Determining pediatric fluid responsiveness by stroke volume variation analysis using ICON® electrical cardiometry and ultrasonic cardiac output monitor: A cross-sectional study.

PURPOSE: The purpose is to determine the adequacy fluid responsiveness by the validity and cut off point of stroke volume variation (SVV) usingelectrical cardiometry, ICON® (Osypka Medical, Berlin, Germany) and ultrasonic cardiac output monitor (USCOM) and to recognize cut off point of tidal volume in shock children with mechanical ventilation.
MATERIALS AND METHODS: A cross-sectional study was conducted from March 2017 to September 2017 in a single center. The selection of subject through consecutive sampling. Measurements of SVV and stroke volume (SV) using USCOM and ICON were performed before and after fluid challenge. The tidal volume of individuals was measured and recorded.
RESULTS: Analysis was performed in 45 patients with median age of 14 months and 62.2% of male population. It showed that the sensitivity and specificity of ICON were 58% and 74%, respectively. The optimal cut off point of SVV using ICON was 16.5% and the area under the curve (AUC) value was 53% (95% confidence interval [CI] 35.9%-70%), P > 0.05 and cut off point of SVV using USCOM was 33.5% with the AUC value was 70% (95% CI 52.9%-87.7%), P < 0.05. The optimal cut off point of tidal volume to fluid responsivenes was 6.8 ml/kg BW and the AUC value was 44.6% (95% CI 27.4%-61.9%), P > 0.05.
CONCLUSION: This study showed that electrical cardiometry (ICON) is unable to assess preload and the response of fluid resuscitation in children. Copyright:

INTRODUCTION

The main problem in critical patients who need fluid resuscitation is the assessment of fluid adequacy (preload) and response to fluid administration (fluid responsiveness).[1] Proper fluid administration is one of the most important indicators in the management of pediatric emergency. Excess or lack of fluid administration can increase a morbidity and mortality.[2] Therefore, accurate monitoring is needed so that morbidity and mortality due to incorrect amount of fluids can be reduced.[345]

Two methods commonly used to assess fluid adequacy and response to fluid administration in patients with unstable hemodynamics are static and dynamic methods. Dynamic methods are more accurate than static methods and currently more widely used in intensive room because they can describe the true hemodynamic state that always changes every time.[23456] Static methods include measurement of central venous pressure, pulmonary artery occlusion pressure, and diameter of inferior vena cava diameter. While dynamic methods include the measurement of stroke volume variation (SVV),[7] pulse pressure variation, index of inferior vena cava collapse, and systolic blood pressure variation.

The assessment of fluid adequacy and response to fluid administration by dynamic methods is done by measuring SVV.[8] SVV is a diagnostic value in predicting fluid responsiveness in various settings.[7] One of the methods for assessing SVV quickly, easily, and noninvasively is using ultrasonic cardiac output monitor (USCOM). USCOM has a good validity and has been tested when compared to Fick and thermodilution invasive technique. Another noninvasive hemodynamic monitoring tool is with electrical cardiometry (ICON). In some studies, ICON is as good as echocardiography and thermodilution methods. Examination using ICON is relatively easy, so it can be done by both doctors and nurses. This is a distinct advantage to be able to immediately assess hemodynamic disorders in pediatric patients.[291011121314151617] SVV by cardiac velocimetry succeeded to predict fluid responsiveness with cut off value of 11.5%, sensitivity 100%, specificity 70.0%, and P < 0.001.[18]

SVV has a diagnostic value to predict response to fluid administration in children with mechanical ventilation.[19] Whereas in patients, who do not use a mechanical ventilation, SVV has a low reliability so it cannot be relied on.[20] This study aimed to determine the validity of SVV as measured by ICON and USCOM in assessing response to fluid administration. We also determined the stroke volume (SV) sensitivity and specificity as measured by ICON with an increase of SV ≥ 10% after fluid challenge as measured by USCOM.[21] This study also aimed to find out the area under the curve (AUC) and the optimal cut off point of SVV before fluid challenge as measured by ICON and USCOM to determine the response to fluid administration. Another aim of this study was to determine the AUC and the optimal cut off point of the tidal volume of patients who use mechanical ventilation to respond to fluid administration.

MATERIALS AND METHODS

This study was approved by the committee on human right related to research involving human subjects, Faculty of Medicine University of Indonesia, Dr. Cipto Mangunkusumo Hospital and all participants parent or guardian provided informed written consent (approval number: 109/UN2.F1/ETIK/2017). It was a cross-sectional study comparing USCOM and ICON for the measurement of fluid responsiveness based on SVV analysis on increased SV in critically ill children. The study was performed at a pediatric intensive care unit and emergency room in a tertiary care academic center. It is a national referral center for pediatric subspecialty care in Indonesia.

Critically ill children, aged 1 month–18 years, who were admitted with shock condition which required fluid resuscitation and mechanical ventilation from March 2017 to September 2017, were screened for enrollment.[22] Decision of admission was made by proxy of severity illness and based on the decision of attending pediatric intensivist. The exclusion criteria were children with congenital heart diseases, congestive heart failure, renal failure, and arrhythmia or an abnormal heart rhythm. Sampling is done by consecutive sampling until the specified time limit. Approximately 45 participants would suffice with acceptable precision. The independent variable was SVV and dependent variable was the change in SV. The tools used in this study were a vital sign monitor (Philips Intellivue M60), USCOM (USCOM 1A, USCOM Pty Ltd, Coffs Harbor, NSW, Australia), and ICON (ICON Electrical Cardiometry, Cardiotronic USA).

Individuals who met the inclusion criteria were carried out at a physical examination (weight, height, and body surface area) and measurement of SV and SVV using USCOM and ICON. Then, the fluid challenge was given using crystalloid liquid or colloidal liquid in the amount of 10 ml/kgBW for 15–30 min. After the fluid challenge was given, the measurement of SV and SV variation were taken again. The measurement using USCOM and ICON was carried out simultaneously. USCOM and ICON measurement were obtained by five pediatric intensivists who were blinded from the simultaneous data using two data recorders (interobserver reliability testing).

In this study, all individuals were divided into two groups based on the response to fluid administration, i.e., the group did not respond to fluid challenge (there was an increased in SV < 10%) and the group who responded to fluid challenge (there was an increased in SV >10%) measured using USCOM. If the SV increased by >10%, the patient was considered to be volume responsive and fluid challenges were repeated until no further increase in SV was noted. At this point, the patient was considered to be unresponsive to fluid. No further fluid challenges were given unless the SV decreased by >10%.[21] The determination of ICON sensitivity and specificity was an increased SV > 10% as measured by USCOM using a 2 × 2 table analysis. In addition, an AUC value and a cut off point for optimal SVV (before fluid challenge) were determined as measured by ICON and USCOM. AUC and the optimal cut off point of the tidal volume of patients who using mechanical ventilation on the response to fluid administration were also measured.

The collected data were processed and analyzed using SPSS for Windows 20.0 software (IBM, Armonk, NY, USA). Data were analyzed descriptively, calculating the frequency distribution and proportion to determine the characteristics of the research subjects.[23] If the distribution is normal thus a paired t-test can be carried out, but when the distribution is not normal, then a McNemar's test was conducted to differentiate clinical hemodynamic parameters before and after fluid challenge. The Wilcoxon test was also carried out to compare the hemodynamic parameters of SV and SVV before and after fluid challenge.

RESULTS

The analysis was performed in 45 patients with median age of 14 months (1 month–204 months) and 62.2% of male population. The median body weight is 9 kg (3.3 kg–45 kg) and the median body surface area is 0.4 m2(0.2 m2–1.32 m2). All the patients were mechanically ventilated with the most mode of mechanical ventilation was the pressure-control mandatory ventilation (66.7%) with median tidal volume was 7 ml/kg BW. The most common indications for fluid challenge were due to hypovolemic shock (60%). The most diseases which required fluid challenge were postsurgery, respiratory problems, and diarrhea [Table 1].

Table 1

Characteristic of study subjects

Characteristics	Value
Age (months), median (minimum-maximum)	14 (1-204)
Gender
Male	28 (62.2)
Female	17 (37.8)
Body weight (kg), median (minimum-maximum)	9 (3.3-45)
Body surface area (m2), median (minimum-maximum)	0.4 (0.2-1.32)
Tidal volume (ml/kg BW), median (minimum-maximum)	7 (5-10)
Mode of mechanical ventilation
P-CMV	30 (66.7)
P-SIMV	13 (28.9)
PS-CPAP	2 (4.4)
Mode of mechanical ventilation P-CMV
Respond to fluid resuscitation	20 (44.4)
Not respond to fluid resuscitation	10 (22.3)
Mode of mechanical ventilation P-SIMV
Respond to fluid resuscitation	4 (8.9)
Not respond to fluid resuscitation	9 (20)
Mode of mechanical ventilation P-SIMV
Respond to fluid resuscitation	2 (4.4)
Not respond to fluid resuscitation	0
Indication for fluid challenge
Hypovolemic shock	27 (60)
Septic shock	18 (40)
Diseases and disorders
Tetanus	6 (13.3)
Respiratory disorder	8 (17.8)
Diarrhea	7(15.6)
Encephalitis/neurology disorder	1 (2.2)
Sepsis	4 (8.9)
Hepatitis fulminant	2 (4.4)
Postsurgery	17 (37.8)

P-CMV: Pressure-control mandatory ventilation, P-SIMV: Pressure-synchronized intermittent mandatory ventilation, PS-CPAP: Pressure support-continuous positive airway pressure

Most individuals experienced significant clinical improvement, i.e., decreased the number of individuals who experienced tachycardia, improved pulse strength, decreased the number of individuals with cold acral palpation in extremities, improved in CRT (Capillary Refill Time) and diuresis after fluid challenge. Only hemodynamic parameters of systolic blood pressure > P5 did not have significant differences after fluid challenge [Table 2].

Table 2

Clinical profile before and after fluid challenge

Clinical parameter of shock	Fluid challenge
	Before	After	P
Tachycardia	40 (88)	32 (71)	0.008
Pulse strength (weak pulse/not palpable)	40 (88)	13 (28)	<0.0001
Extremity (cold acral)	37 (82)	15 (33)	<0.0001
CRT >2 s	36 (80)	15 (33)	<0.0001
Systolic blood pressure >P5	33 (73)	39 (86)	0.146
Diuresis <1 ml/kgBW/h	21 (46)	8 (17)	<0.0001

CRT: Capillary refill time

Hemodynamic parameters measured using the electrical cardiometry (ICON) method and those measured using the Doppler method (USCOM) showed a significant increase in SV after fluid challenge. The hemodynamic parameters of SVV measured using ICON method showed a significant decreased after fluid challenge, with a median of SVV was 16% to 13%. Whereas SVV parameters measured using the Doppler method (USCOM) did not show a significant decrease after fluid challenge with a median of SVV was 37% (before fluid challenge) decreased to 32% (after fluid challenge). The purpose of this study was to determine the ICON validity in assessing the response of fluid administration with gold standard SV increased ≥ 10% as measured by USCOM. To determine ICON validity, an analysis was done using 2 × 2 tables. The results of the analysis showed that ICON sensitivity was 58%, while the specificity was 74% [Table 3].

Table 3

Response of fluid administration (increased of stroke volume ≥10%) measured using ICON and ultrasonic cardiac output monitor

Response of fluid administration (increased of SV >10%) measured using ICON	Response of fluid administration (increased of SV >10%) measured using USCOM		Total
	Respond to fluid administration	Not respond to fluid administration
Respond to fluid administration	15	5	20
Not respond to fluid administration	11	14	25
Total	26	19	45

USCOM: Ultrasonic cardiac output monitor, SV: Stroke volume

In this study, determining the optimal cut off point of SVV before fluid challenge as measured by ICON and USCOM will be carried out statistically by looking for the cut off point of sensitivity and specificity through AUC value and determining the coordinate from receiver operator curve (ROC). The AUC value (determined by ROC method) of SVV before fluid challenge as measured by ICON was 53% (95% IK 35.9%–70%), P > 0.05 [Figure 1]. This value meant that the SVV value measured by ICON was unable to distinguish groups of patients who respond to fluid resuscitation and who did not respond to fluid resuscitation. The optimal cut off point of SVV before fluid challenge as measured by ICON toward increased of SV ≥10% as measured by USCOM was 16.5%. The SVV before fluid challenge >16.5% as measured by ICON had a sensitivity of 54% and specificity of 58%.

Figure 1

Receiver operator curve of stroke volume variation before fluid challenge as measured by ICON

The AUC value (determined by ROC method) of SVV before fluid challenge as measured by USCOM was 70% (95% IK 52.9%–87.7%), P < 0.05 [Figure 2]. The optimal cut off point of SVV before fluid challenge toward increased of SV ≥10% after fluid challenge as measured by USCOM was 33.5% and had a sensitivity of 73% and specificity of 68%.

Figure 2

Receiver operator curve of stroke volume variation before fluid challenge as measured by ultrasonic cardiac output monitor

The AUC value of tidal volume in determining the response of fluid administration (increased of SV ≥10%) as measured by USCOM was determined by the ROC curve was 44.6% (95% CI 27.4%–61.9%), P > 0.05 [Figure 3]. Tidal volume value when used to determine the response to fluid administration in 100 patients the correct conclusions were obtained in 44 patients. The optimal cut off point of tidal volume was ≥6.8 ml/kg BW and had a sensitivity of 46% and specificity of 42% in determining the response to fluid administration based on the SVV analysis.

Figure 3

Receiver operator curve of tidal volume (response to fluid administration) as measured by ultrasonic cardiac output monitor

In this study, interobserver reliability testing was conducted for USCOM SV examination and statistically analyzed using the Bland–Altman test. From the results of Bland–Altman Analysis, there were no differences in USCOM SV examination between examiners.

DISCUSSION

Fluid resuscitation in patients with inadequate tissue perfusion can prevent tissue hypoxia and more severe organ damage while early resuscitation can also improve patient outcomes.[24] However, in several clinical studies, only 45%–50% of patients with unstable hemodynamics who responded to fluid resuscitation.[225] The study showed that the most majority of pediatric patients who needed fluid resuscitation was male (62.2%). This is consistent with study conducted by Labib et al. who examined the relationship of passive leg raising and fluid challenge in pediatric patients after surgery and the majority of the study population was male (62.5%) using electrical cardiometry method.[26]

A patient is said to respond to fluid resuscitation, if there is an increase of SV ≥ 10% after the fluid challenge. Conversely, a failure to increase SV after the fluid challenge indicates that the circulatory system has not responded to fluid administration.[2127] This is consistent with our study where there was a change of SV in shock patients after fluid challenge. SV parameters were measured by the method of electrical cardiometry (ICON) and Doppler method (USCOM) demonstrated a significant increase of SV after fluid challenge. A patient who is experiencing a shock is also said to have a response to fluid administration if there is a decrease in SVV parameters after the fluid challenge.[28] The study also found a significant reduction in SVV after the fluid challenge as measured by ICON.

The results of this study showed that sensitivity of ICON was 58% while specificity of ICON was 74%. Low sensitivity value indicates that the response of fluid administration as measured by ICON has a poor diagnostic value. There are differences in working principles between ICON and USCOM. The calculation of contractility in ICON is strongly influenced by the basic impedance (Z0). The basic impedance is largely determined by thoracic fluid, including the volume of blood in the thorax. This will affect the calculation of contractility in patients who have experienced pulmonary edema or pleural effusion.[29] Many factors influence the calculation of SV in ICON. For example, to calculate patient's contents, it is necessary to calculate the patient's body mass which strongly influenced by the nutritional status of the patient. The calculation of corrected flow time is influenced by the calculation of R-R interval on the ECG, such as patients who have arrhythmias or myocarditis.[29] This calculation is very different from the calculation of SV on USCOM.

On USCOM, the principle of SV measurement is based on the principle of Doppler. USCOM is a direct derivative of echocardiography that uses continuous wave of Doppler ultrasonography. The working principle of USCOM is to measure velocity time integral (VTI) from cardiac ejection flow and heart rate. In USCOM, the accuracy of weight and height data from patients is needed for SV measurement. In addition, in order to obtain optimal VTI results, it is also necessary to have the proper placement skills of USCOM probes, so they will get the optimal VTI.[30] In our study, the AUC value of SVV as measured by USCOM was 70% with an optimal cut off point of SVV before fluid challenge was 33,5% (a sensitivity of 73% and specificity of 68%). This result is similar with a previous study which is done in the same hospital before.[31]

In our study, the AUC value of SVV before fluid challenge as measured by ICON was 53%. This value meant that the SVV value measured by ICON was unable to distinguish groups of patients who respond to fluid resuscitation and who did not respond to fluid resuscitation. Our results are consistent with the study of Vergnaud et al. who obtained AUC value of SVV 57% for patients younger than 3 years with a sensitivity of 91% and specificity of 40%. This condition is caused because myocardium in children aged >3 years responds less to adequate fluid and is more susceptible to fluid overload. In addition, cardiovascular physiology experienced a drastic change in the 1st year of life, but thereafter did not change much.[32] The study samples were not different from the study conducted by Vergnaud et al. where the median age in our study was 14 months.[32] On the other hand, factors that causes differences in diagnostic accuracy values in our study and another study are physiological differences between children and adults, especially in the elasticity of blood vessels, compliance of the chest wall and heart and the difference in aortic impedance that is very influential on SVV.[17]

SVV had a diagnostic value in predicting the response to fluid administration in children if the patients used mechanical ventilation.[19] Whereas in patients who did not use mechanical ventilation, SVV parameters were not reliable.[20] Suehiro and Okutani revealed that SVV could be used as a predictor of response to adequate fluid administration in patients who use mechanical ventilation with a minimum tidal volume of 8 ml/kg BW, whereas SVV did not have the predictor value of response to adequate fluid administration in patients who use mechanical ventilation with a tidal volume of 6 ml/kg BW.[33] This is consistent with the results of our study, the tidal volume value in determining the response to fluid administration using SVV parameters having AUC determined by the ROC method was 44.6% with the optimal cut off point of the tidal volume was 6.8 ml/kg BW (a sensitivity of 46% and specificity of 42%). This value means that the tidal volume does not have a predictor value in distinguishing patients who respond to fluid resuscitation and who did not respond to fluid administration.

Our research has several limitations including the heterogeneous subjects with various characteristics and diseases. There are several factors that influence hemodynamic examination using ICON, including patient's muscle mass, nutrition, heart rhythm, and the presence of myocarditis. This can affect the outcome of ICON examination results, for example patients with the same age but with different nutritional status. This is the first study in Indonesia which is examined the validity of ICON in determining the response of fluid resuscitation in pediatric patients whereas ICON has been widely used in many hospitals in Indonesia as a tool to assess the hemodynamics of pediatric patients.

CONCLUSION

This study showed lower sensitivity and specificity of ICON compared to USCOM to determine the validity of SVV calculation to assess the preload condition. The main explanation is several factors can influence the impedance of electrical impulse which is crucial for the SVV calculation. ICON also has a lower AUC value compared to USCOM to calculate the SVV cut-off point before the fluid challenge performed. SVV cut-off point is important to assess fluid resuscitation response after fluid challenge. Thus, it is concluded that ICON is unable to assess preload and fluid resuscitation responses in children accurately. These result should be a consideration of many hospitals which use ICON to assess the hemodynamics of pediatric patients.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

Research quality and ethics statement

This study was approved by the Institutional Review Board / Ethics Committee. The authors followed applicable EQUATOR Network (http://www.equator-network.org/) guidelines during the conduct of this research project.