterized by an increased contractile state, increased sensitivity to afterload, and increased oxygen demands at an increased heart rate or preload state, which is less clear for older children . While adults in shock usually present with a high cardiac index and low SVR, children can present similarly with warm shock, with high cardiac output (CO) and SVR. Children may also present with cold shock, with low CO and high SVR. They can lose up to 35% of their circulating volume before blood pressure drops . Hypotension is a late sign of circulatory failure in children and is however rapidly followed by deterioration of the child’s condition.

Hemodynamic monitoring in children

A number of monitoring techniques, both invasive and non-invasive, aid the physician when caring for the hemodynamically compromised child . Hemodynamic monitoring can help detect the lost equilibrium between oxygen supply and demand. At a macrocirculatory level, CO plays a central role in oxygen delivery. CO can be assessed with pulmonary artery catheters, which, although the golden standard, are invasive and labor intensive, and yield disputable results in adults . Moreover, catheters are hard to insert in small children and findings are difficult to interpret in children with cardiovascular shunting. The noninvasive bioimpedance technique, Doppler echocardiography and the more invasive indicator dilution methods are not commonly used in children due to inaccuracy, the requirement of excessive training, and intrusiveness . Other methods like electrical cardiometry and arterial pulse pressure recordings are not yet sufficiently validated in children . Additionally, CO can be assessed with clinical indicators, namely heart rate, blood pressure, pulsations, central venous pressure, urine output, and end-tidal CO ₂ . Although considered routine procedure, none of these indicators are solid enough for clinicians to predict hemodynamic status or fluid responsiveness in children .

Oxygen delivery can be assessed with biochemical markers such as serum lactate, mixed venous oxygen saturation (SvO ₂ ), and central venous oxygen saturation (ScvO ₂ ). Abnormal lactate levels after cardiac surgery may indicate decreased oxygen delivery and can predict poor outcome .

However, increased serum lactate is often a late sign of decreased oxygen delivery. SvO ₂ and ScvO ₂ are useful markers for oxygen consumption after pediatric cardiac surgery and decreased values are strongly associated with adverse outcome; however, measurements of SvO ₂ require a pulmonary artery catheter . Additionally, targeting ScvO ₂ alone in goal-directed therapy did not improve outcome in critically ill children . Only together with other parameters, intermittent use of ScvO ₂ was likely to reduce mortality and improve organ dysfunction in children in septic shock .

Clinical indicators of perfusion of the skin are capillary refill time (CRT) and temperature gradients . Decreased perfusion of the skin vasculature can cause a prolonged CRT and increased temperature gradients. In contrast to the emergency department, in the ICU, CRT shows no correlations with cardiac index, central venous pressure, stroke volume index, and SVR in children after cardiac surgery . Temperature gradients are a useful tool to assess changes in peripheral blood flow, but their reliability is diminished by the use of vasoactive drugs, cold ambient temperature, and (therapeutic) hypothermia and their relationship to systemic parameters are unclear . Another way of assessing the metabolic needs of tissues is with tissue pCO ₂ , derived from techniques using electrodes, probes, and tonometry. Increase of tissue pCO ₂ is the repercussion of a mismatch between tissue perfusion and metabolism . Multiple techniques are available to assess tissue pCO ₂ , but none has been sufficiently validated in children .

Macrocirculatory monitoring techniques may be inadequate to detect altered tissue perfusion and oxygenation, especially when HC is lost. Further, the abovementioned techniques can seldom predict deterioration. With monitoring of the microcirculation, loss of HC can be directly visualized. Thus, combining microcirculatory monitoring and HC can be a valuable addition to the existing monitoring techniques.

Bedside monitoring of microcirculation and hemodynamic coherence in children

The microcirculation plays an essential part in the delivery of oxygen and nutrients to tissues and cells. It consists of all vessels smaller than 100 μm, i.e. arterioles, capillaries and venules. Starting with orthogonal polarization spectral (OPS) imaging, sidestream dark field (SDF) imaging and more recently incident dark field (IDF) imaging are video-microscopy techniques that allow the direct visualization of the microcirculation on tissue and organ surfaces . With handheld cameras, perfusion of individual vessels of different tissue surfaces can easily be determined at the bedside. In preterms, microcirculation of the skin can be investigated . In term neonates and children, the buccal and sublingual areas are the most investigated and easily accessible surfaces , but avoiding movement artefacts is challenging in the awake and uncooperative child, in particular at toddlers age.

Variables of vessel density, microvascular blood flow, and heterogeneity of blood flow can easily be visualized and assessed to judge the state of tissue perfusion. In 2007, consensus was reached on the variables to quantify the microcirculation . An update of this consensus is being prepared. Vessel density can be quantified by total vessel density (TVD), the density of all vessels, and FCD. Perfusion can be determined with the proportion of perfused vessels (PPV) and microcirculatory flow index (MFI), a semi-quantitative measurement of the predominant flow pattern. Lastly, heterogeneity of blood flow between the different measured areas can be calculated with the heterogeneity index (HI). All of these have been used in neonates, infants, and older children .

Loss of hemodynamic coherence in critically ill children

Research in adults has revealed multiple pathological circumstances where microcirculatory alterations were found despite normal systemic hemodynamics. Ince described the dissociation of the macrocirculation and microcirculation as loss of HC and distinguished four types, all showing a decreased FCD . These conditions are illustrated in Fig. 1 and have also been observed in children.

Microcirculatory alterations associated with loss of hemodynamic coherence. Image adapted from . Type 1: Heterogeneity of microcirculatory perfusion as seen in sepsis, with perfused vessels found next to the obstructed vessels without blood flow. This results in heterogeneity of oxygen diffusion distances, impeding oxygen delivery. Type 2: Dilution of microcirculatory blood due to hemodilution, reducing the number of RBC in the capillaries and resulting in less oxygen delivery. Type 3: Constriction or tamponade because of manipulation of systemic hemodynamics with vasopressor therapy or therapeutic increase of the fraction of inspired oxygen. Type 4: Edema due to fluid resuscitation. Fluid administration can lead to tissue swelling, which increases the oxygen diffusion distance and interferes with oxygen delivery.

To describe HC, a limited combination of the following variables has been proposed to describe microcirculatory oxygen delivery: tube hematocrit (Ht), discharge Ht, FCD, and HI . Tube Ht describes the volume of RBC in the capillaries; discharge Ht combines flow with tube Ht to describe the delivery of oxygen per capillary, portraying the convection of blood through the capillaries . FCD represents the density of vessels containing RBC, which is related to the diffusion distance of oxygen between capillaries and cells. HI describes the heterogeneity of flow in perfused vessels in a region, which influences oxygen delivery . Table 1 describes the different types of loss of HC using these variables.

The first type of loss of HC is characterized by heterogeneity of microcirculatory perfusion. Here perfused vessels are found next to obstructed vessels without blood flow, resulting in functional shunting . Just like multiple studies have shown this heterogeneity in septic adults, this type has also been found in children with sepsis. Paize et al. observed 20 children with meningococcal septic shock with SDF imaging . Compared to controls, these children showed an abnormal flow pattern and a significant reduction in MFI, TVD, PPV and FCD despite normal cardiac indices. The severity of illness, as measured by the Pediatric Risk of Mortality III score, the Pediatric Logistic Organ Dysfunction score, and the Glasgow Meningococcal Septicaemia Prognostic Score, was related to the degree of abnormality of MFI. Here, clinical recovery was associated with microcirculatory improvement, and before extubation, the microcirculation was similar to that of controls. Top et al. also found circumstantial evidence for this type of persistent microcirculatory alterations in children with sepsis . With OPS imaging, 21 children with septic shock were measured in the buccal area, showing a persistent decrease in FCD and MFI in non-survivors. In contrast, survivors showed higher, yet still decreased FCD and MFI compared to non-survivors. FCD and MFI improved during clinical recovery. Although not included as a measured variable, these studies described blood flow as abnormal. This heterogeneity of microcirculatory perfusion results in heterogeneity in oxygen diffusion distance, impeding oxygen delivery.

Type 2 is defined by hemodilution, found for example after cardiac surgery. Nussbaum et al. found significant microcirculatory alterations in children after cardiac surgery on cardiopulmonary bypass . By visualizing the microcirculation in the ear conch of 40 children with SDF imaging, they found a significant decrease in MFI and PVD after surgery, regardless of normal systemic hemodynamics. These variables normalized to baseline values within 24 hours. These alterations can partly be explained by perioperatively-administered fluids. By correcting hypovolemia with crystalloids, hemodilution reduces the number of RBC in the capillaries, reflected by a lower Ht, resulting in less oxygen delivery.

Stasis or tamponade in the microcirculation has been described as type 3 loss of coherence. This can result from manipulation of systemic hemodynamics with vasopressor therapy or therapeutic increase of the fraction of inspired oxygen. Buijs et al. explored the effects of catecholaminergic drugs on the buccal microcirculation in 28 children with congenital diaphragmatic hernia requiring inotropic support using SDF imaging . Microcirculatory alterations were present before start of catecholaminergic therapy. The researchers demonstrated that dopamine has a positive effect on blood pressure and heart rate in children with congenital diaphragmatic hernia, but does not improve any of the microcirculatory variables. A subgroup of patients also received norepinephrine or epinephrine. Again, blood pressure and heart rate increased, but microcirculatory variables did not improve. This study showed that recovery of systemic hemodynamics in children does not always simultaneously improve the microcirculation. This may be explained by norepinephrine-induced stasis in the microcirculation, which prevents improvement of the microcirculation when the macrocirculation improves. In adults treated with norepinephrine microcirculatory blood flow could not be restored and even showed a trend to decrease . Type 3 alterations were also found in adults receiving oxygen therapy who developed hyperoxia, but this has not yet been observed in children .

Lastly, type 4 loss of coherence is characterized by edema due to fluid resuscitation. The importance of early aggressive fluid resuscitation in pediatric septic shock has already been established 25 years ago, and this approach has since been one of the principal therapies in the treatment of septic shock . Recent studies however have shown the dangers of this approach and have questioned the effectiveness of fluid administration in children . Maitland et al. randomly assigned 3141 African children with severe sepsis due to malaria to receive or not receive fluid boluses . Children receiving fluid boluses developed pulmonary edema and increased cranial pressure. More importantly, a liberal fluid strategy increased the 48-hour mortality rate in these children. These findings need not be the reason to abandon fluid resuscitation, but the gaps in our understanding of fluid responsiveness and the possible adverse consequences of fluid therapy need our attention. At a microcirculatory level, fluid administration can lead to tissue swelling, which increases the oxygen diffusion distance and interferes with oxygen delivery.