Inferior Vena Cava/Abdominal Aorta Ratio as a Guide for Fluid Resuscitation.

INTRODUCTION: The fluid therapy is crucial in the treatment of critically ill children. Inadequate or excessive fluid resuscitation leads to increased mortality and morbidity, thus necessitating an accurate parameter for predicting fluid responsiveness when conducting fluid resuscitation. The inferior vena cava/abdominal aorta (IVC/Ao) ratio is suggested as a good guide for fluid resuscitation. However, the cutoff value for predicting fluid responsiveness in children has not been established. Is IVC/Ao ratio can be used to predict fluid responsiveness?
METHODS: The objective was to determine the accuracy and a cutoff value of IVC/Ao in predicting fluid responsiveness. A prospective cross-sectional study was conducted in the emergency room and the pediatric intensive care unit of the tertiary hospital from March to August 2017. We consecutively enrolled all critically ill children aged 1 month to 18 years' old who were hemodynamically unstable (shock). Measurements of IVC/Ao with ultrasound and stroke volume with ultrasound cardiac output monitor were obtained before and after fluid challenge.
RESULTS: Of 167 subjects enrolled in this study, only 58 subjects were included, most of whom were male (58.6%) and ranging in age from 1 to 11 months (32.8%). The mean IVC/Ao ratio before the fluid challenge in the fluid responsive group was 0.70 ± 0.053. The best cutoff of the IVC/Ao ratio is 0.675 with area under the curve 70.8% (95% confidence interval of 54.6%-87%), 75.7% sensitivity, and 61.9% specificity for predicting significant fluid responsiveness.
CONCLUSION: The measurement of IVC/Ao is an accurate, sensitive, and specific parameter to predict fluid responsiveness. The best cut-off for the IVC/Ao ratio is 0.675. Copyright:

INTRODUCTION

A circulatory failure, such as sepsis and trauma post major surgery, often occurs in critically ill children who are treated in intensive care or emergency rooms. Proper administration of intravenous fluids is an important part of treating circulatory failure in children and affects the clinical outcome of the management of the child's emergency. To this day, the fluid therapy strategy, including both the form and the amount, has remained controversial.[12]

Giving a small amount of fluid will cause tissue hypoperfusion and worsen organ dysfunction and cause ischemia. On the other hand, excess fluid interferes with oxygen delivery, exacerbates treatment outcomes, increases complications, and extends both the length of stay in intensive care and mortality.[23]

Monitoring hemodynamics to predict fluid responsiveness in pediatric emergencies is a challenge often faced by health workers in intensive care units or emergency departments (EDs), which is why it is so important to have accurate techniques to predict a patient's response to fluid administration.[23] There are two ways to predict fluid responsiveness to optimize the administration of resuscitation fluids: dynamic (i.e. systolic pressure variation, pulse pressure variation, stroke volume [SV] variation, pulse oximetry plethysmographic waveform amplitude, caval index, collapsibility superior vena cava index, distensibility internal jugular vein, subaortic velocity-time integral [VTI] variation, aortic velocity variation, carotid artery peak velocity variation, and inferior vena cava/abdominal aorta [IVC/Ao] ratio) and the other static (i.e. central venous pressure [CVP], pulmonary artery occlusion pressure, left ventricular end-diastolic area, left ventricular end-diastolic volumes, and inferior vena cava parameter).[456] Dynamic methods have been shown to be superior for predicting fluid responsiveness and is currently more widely used.[34]

All of these methods have a limitation because they are invasive, require high expertise, and cost a lot of money. In an attempt to find parameters that are dynamic, easy to do, noninvasive, and inexpensive, we identified the ratio of the inferior vena cava diameter and the abdominal aorta diameter, assessed using ultrasonography (USG), as a tool to determine fluid responsiveness, including caval index and inferior vena cava-abdominal aorta ratio. The caval index as a hemodynamic monitor has been widely used, but the use of the IVC/Ao ratio to predict fluid responsiveness has not been thoroughly studied. There is only one study in Poland in young adult patients that uses this parameter to assess hypovolemic shock, even though we have known that pediatric patients are different from adults both anatomically and physiologically.[7] Therefore, this study was conducted to validate the use of the IVC/Ao ratio and also to find out the optimal cutoff value, as a response parameter for fluid administration in critically ill children with impaired fluid homeostasis. We hypothesized that the IVC/Ao is good in predicting fluid responsiveness.

METHODS

Study design

This is a diagnosis test assessment, which is an observational, prospective, cross-sectional study, conducted at the pediatric intensive care unit (PICU) and ED of Cipto Mangunkusumo Hospital, a tertiary teaching university hospital, during a 6-month period of study from March to August 2017. This study was approved by the committee on human rights related to research involving human subjects, Faculty of Medicine University of Indonesia (approval number: 143/UN2.F1/ETIK/2017), and all the subjects' parents gave consent.

Study population

We consecutively enrolled all critically ill children aged 1 month to 18 years' old who were hemodynamically unstable (shock), following our two inclusion criteria, were admitted to the ED or PICU, and who needed a fluid challenge (according to the attending physician's assessment). The exclusion criteria were patients with congenital heart disease, heart failure, the disorder of venous return, and an increase of abdominal pressure. We needed a minimum of 54 subjects according to the statistical calculations.

Data collection and methods of measurement

All the critically ill children who came to the ED or PICU were assessed by the attending physician regarding their need for the fluid challenge. We recorded the identity, working diagnosis, vital signs (using monitor Philips Intellivue M60), and physical examination. After, we obtained consent from the child's parent.

Besides taking ultrasound measurements of the IVC and aorta using two-dimensional echocardiography, SV measurements using an ultrasound cardiac output monitor (USCOM) were taken simultaneously before and immediately after the fluid challenge was administered. USG was performed using Phillips HD 11XE, using the curve probe. The child was laid in the supine position, and the attending fellow (a pediatrician who wants to be a pediatric intensivist) of the PICU (who were on duty) did the ultrasound and USCOM.

Inferior vena cava/abdominal aorta ratio

The transducer was placed below the processus xyphoideus. The diameter in the IVC was measured during the expiratory phase of the respiratory cycle. The diameter of the Ao was obtained during the systole of the cardiac cycle.

Stroke volume measurement using the ultrasound cardiac output monitor

The transducer was placed on the suprasternal notch to find the aortic blood flow signal.

The USCOM probe was moved slowly around the suprasternal notch and stopped when we found a clear picture of the initial systolic to end-systolic, full systolic time, and sharp velocity-time curve (VTI) from the Doppler flow curve, which gives a clear sound. The curve will be presented on the monitor screen as a velocity–time curve. After we obtained the best curve image with the clearest sound, we pressed the freeze button to stop the measurement, which was then stored and SV was calculated automatically by the USCOM device.

Outcome measure

For each patient, we measured the ratio of the inferior vena cava diameter to the abdominal aorta diameter and SV two times, before and after fluid challenge (ringer lactate 10 ml/kg body weight over 20 min). After the fluid challenge was given, we categorized the outcome as fluid responsive if we found an increase of SV >10%. The number of the fluid challenge is only one time.

Statistical analysis

The research data obtained were processed using SPSS for Windows 20.0 software (IBM, Armonk, NY, USA). Data were presented in the form of narratives, tables, and images.

Numerical data were tested for normality by Kolmogorov–Smirnov where data with normal distribution were presented in mean and standard deviation, while those that were not normally distributed were presented in the median and range. The means that were normally distributed were analyzed using the t-test, and abnormal means were analyzed using the Wilcoxon test.

Univariate analysis was carried out for all variables, followed by multivariate analysis for meaningful variables.

Changes in the IVC/Ao ratio before and after the fluid challenge made a receiver operating characteristic (ROC) curve and was used to find the optimal cutoff value and calculated sensitivity and specificity.

RESULTS

During the study period, there were 3682 children who came to the ED and PICU of Cipto Mangunkusumo General Hospital. Out of 167 critically ill children, 47 children died ≤2 h of treatment and 58 children were subjects of the study. Of these subjects, most were male (58.6%), with the most represented age range between 1 and 11 months (32.8%). The most common nutritional status is malnutrition, with a total of 32 subjects (56.9%). Initial physical examination of the majority of subjects showed symptoms of increased heart rate (87.9%), tachycardia/nonpalpable pulse (91.4%), cold perfusion (87.9%) with prolongation of CRT >2 s (82.8%), and systolic blood pressure above the 5th percentile (86.2%) [Table 1].

Table 1

Characteristic subject

Characteristic	Sample, n (%)
Sex (male)	34 (58.6)
Age (months)
1-11	19 (32.8)
12-24	13 (22.4)
25-60	12 (20.7)
61-144	5 (8.6)
144	9 (15.5)
Nutritional status
Malnourished	32 (56.9)
Initial physical diagnostic
Heart rate (tachycardia)	51 (87.9)
Pulse (weak)	53 (91.4)
Perfusion (cold)	51 (87.9)
CRT >2”	48 (82.8)
Blood pressure systole >P5)	42 (72.4)
Urine production (<1 ml/kg/h)	50 (86.2)
Type of shock
Hypovolemia	43 (74.1)
Septic	15 (25.9)
Mechanical ventilation	30 (51.7)
Use of vasoactive agent
No	45 (77.6)

In this study, 37 subjects were fluid responsive and 21 subjects were nonfluid responsive after given a fluid challenge, based on whether the SV increased >10%.

There was a statistically significant difference in the parameters SV, cadiac index (CI), and IVC/Ao ratio before fluid administration between fluid responsive and nonfluid responsive subjects, with P = 0.004, 0.021, and 0.000, respectively [Table 2].

Table 2

Hemodynamic characteristics before fluid administration between fluid responsive and nonfluid responsive subjects

Parameter	Fluid responsive (n=37)	Nonfluid responsive (n=21)	P
Heart rate (beats/min), mean±SD	157.4±25.2	158.1±29.2	0.200
MAP (mmHg), mean±SD	61.9±18.47	65.7±13.07	0.200
SV (mL), median (range)	26 (10-38)	34 (8.9-73)	0.004*
CI (L/min/m2), median (range)	3.7 (2.1-7.9)	5.3 (1.6-13)	0.021*
IVC/Ao ratio	0.70 (0.053)	0.62 (0.112)	0.000***

*Mann-Whitney test, **T-test. MAP: Mean artery pressure, SV: Stroke volume, CI: Cardiac index, IVC/Ao: Inferior vena cava/abdominal aorta

The IVC/Ao ratio after the fluid challenge showed a significant increase (P = 0.010) in the fluid responsive subject group, with a median value of 0.17 (0.17–0.32). The mean IVC/Ao ratio before the fluid challenge in the fluid responsive group was significantly lower, with a value of 0.70 ± 0.053 compared to 0.86 ± 0.090 after administering the fluid challenge (P = 0.000).

There were 30 subjects (52.6%) using mechanical ventilation during the study. The most common mode used in this study is pressure control mode and synchronized intermittent mandatory ventilation. Tidal volume used is between 5 and 8 ml/kg body weight, whereas the highest peak end-expiratory pressure is 7 and the lowest is 5.

Of the 13 subjects (22.8%) who used inotropic (the inotropic had been administered prior to the study), the most used ones are dobutamine and dopamine.

For the majority of subjects (53, or 91.4%), the output was resolved within 6 h.

A low inferior vena cava diameter value before giving the fluid challenge showed an increase in SV of more than 10%.

By using the method, the area under the curve (AUC) value is 70.8% (95% CI of 54.6%– 87%), P = 0.009 [Figure 1]. The optimal intersection point on the curve shown in the coordinate ratio of the IVC/Ao ratio before giving the fluid challenge is ≥0.675 [Figure 2] and has a sensitivity and specificity of 75.7% and 61.9%, respectively [Table 3].

Figure 1

The receiver operating characteristic curve of the inferior vena cava/abdominal aorta ratio before the fluid challenge

Figure 2

Intersection curves of the inferior vena cava/abdominal aorta ratio before the fluid challenge

Table 3

Sensitivity and specificity of the inferior vena cava/abdominal aorta ratio before the fluid challenge

Variable	Value	Fluid responsiveness		Sensitivity (%)	Specificity (%)	PPV (%)	NPV (%)
		Yes	No
Ratio diameter IVC/Ao before fluid challenge	≤0.675	28	8	75.7	61.9	77.8	59.1
	>0.675	9	13

IVC/Ao: Inferior vena cava/abominal aorta, PPV: Positive predictive value, NPV: Negative predictive value

DISCUSSION

The most frequently used assessment of body fluid status is based on anamnesis and physical examination. Although it is very helpful, if only based on these two things, the results obtained are less accurate, this can be seen in our study result that showed almost all the subject have a sign of henodinamycally unstable but not all subjects are fluid responsive, while other additional examinations using hemodynamic monitoring devices require a longer time to get the result because they are invasive, expensive, require high expertise, and or not available in every place. Ultrasound is a noninvasive tool, cost-effectiveness, easy to use, portability, and is available in every places even in rural hospitals (in our country, ultrasound is available in almost every ED and PICU, even in primary health-care centers).[78]

The use of ultrasound as a tool to predict volume status has been widely confirmed by previous studies in both children and adults. Like a study conducted by Levine et al., 2010, in 52 children with diarrhea and/or vomiting who were diagnosed with clinical dehydration from WHO criteria in a hospital in Rwanda, a collapsibility IVC ultrasound and body weight measurements were performed before and after fluid administration where an increase in body weight of more than 10% after fluids is expressed as severe dehydration, IVC inspiratory collapse has a high sensitivity of 93%.[7]

The use of the IVC/Ao ratio as an ultrasound element is relatively new to assess volume status. Technically, the sonographic assessment of the IVC/Ao diameter ratio is an easy test to perform and can be performed effectively by doctors who are not experienced in the field of sonography. According to the research by Durajska et al., 2014, after 4 h of training, a person without experience doing ultrasound can perform an IVC and Ao diameter examination with a level of accuracy comparable to experienced examiners and an ultrasound examination of the IVC and Ao diameters is not a complex examination.[8] The diameters of the inferior vena cava and abdominal aorta differ in each child due to several factors such as age, sex, body weight, height, and body surface area.[910] The IVC diameter is greatly affected by changes in respiration (changes in intrathoracic and intra-abdominal pressure) and the state of fluid deficiency. The Ao diameter, however, is said to be quite stable in every child even in a state of dehydration because the Ao diameter has low compliance when compared with compliance from IVC, according to previous research. In a 2005 study, it was found that in patients with hypovolemia shock, the Ao diameter remained constant despite a large amount of blood loss.[11] This is the basis of this study: the diameter of the aorta serves as an internal control for each child.

The study showed that the mean IVC/Ao ratio before the fluid challenge in the fluid responsive group was significantly lower, with a value of 0.70 ± 0.053 compared to 0.86 ± 0.090 after administering the fluid challenge. The results of this study are consistent with a previous study in 2009 that showed that the average IVC/Ao ratio before fluid hydration is 0.75, which is significantly different than the measurement of 1.09 after hydration.[12]

The results of a 2011 study of 24 pediatric patients with dehydration supported previous studies where the mean IVC/Ao ratio before bolus was 0.78 ± 0.3 and increased significantly to 1.05 ± 0.45 after administration of a fluid bolus.[11]

The results of this study indicate that the IVC/Ao ratio – both the value before giving fluid and the percentage change in value – can be statistically significant and help to predict fluid responsiveness. Regarding the IVC/Ao ratio before giving the fluid challenge (P = 0.009), the AUC value was 70.8 where the optimal cutoff at 0.675 has sensitivity and specificity of 70.3% and 61.9%, respectively.

Research conducted in 2010 on 112 subjects aged <18 years in U. S. hospitals showed that the IVC/Ao ratio significantly predicted dehydration (P < 0.05) where the AUC was 73% with 86% sensitivity and 51% specificity at the optimal cutoff point of 0.88. In another study conducted in 2014, in 113 patients with suspected complaints of diarrhea or dehydrated vomiting in Hasbro Children's Hospital on Rhode Island, the results identified an AUC value for the IVC/Ao ratio of 72% that significantly predicted dehydration, with a sensitivity of 67% and specificity of 71% at the optimal cutoff point of 0.8.[13]

Contrary to the results of this study, a study conducted in 2009 that examined 36 dehydrated subjects aged 6 months–16 years showed an AUC value of 91%, with an optimal cutoff point of 0.72 and very low sensitivity and very high specificity.[12] In addition, contrasting with this study is a study conducted in 2013 that examined 51 pediatric patients aged <21 years who entered the PICU requiring CVP for hemodynamic monitoring and who had a simultaneous measurement of the IVC and aorta using USG. Where dehydration in the subject was established by CVP ≤8 mmHg, caval index >50%, or IVC/Ao ratio ≤0.8, a statistical test of these values yielded very low sensitivity (17.6%) and high specificity (80.8%).[1415]

There may be differences in the current study's results compared to previous research because the present study did not use a gold standard; it only used children who were considered normal.[12] The other reason why previous studies do not match ours maybe that CVP was used for comparison, whereas most subjects (67%) used ventilators and patients following abdominal or thoracic surgery.[14] Positive pressure ventilation can change the pressure gradient in the chest and abdominal cavity, thereby affecting the ability of the IVC to shrink (reduce collapsibility) because the IVC is very easily influenced by outside pressure.[15161718] In addition to this study, 65% of CVP subjects were placed in the femoral vein,[14] which caused research results to be biased due to more external influences on the femoral vein compared to when placed in the subclavian or internal jugular veins. Subjects in this study also used sedation and vasopressors, which could affect the measurement of IVC and CVP diameters.[151617] That's why we have to be careful when using the IVC/Ao ratio as a hemodynamic parameter in a patient with high abdominal pressure.

Study limitation

(1) Subjects with heterogeneous diagnosis and (2) we did not do a kappa test for each ultrasound operator, but all the operators had undergone certified ultrasound examination training.

CONCLUSION

Based on this study, the optimal cutoff value for the IVC/Ao ratio before the fluid challenge is 0.675. This demonstrates good accuracy. The sensitivity and specificity of using the IVC/Ao ratio before the fluid challenge are also good.

Research quality and ethics statement

This study was approved by the Institutional Review Board / Ethics Committee, University of Indonesia approval no #143/UN2.F1/ETIK/2017. The authors followed applicable EQUATOR Network (http://www.equator-network.org/) guidelines during the conduct of this research project.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.