Fluid Overload in Critically Ill Children

Background: A common practice in the management of critically ill patients is fluid resuscitation. An excessive administration of fluids can lead to an imbalance in fluid homeostasis and cause fluid overload (FO). In pediatric critical care patients, FO can lead to a multitude of adverse effects and increased risk of morbidity. Objectives: To review the literature highlighting impact of FO on a multitude of outcomes in critically-ill children, causative vs. associative relationship of FO with critical illness and current pediatric fluid management guidelines. Data Sources: A literature search was conducted using PubMed/Medline and Embase databases from the earliest available date until June 2017. Data Extraction: Two authors independently reviewed the titles and abstracts of all articles which were assessed for inclusion. The manuscripts of studies deemed relevant to the objectives of this review were then retrieved and associated reference lists hand-searched. Data Synthesis: Articles were segregated into various categories namely pathophysiology and sequelae of fluid overload, assessment techniques, epidemiology and fluid management. Each author reviewed the selected articles in categories assigned to them. All authors participated in the final review process. Conclusions: Recent evidence has purported a relationship between mortality and FO, which can be validated by prospective RCTs (randomized controlled trials). The current literature demonstrates that “clinically significant” degree of FO could be below 10%. The lack of a standardized method to assess FB (fluid balance) and a universal definition of FO are issues that need to be addressed. To date, the impact of early goal directed therapy and utility of hemodynamic parameters in predicting fluid responsiveness remains underexplored in pediatric resuscitation.


INTRODUCTION
In a critically ill patient, fluid balance is imperative in management and maintaining homeostasis. More often than not, patients are resuscitated with fluids to maintain adequate intravascular volume. A common issue seen in pediatric intensive care units is fluid imbalances and hemodynamic instability. Aggressive fluid administration can lead to fluid overload (FO), a condition in which there is a positive fluid balance in the patient. This phenomenon has been associated with a multitude of unfavorable effects and can further complicate the patient's condition. While FO itself is no indicator of mortality, the adverse effects of FO on an already at risk population puts them at an increased risk of morbidity and mortality. Research on this commonly practiced standard of care has shown that inundation is not always the best method for hemodynamic optimization. Current literature shows that there is a lack of a standard definition of fluid overload and a standard guideline by which to direct fluid therapy. The purpose of this study was to review the literature highlighting the impact of FO on a multitude of outcomes in critically-ill children, analyzing the causative vs. associative relationship of FO with critical illness, and addressing the current pediatric fluid management guidelines.

METHODS
The literature search was conducted using PubMed/Medline, and Embase databases for studies on fluid overload in pediatric critical care patients <18 years, from the earliest available date until June 2017. A total of 477 articles were obtained; after duplicate removal and review, 48 articles were selected for data extraction (Figure 1). The initial literature review consisted of two authors who independently reviewed the articles. The manuscripts of studies deemed relevant to the objectives of this review were then retrieved and associated reference lists handsearched. Articles were categorized into FO associated with renal replacement therapy, extracorporeal membrane oxygenation, cardiac surgery, neonates, respiratory tract disease, and sepsis. All relevant publications were considered for data extraction. Tables were created summarizing the outcomes of the included studies.

PATHOPHYSIOLOGY OF FLUID OVERLOAD
Current understanding of FO is contingent upon the revised Starling law which ascribes an important role to EGL (endothelial glycocalyx) in maintaining an intact vascular biology. EGL, a web of membrane-bound glycoproteins and proteoglycans on the luminal surface of endothelium, binds to plasma proteins and effectively excludes proteins from the sub-endothelial cleft, thereby resulting in a local oncotic gradient between plasma and cleft which opposes trans-capillary efflux (1). The law states that a constant oncotic gradient present along capillaries opposes pressure-mediated fluid efflux without ever causing fluid reabsorption from interstitium (except for vascular beds in renal tubules and intestines), and almost all vascular refilling from interstitium occurs via lymphatics (2). EGL breakdown can be triggered by various processes and stresses such as sepsis, surgery, postischemic states, and hyperglycemia (2)(3)(4)(5)(6)(7). Once the EGL is damaged, fluid efflux becomes primarily dependent on capillary hydrostatic pressure and the manipulation of plasma oncotic pressure with colloids does not encourage any vascular refilling. In addition to fluid and protein extravasation, EGL disruption potentiates ongoing inflammatory responses by exposing endothelial cell-adhesion molecules and may account for pathologies with interstitial edema such as acute lung injury (4). In experimental and perioperative settings, volume expansion has been shown to hasten glycocalyx disintegration and reduce lymphatic drainage through release of natriuretic peptides. In addition to cleaving membrane-bound proteoglycans and glycoproteins (most notably syndecan-1 and hyaluronic acid) off endothelial glycocalyx, they also reduce lymphatic drainage by inhibiting lymphatic propulsive motor activity (3,4). This worsens FO because interstitial fluid is primarily channeled back into circulation via lymphatics.
Of note, some data regarding glycocalyx and natriuretic peptides is partly theoretical, and partly derived from adult studies and there is no pediatric-specific data in this context (8)(9)(10).

QUANTIFYING FLUID OVERLOAD
Despite the burgeoning evidence regarding association between FO and adverse patient outcomes, there is no consensus on the optimal method for calculating FO. Lombel et al. highlighted that depending upon the definition used, there was a significant variability in the number of patients identified with >10% FO (11).
Most pediatric studies utilize the FB method introduced by Goldstein et al. (12). The formula is: %FO = daily fluid intake (L) −total output (L) Baseline body weight (kg) x 100 More practical WB(weight based) methods can be alternatively used to estimate degree of FO. The formulas are: %FO = CRRT initiation weight kg − hosptial admission weight (kg) hosptial admission weight (kg) x 100 or %FO = CRRT initiation weight kg − ICU admission weight (kg) ICU admission weight (kg) x 100.
These are particularly useful in neonates because of uncertainties in estimating fluid input via breastmilk, fluid lost into diapers/bedsheets, and insensible losses apart from losses which can be markedly influenced by a number of factors (13). ICU admission weights when taken as baseline might under-estimate degree of FO in comparison to hospital admission weight. This discrepancy may also delay CRRT (continuous renal replacement therapy) initiation as most clinicians still consider 10% as the threshold to intervene (14). Multiple studies in both adult and pediatric literature have demonstrated poor agreement between changes in body weight and FO ( Table 2). Nonetheless, reports have shown correlation between the two and a similar predictive potential of mortality (15). These inconsistent results warrant further research to recognize the most appropriate method of FO measurement.

IS FLUID OVERLOAD EVER INDEPENDENT OF AKI?
FO has detrimental effects on all organs, notably kidneys (2) (Figure 2). It is not just an epiphenomenon in AKI because FO per se renders an individual vulnerable to AKI and often predates it (23,24). The development of abdominal hypertension and abdominal compartment syndrome impairs renal perfusion, and can culminate in a vicious cycle of AKI and FO (25). Deferral of AKI diagnosis correlates with the degree of fluid overload ( Table 1). The risk of FO in this setting is that it can increase the volume of distribution of serum creatinine, a measure of AKI, leading to an overestimation of kidney function and a delayed diagnosis of AKI (26,27). In scenarios where FO was not accounted for, patients with a decreasing trend in sCr (serum creatinine) had increased mortality than patients with high sCr and severe AKI (28).
The following equation can be used to adjust sCr for fluid accumulation: adjusted Cr = sCr [1+ (Net fluid balance/Total body water)], where TBW = 0.6 × weight (kg) (29). However, use of the constant 0.6 in estimating TBW may not be accurate in all patients as it can be affected by age, sex, and various pathologic states (heart failure, cirrhosis, nephrosis).
Re-analysis of FACTT (Fluid and Catheter Treatment Trial) showed that presence of covert AKI was independently associated with a higher risk of death. Incidence of AKI was greater in the conservative arm before adjustment but in liberal arm post-adjustment (30). Secondary analysis of PICARD (Program to Improve Care in Renal Disease) cohort also revealed that correction for %FO uncovered missed diagnoses and reclassified AKI in some patients. Although no mortality differences were noted, the need for subsequent dialysis was higher in the misdiagnosed patients (27). Basu et al. reiterated these findings in 92 infants undergoing arterial switch operation wherein correction of sCr for FO not only increased AKI prevalence but also strengthened its association with postoperative morbidities (29).

EPIDEMIOLOGY AND OUTCOMES
Fluid Overload and Adverse Outcomes: Does Current Evidence Endorse Causality?
With the exception of select adult trials and recent findings of Grams et al. (31)(32)(33)(34)(35)(36), no pediatric study to date has demonstrated FO as a causal contributor to mortality rather than just a marker  (35). Few pediatric studies have in fact shown that FO is a significant co-morbidity that is highly associated with adverse outcomes only in patients with limited disease unlike patients with severe disease in whom the ultimate outcome is independent of %FO, suggesting that the disease severity itself is the primary determinant (37)(38)(39).

Renal Replacement Therapy
There are many indications for initiating renal replacement therapy (RRT) in a critically ill patient, FO being one of them. RRT provides control in fluid management and allows for the ability to achieve a negative fluid balance (40). Association between FO and mortality was first observed by pediatricians in critically ill children with AKI requiring RRT. Several pediatric studies have demonstrated an independent association between FO at the time of continuous renal replacement therapy (CRRT) initiation with mortality (41)(42)(43)(44)(45). A study by Goldstein looked at 116 pediatric intensive care unit patients with multiorgan dysfunction who received CRRT. The results showed a significant survival different in patients who were in a state of FO while on CRRT vs. those who are able to achieve their dry weight (36 vs. 76%) (43). The results from the ppCRRT Registry stratifying 297 patients into <10%, 10-20%, and >20% FO groups remain most comprehensive in this regard (41). Analysis revealed that presence of FO at CRRT initiation was associated with raised  -Correlations were strong between fluid input minus output/adjusted fluid input minus output and body weight change -Agreement between was poor between fluid input minus output/adjusted fluid input minus output and body weight change during the first 24 h after PICU admission -Since daily body weight is not particularly difficult to measure, fluid input minus output/adjusted fluid input minus output calculations may be reserved for the most severely ill patients in whom body weight measurement is strictly contraindicated Perren et al. (19) Correlation and Bland Altman agreement was poor between -body weight change and cumulative fluid balance in a cohort of ICU patients Mank et al. (20) -Body weight was deemed to be a more accurate, safe and reliable parameter for monitoring fluid retention in patients undergoing hyperhydration prior to chemotherapy -Correlation between body weight change and cumulative fluid balance was relatively low Eastwood (21) Body weight gain in post-cardiac surgery patients was falsely undermined by the fluid balance method  odds for mortality (aOR 1.03) after controlling for the severity of illness, with the maximum risk noted in the cohort with >20% FO(aOR 8.5,95% CI 2.8-25.7). The RENAL study (Randomized Evaluation of Normal vs. Augmented Levels of renal replacement therapy) showed that a negative mean daily FB was significantly associated with reduced mortality (p < 0.0001) and increased RRT-free days (p = 0.0017) (46). Non-surviving CRRT recipients (n = 610) in the PICARD study also had a significantly higher %FO at CRRT initiation (p = 0.01) and cessation (p = 0.004) (47). A subsequent re-analysis highlighted for the first time an increased risk of sepsis in patients after AKI. Authors speculated that FO compromises epithelial integrity by causing gut edema, thus favoring bacterial translocation (48).

Extracorporeal Membrane Oxygenation
Patients on ECMO are vulnerable to develop significant degree of FO due to the inflammatory milieu triggered by their critical illness, exposure to extracorporeal circuit and iatrogenic administration of fluids to minimize venous access insufficiency. The association between FO and mortality in children on ECMO has been reaffirmed in various pediatric and adult cohorts. Per literature, the need for initiating RRT arises in almost 50% ECMO patients (49,50). Selewski et al. recently showed that median FO at CRRT initiation and discontinuation were significantly lower in ECMO survivors (49). FO reduction to <10% at CRRT conclusion was not associated with improved survival, emphasizing that early intervention is the key to enhanced survival. Although a growing body of evidence favors concomitant use of CRRT in ECMO patients, reports on this issue have been contentious (51,52). In another study, Selewski showed that the degree of peak FO was associated with significant increases in duration of ECMO. Patients who died during hospitalization had higher FO at ECMO initiation, along with higher peak FO. Also noted was a graded increase in both ECMO and hospital mortality by 10% interval increases in FO ( Table 4) (40).

Cardiac Surgery
In neonates with congenital heart disease undergoing cardiac surgery, fluid management is of particular importance. These patients are administered large amounts of fluid pre, intra, and post-operatively to prevent hypotension. The mainstay of management in these patients is diuretic therapy, with furosemide being the most widely administered drug (16). FO in these patients can also result from AKI (incidence = 30-50%) and systemic inflammation stemming from cardiopulmonary bypass (CPB) (53). FO in this cohort is significantly associated with increased in-hospital mortality (15,54,55); longer ICU and hospital stay, prolonged inotrope and ventilator support, and delayed sternal closure (56)(57)(58)(59). Novel FO mitigating strategies, namely passive peritoneal drainage and modified ultrafiltration techniques are increasingly being employed. Given the high risk status of CHD patients, many are considered for preventative use of PD to reduce fluid accumulation. In some institutions the practice of inserting PD catheters intra-operatively is considered routine (16). Several studies have purported that prophylactic PD (peritoneal dialysis) achieves a net negative FB, ameliorates post-CPB inflammation via cytokine removal and reduces inotrope requirement (60)(61)(62). However, few single center studies do not reaffirm above findings (63). Most recent data on this population suggests that the potentially deleterious FO cut-off is likely <10% as studies have demonstrated equally poor outcomes with a FO level as low as 5% (54,59).

Neonates
Association between FO and BPD (bronchopulmonary dysplasia) has been an active area of research. Studies document an association between FB and weight loss in the first few weeks of life and subsequent development of BPD in preterm neonates (64)(65)(66). In the neonatal period, a physiologic contraction of the extracellular fluid compartment leads to post-natal weight loss and a negative fluid and sodium balance. When this process is hindered by excess fluid intake, the retention of extracellular fluid can lead to fluid accumulation in pulmonary interstitial tissue. This puts the neonate at risk of barotrauma leading to BPD (64). Askenazi et al. reported that neonates with AKI had a higher maximum weight change in first 4 days of life (p = 0.05), higher risk of death and were more likely to require prolonged MV (p < 0.03) and oxygen support than neonates without AKI even post-adjustment (67). These studies suggest that ameliorating FO and AKI can reduce mortality and their sequelae (BPD and chronic kidney disease) which incur long term morbidity and excessive medical costs ( Tables 3, 4). In the treatment of BPD, diuretics have been used; though studies have only shown short term effects without any long term benefits (80). Literature has suggested the possible benefit of practicing fluid restriction in preterm infants to prevent BPD. However, recent systematic review showed no evidence that supported fluid restriction in infants would prevent the subsequent development of BPD (81).

Acute Respiratory Distress Syndrome
Several studies have demonstrated association between FO and worse respiratory outcomes (72)(73)(74). The adult FACTT trial showed that mean CFB (p < 0.001), VFDs (ventilator-free days) and ICU-free days were significantly lower in the conservative arm (31). Similarly, Valentine et al. noted that daily FO on D1-3 and CFB on D1-7 were higher in children with FO compared to adults in FACTT conservative arm (p < 0.001, each day) but similar to adults in the liberal arm, and promulgated the use of a Bayesian pediatric trial mirroring FACTT. No association with mortality was noted (73). A simplified fluid restrictive strategy employed in "FACTT lite" study also led to significantly lower FO than FACTT liberal protocol and no difference in VFDs compared to FACTT conservative protocol (82). Interestingly, mortality has been shown to correlate with increasing CFB (46), but not to a restrictive fluid management (31,82).

USE OF EARLY GOAL DIRECTED RESUSCITATION-HAS THE IMPASSE ENDED?
The early goal directed therapy (EGDT) has been an area of intense research since its inception in the landmark Rivers' study wherein a mortality benefit was sought using protocolized (maintaining a CVP of 8-12 mmHg and SvO 2 ≥70%) fluid resuscitation within first 6 h in the treatment of septic patients (83). An early up-front aggressive resuscitation was effective but late and excessive fluid administration was associated with worse survival at hospital discharge, 28 and 60 days. The critical difference emphasized by this protocol was the timing of the administration since the overall amount of fluid administered was equal in the two groups: a corollary to the conclusions drawn by Arikan et al., who recently evaluated outcomes in a controlled before and after study using a protocol driven resuscitation bundle in children with septic shock (84). Even though the total volume of fluid boluses did not differ between the two groups, thoughtful fluid administration earlier during the course of illness decreased mortality, AKI incidence, and need for RRT in comparison to historical controls. This time-sensitive nature of shock resuscitation has been stressed upon many times due to evidence of higher mortality in children with a delay in reversal of shock (85)(86)(87)(88).
A number of studies, however, have questioned the efficacy of EGDT. Three multicenter RCTs-ProCESS, ProMISE and ARISE, did not show a statistically significant reduction in 90d mortality with EGDT but reported mortality was lower than the anticipated rate (89)(90)(91). Current surviving sepsis campaign guidelines (SSCG) do not endorse EGDT but still permit the use of previous targets as no harm was documented in these trials (92). Subsequent meta-analyses have yielded inconclusive results  and equipoise still exists (93). Nevertheless, recruitment of less sick patients, treatment crossover due to non-blinding, and early institution of aggressive antibiotic therapy could haves biased the results of these negative trials (94). Workman and colleagues recently showed that compliance with SSCG was not associated with better outcomes compared with usual therapy, which was administered more slowly in the emergency department (95). Since all children were treated rapidly, they reported low morbidity and mortality, and underscored the importance of rapid recognition and treatment of septic shock.

FLUIDS: TYPE IN CRITICALLY ILL CHILDREN?
In children with diabetic ketoacidosis, NS (normal saline) is the preferred resuscitating fluid but development of significant hyperchloremic acidosis necessitates a switch to hypotonic fluids such as 1/2NS or more physiological choices with lower chloride but high sodium concentrations such as LR (lactated Ringer's), Hartmann's solution, or plasmalyte (96). In pediatric acute respiratory distress syndrome, the PALICC recommends monitoring and titrating fluid balance to maintain adequate intravascular volume, while aiming to prevent positive fluid balance (97). Mechanical ventilation in acute lung injury reduces the amount of maintenance fluids, so in these patients the preferred choice is a dextrose containing solution with NS (or one of the other more physiological solutions) at two-thirds of the maintenance rate that is used recommended for nonmechanically ventilated patients (96).
In cases of pediatric sepsis, different studies have shown varying and conflicting data on which fluids provide adequate resuscitation without increasing mortality. The SSCG recommend crystalloids as the initial resuscitation fluid of choice, with addition of albumin only if substantial amounts of crystalloids are required (98). However, UK investigators obtained outstanding results with use of albumin boluses in resuscitation of children with meningococcal septic shock (99). In an RCT, Maitland et al. demonstrated a mortality benefit in malarial septic shock with albumin when compared to crystalloids (100). Nevertheless, evidence strongly recommends against using hydroxyethyl starch in adults and children due to an acute increase in RRT and red cell transfusion requirement (92,98).
In a recent large matched cohort study on pediatric sepsis, LR resuscitation was not associated with reduced mortality, AKI, or dialysis even when matched by fluid volume and proportionate LR utilization (101). Mortality increased with larger fluid volumes and decreased with a greater proportion of fluid given as LR, but there were no differences between LR and NS groups after matching within volume quartiles, by proportionate LR utilization, or in the separate matched analysis of LR-only patients. However, LR was preferentially used either as first-line fluid in patients with lower illness severity or as an adjunctive fluid in patients who received large amounts of fluid, and the matching algorithm was least effective in the most severely ill patients receiving the largest total fluid volumes. This  14 ± 8%, p = 0.02) methods was associated with composite poor outcome -FO <10% was associated with a good outcome by both FB and weight based methods (p = 0.02,0.01) but the association did not reach statistical significance after multivariate analysis    -Neonates without PDA who received prophylactic indomethacin had lower urinary output, lower weight loss (p = 0.012) and higher FiO 2 requirement (p < 0.0001) by the end of first week -Incidence of BPD was similar in PDA+ neonates irrespective of indomethacin prophylaxis but was significantly higher in PDA-infants who received indomethacin (p = 0.015) -Indomethacin prophylaxis reduces the incidence of PDA but not that of BPD

MANAGEMENT OF FLUID OVERLOAD
Early FO has garnered much attention as it seems to be a more decisive factor than late FO in predicting survival because the peak of illness is typically in first few days post-admission and the CFB curve generally flattens out after this period (58,59,71,75,76). Hence early fluid resuscitation is most likely beneficial, whereas early FO is not. Many studies advocate use of standard concentration drug infusion to prevent iatrogenic FO because many drugs are delivered as dilute solutions which significantly add to the daily fluid intake. This effect can be drastic in infants who are unnecessarily exposed to excess fluid through weight-based infusions (102). The concerns about perioperative hyponatremia in children with use of hypotonic maintenance fluids have been addressed lately via a switch to isotonic fluids (96,103,104). The lack of a specific recommendation on optimal timing of RRT institution still plagues clinicians. CRRT remains the modality of choice in pediatric critical care. The American College of Critical Care Medicine recommends that a threshold of 10% should be construed as the time for intervention (14). An expert discussion of the pharmacologic management of FO in critically ill patients was recently published (105). Furosemide effectively achieves diuresis in critically ill children, but RCTs have failed to demonstrate any benefit of diuretics in preventing or treating AKI (106,107). The furosemide stress test can be employed to assess the response to diuretics as resistance only delays RRT initiation (108). Aminophylline and fenoldopam have received medical attention owing to their potential in ameliorating AKI. The KDIGO guidelines recommend the use of a single dose of prophylactic theophylline in infants with severe perinatal asphyxia who are vulnerable to develop AKI (109). Use of these agents in post-cardiac surgery infants has also resulted in an improved excretory function and a reduction in urinary AKI biomarkers (110,111).

ONE SIZE DOES NOT FIT ALL -IS IT WORTHWHILE TO PREDICT FLUID RESPONSIVENESS?
The prediction of FR (fluid responsiveness) in critically ill children remains largely unembraced to date despite its potential role in obviating FO as only 40-60 % of patients typically respond to fluids by increasing the cardiac output by more than 10-15% (112). The hemodynamic predictor variables have been classified as static and dynamic. Static variables are volumetric and include preload measures based on a single observation in time, namely maximal end diastolic volume, CVP or pulmonary artery occlusion pressure. Dynamic variables consider the cyclical changes in preload occurring during ventilation which are apparent as changes in stroke volume via the Frank-Starling law. These include systolic pressure variation (SPV), pulse pressure variation (PPV), and stroke volume variation (SVV).
Respiratory variation in aortic blood flow peak velocity has been shown to be the only variable predictive of FR in children (112,113). A recent systematic review demonstrated the limited predictive ability of static and dynamic variables in pediatric patients (112). Static variables did not predict FR in children, similar to the evidence in adults. However, dynamic variables based on arterial blood pressure also did not predict FR unlike adults. This probably could be due to age-related differences in vascular mechanical properties which affect the arterial waveform behavior. Evidence for dynamic variables based on plethysmography was deemed inconclusive. The insensitivity of these variables can be ascribed to numerous reasons, namely low tidal volume ventilation, spontaneous respirations, arrhythmias and valvular regurgitation. Inclusion of baseline contractility may improve the predictive potential of these variables. Standardized volume loading (e.g., 10 ml/kg) may not lead to consistent changes in ventricular EDV due to inter-individual variance (114). There is a need to clarify how differences in lung, vascular, and cardiac compliances affect predictive potential of dynamic variables in children as volume loading should be tailored for only those patients who are likely to increase stroke volume.

LIMITATIONS
Literature on fluid resuscitation in pediatric patients is vast but there is no general consensus on the type and dose of resuscitation fluid. Though there is an association between fluid overload and the risk of BPD in a neonate, there is no consensus or guideline on whether fluid restriction or the management of FO will reduce BPD incidence. This is just one example showing the current lack of knowledge on fluid management in children. Further studies and well-designed RCTs are necessary to provide physicians with optimal guidelines. The limited predictability of fluid responsiveness in children along with limited knowledge on the utility of early goal directed therapy further complicates the management of FO. There is a need for more studies to better clarify the effects of FO and create a standardized set of guidelines to assist physicians with management and decrease adverse outcomes.

CONCLUSIONS
Managing fluid overload in a critically ill pediatric patient is no simple task and poses a multitude of complications. To begin with, the lack of a proper consensus on a method for calculating FO due to the significant number of uncertainties in estimating intake and output makes management a challenge. While FO itself is not a direct marker of mortality, the adverse effects of FO lead to patients being vulnerable to an increased risk of morbidity and mortality. FO puts patients at a risk of being underdiagnosed with AKI and delays treatment, raises odds for mortality associated with CCRT complications, can lead to increased hospital and ICU stay, and prolonged ventilator support in the critically ill population. Given the current lack of knowledge on fluid management in children, further studies and RCTs are necessary to provide guidelines to physicians.