Evaluation of Fluid Overload by Bioelectrical Impedance Vectorial Analysis

Summary

In this study, we demonstrate how to evaluate the presence of fluid overload through bioelectrical impedance vectorial analysis (BIVA) and the impedance ratio measured using tetrapolar multi-frequency equipment in patients admitted to the emergency department. BIVA and impedance ratio are reliable and useful tools to predict poor outcomes.

Abstract

Early detection and management of fluid overload are critically important in acute illness, as the impact of therapeutic intervention can result in decreased or increased mortality rates. Accurate fluid status assessment entails appropriate therapy. Unfortunately, as the gold standard method of radioisotopic fluid measurement is costly, time-consuming, and lacks sensitivity in the acute care clinical setting, other less-accurate methods are typically used, such as clinical examination or 24 h output. Bioelectrical impedance vectorial analysis (BIVA) is an alternative impedance-based approach, where the raw parameter resistance and reactance of a subject are plotted to produce a vector, the position of which can be evaluated relative to tolerance intervals in an R-Xc graph. The fluid status is then interpreted as normal or abnormal, based on the distance from the mean vector derived from a healthy reference population. The objective of the present study is to demonstrate how to evaluate the presence of fluid overload through bioelectrical impedance vectorial analysis and the impedance ratio measured with tetrapolar multi-frequency equipment in patients admitted to the emergency department.

Introduction

Fluid overload (FO), defined as an excess of total body fluid or a relative excess in one or more fluid compartments¹, is frequently observed in critically ill patients and is associated with higher morbidity and mortality^1,2,3. The range of alterations in hydration status is wide; can indicate renal, cardiac, or hepatic failure; and/or maybe the result of excessive oral intake or iatrogenic error⁴. Routine assessment of hydration status is challenging in emergency departments, as the gold standard of radioisotopic volume measurement requires specialized techniques, is costly and time-consuming, and may fail to identify early disturbances in hydration status. Hence, other less-accurate methods are generally used, including clinical examination and accumulated fluid balance (volume in mL in 24 h)⁵. Accurate and sensitive determination of fluid volume status is necessary to help clinicians in controlling body fluids, managing intravenous fluid administration, and maintaining hemodynamic stability, thus allowing patients to receive early treatment^3,5,6. Errors in the volume assessment can lead to a lack of necessary treatment or to implementation of unnecessary therapy, such as excess fluid administration, both of which are related to increased hospitalization costs, complications, and mortality⁴.

Interest has recently increased in bioelectrical impedance analysis (BIA), which has been considered an alternative method for the classification of an individual's hydration status. BIA is a safe, non-invasive, portable, quick, bed-side, and easy-to-use method, designed for the estimation of body compartment composition. The analysis measures the opposition generated by soft tissues to the flow of an injected alternating electric current into the body (800 µA), through four surface electrodes placed on the hands and feet. Total body water estimated by BIA has been shown to have a high correlation with that obtained by deuterium dilution (r = 0.93, p = 0.01)⁷.

Phase-sensitive BIA devices evaluate the direct measurement of phase angle and impedance (Z₅₀), obtaining the resistance (R) and reactance (Xc) in single-frequency mode (50 kHz) or multi-frequency mode (5 kHz to 200 kHz)⁸. Dividing the R and Xc values by the subject's height (in m) squared-to control for inter-individual differences in conductor length-and plotting them in an R-Xc graph is the method used in bioelectrical impedance vector analysis (BIVA) to estimate the fluid status. BIVA is an alternative impedance approach, developed by Piccoli et al.⁹, which uses the spatial relationship between R (i.e., the opposition to the flow of an alternating current through intra- and extra-cellular ionic solutions) and Xc to assess soft tissue hydration, independent of the multiple-regression prediction equations generated in limited and specific samples¹⁰. Therefore, the classification of fluid status is more precise and accurate than the quantification of total body water. The R and Xc values of a subject produce a vector whose position can be evaluated relative to tolerance intervals in the R-Xc graph, which can be interpreted as indicating normal or abnormal hydration, based on the distance from the mean vector derived from a healthy reference population^11,12,13.

In a previous study, we compared different bioelectrical impedance analysis parameters for the detection of fluid overload and prediction of mortality in patients admitted to an emergency department (ED) and demonstrated that BIVA (relative risk = 6.4; 95% confidence interval from 1.5 to 27.9; p = 0.01) and impedance ratio (relative risk = 2.7; 95% confidence interval from 1.1 to 7.1; p = 0.04) improved the estimation of the probability of 30-day mortality³.

Fluid overload can also be estimated using the impedance ratio (imp-R), which is the ratio between impedance measured at 200 kHz and impedance measured at 5 kHz obtained by the multi-frequency bioelectrical impedance equipment. Imp-R considers conduction in total body water (Z₂₀₀) and in extracellular water fluid spaces (Z₅). The penetration of a current into cells is frequency-dependent and, the 200/5 kHz ratio describes the ratio of greater to lesser current entry into cells^3,8. If the difference between these two values decreases over time, it may indicate that the cells are becoming less healthy¹⁴.

Imp-R values ≤0.78 in males and ≤0.82 in females have been observed in healthy individuals¹⁵. Values nearer to 1.0 indicate that the two impedances are closer to each other, and the body cell is less healthy. In the case of critical illness, the resistance of the cell membrane at 5 kHz is reduced, and the difference between the impedance values at 5 and 200 kHz is markedly lower, indicating cellular worsening³. Values > 1.0 suggest device error^16,17. Thus, the objective of the present study is to demonstrate how to evaluate the presence of fluid overload through bioelectrical impedance vectorial analysis, as well as by using the impedance ratio, measured with tetrapolar multi-frequency equipment in patients admitted to the emergency department.

Protocol

The following protocol was approved (REF. 3057) and follows the guidelines of the human research ethics committee of Instituto Nacional de Ciencias Médicas y Nutrición SZ. Furthermore, prior consent was obtained from the patients for this study.

NOTE: This procedure is to be used for measuring bioelectrical impedance analysis using tetrapolar multi-frequency equipment (see Table of Materials) and will provide accurate resistance and reactance values at a single frequency of 50 kHz, as well as the ratio between 200 kHz and 5 kHz impedance values (200/5 kHz).

1. Before testing

1. Perform standardization of the person who will perform the measurements, as someone with a qualification in the area or who has extensive experience in making measurements.
2. Ask the patient to refrain from eating for 4 to 5 h before testing.
3. Periodically test the equipment to verify the impedance measurement are as accurate as possible, according to the guidelines provided by the manufacturer, using a test resistor with a known value of 500 Ω (range 496-503 Ω). Make sure the adhesive electrodes correspond to the manufacturer's recommendation.
4. Clean the equipment using a chlorhexidine wipe, and then wash your hands. If the screen of the equipment shows the legend: change the battery, then replace the battery.
5. If the patient is conscious, explain the procedure to them. Obtain the age and accurate measurement of the patient's height (in cm) and introduce these data into the equipment.
6. Remove the shoe and sock from the right foot, as well as any metal objects, such as watches or bracelets worn by the patient. Place the patient in a supine position for 5 min with legs and arms spread out around 45° before taking the measurements, verifying that they are not in contact with any other part of their body. In patients with obesity, in order to avoid contact between the thighs, place a sheet between their legs.

2. Measurement of BIA parameters

1. Clean the surfaces where the electrodes will be placed with a 70% alcohol pad twice. Place two electrodes on the right hand dorsally, one behind the knuckle of the third metacarpophalangeal (middle finger) and the other on the wrist, next to the ulna head carpal joint. It might be helpful to draw an imaginary straight line between the protruding bones on the wrist, and then place each electrode in the center of that line.
2. Place two electrodes on the right foot, one behind the third metatarsophalangeal joint and the tarsal joint on the ankle between the medial and lateral malleoli. To place the electrodes, follow the bones underneath. Ensure that the distance between the electrodes on the foot and hand is at least 5 to 10 cm, according to the size of the hand.
3. Connect the lead wires to the equipment, with the red alligator clip nearest to the nails and the black clip nearest to the ankle or wrist; ensure that the wires do not cross between them.
4. Ensure that the patient is not talking or moving during the measurements, as this will affect the results.
5. The patient's ID will come up on the first screen. Scroll through and change the patient's parameters (sex, age, height, and weight). Ensure that the electrodes are stuck down correctly, and press Enter. It will show: measuring, on the screen. It takes about 6 to 10 s to measure, and a beep will sound when the measurement is complete.

3. Analysis of bioimpedance parameters

1. The equipment will display the raw impedance values (Z) at four different frequencies: 5, 50, 100, and 200 kHz, as well as the resistance and reactance at 50 kHz, which are the values needed to classify a patient with fluid overload.
2. Download the software named BIVA tolerance R-Xc graph¹³ (see Table of Materials) and open it.
3. Observe that the software is in a workbook in a spreadsheet program with seven worksheets: guide, reference populations, point graph, path, subjects, Z-score, Z-graph.
4. Right click on the Reference Population sheet, choose the line of the selected reference population, and copy and paste it into the second row (yellow row).
5. Right click on the Subjects sheet and, in the second row, insert the following data: the subject ID assigned to the patient. In the second column named Seq, always put the number 1; and optionally fill out the surname and name columns. In the sex column, enter F for female patient and M for male patient. In the next two columns, input resistance and reactance at 50 kHz each. Insert the height (in cm) and weight (in kg) in the next two columns.
6. In the Popul Code column, insert the number that appears in the first column of the reference population sheet. In group code, randomly chose a number between 1 and 10 (this number will be required in the Point graph sheet), insert the patient's age in the next column.
7. In the spreadsheet program menu, go to the Complements tab and right-click on the Calculate option to obtain the values of resistance and reactance adjusted by height and phase angle.
8. Right-click on the Point Graph sheet, and observe that 50%, 75%, and 95% tolerance ellipses are drawn for the selected reference population (i.e., the population in the first yellow row at the top of the reference population sheet).
9. In the dialog box, select groups, right-click on the number placed in the group code located in the subjects' sheet, and right-click on OK. Next, the BIVA graph will appear, with the subject vector as a geometric figure (Δ, •, □).
10. Patients with vectors that fall outside the lower pole of the 75% tolerance ellipse will be classified as fluid overload (see Figure 1).
11. Divide Z at 200 kHz by Z at 5 kHz-which reflects the total body water and extracellular water compartment, respectively-in order to obtain the impedance ratio (Imp-R). A value ≥0.85 indicates fluid overload.
    NOTE: In new tetrapolar multi-frequency devices, the R-Xc graph is already included; however, it is important to ensure that the reference population is correct.

Representative Results

As an example of the method presented above, we present the results for two women admitted to the emergency department. Bioelectrical impedance analysis was assessed at admission using a phase-sensitive multi-frequency device (see Table of Materials), and the obtained resistance (R) and reactance (Xc) values were used to calculate the BIVA graph. The results show that patients with overhydration had worse prognoses and clinical characteristics such as SOFA and Charlson index scores, which are related to fluid overload.

In Figure 2, the results plotted with Δ denote a 77-year-old female (height = 155 cm) with normal fluid status and the following bioimpedance results: R = 586.7, Xc = 62.1. The data of clinical variables were as follows: sequential organ failure assessment score (SOFA) = 3; Charlson comorbidity index score = 5; the primary cause of hospital admission = hypotonic hyponatremia secondary to diuretic use and diarrhea; and length of hospital stay = 2 days.

Meanwhile, the results plotted with □ denote a 62-year-old female (height = 149 cm) with fluid overload and bioimpedance results R = 332.6, Xc = 33.6. Data of clinical variables were as follows: SOFA = 16; Charlson comorbidity index = 4; primary cause of hospital admission = septic shock secondary to soft tissue infection; length of hospital stay = 3 days. This patient died due to the progression of refractory shock, with acute respiratory distress syndrome that progressively worsened.

Figure 1: RXc-graph of the bioelectrical impedance vector analysis to classify the fluid status of a patient. Individual vectors below 75% (+2 standard deviation) can be classified as fluid overload Δ. Abbreviations: R = resistance, Xc = reactance, H = height. Please click here to view a larger version of this figure.

Figure 2: RXc-graph with the data of two female patients admitted to the emergency department. Δ is patient in the 50% tolerance ellipse showing a normal fluid status. □ is patient below the 75% ellipse classified with fluid overload. Please click here to view a larger version of this figure.

Discussion

It is important to mention that different bioelectrical impedance analysis (BIA) approaches have been proposed in the published literature, including the use of multiple frequencies at 1-500 kHz (MF-BIA), phase-sensitive single frequency (SF-BIA) at 50 kHz, and spectroscopic BIA at 5 kHz to 2 MHz. Studies have provided inconsistent results, concerning the agreement regarding single- and multiple-frequency BIA equipment⁶, including source current, frequency, total impedance range over which the current is within a specified tolerance, resolution, and accuracy of the displayed impedance, as has been described in the National Institutes of Health (NIH) Technology Assessment Conference Statement²¹. Bioelectrical impedance spectroscopy (BIS) instruments present an important limitation: R and Xc of total body raw data cannot be obtained, which must be calculated or modeled from other segmental impedance parameters, and they seem to present an under-estimation when compared with phase-sensitive single- and multiple-frequency devices^5,22. Therefore, we do not recommend the use of such technology.

On its own, MF-BIA is a phase-sensitive tetrapolar instrument that directly measures phase angle and impedance at different frequencies (5, 50, 100, 200, and 500 kHz), reporting a 0.5% deviation and an accuracy of 500 Ω for each frequency, allowing differentiation between intracellular and extracellular water, based on the principle that: at lower frequencies, the current flows through extracellular water while, at higher frequencies, it flows through a total body of water. As this kind of device provides raw data, the IR can be calculated, as has been previously described^6,23.

It is also important to consider that the type of electrode, and the specific anatomical location of electrodes, in addition to the position of the subject to be measured, can influence the raw bioelectrical values. Thus, one should avoid extrapolating results obtained with different equipment in patients who are decompensated (e.g., heart, renal, or hepatic failure), or who are suffering an acute event or another chronic disease⁶. It is essential to implement a protocol to standardize a method in order to determine fluid overload at admission. Therefore, obtaining a basal fluid distribution status allows for early and appropriate therapeutic approaches to be taken.

Clinical practice guidelines and manufacturers do not recommend BIA assessment to be performed in patients with cardiac implantable electronic devices (CIEDs), such as pacemakers and implantable cardioverter defibrillators, as it can cause electromagnetic interference due to the applied electric current. However, when the low magnitude of the electric current transferred to the body is inferior to the susceptibility limits of the CIED, and with the absence of alterations in its function, BIA is considered safe and can be performed in this group of patients²⁴.

Another consideration to take into account is that BIA and BIVA cannot be performed in patients with any amputation or with abnormal physical structure²¹.

Some limitations of the measurement technique that may not be controlled in the context of patients at emergency admission include fasting time, alcohol consumption, previous physical exercise, and bladder voiding²⁵.

When a fluid overload is detected-and based on the assumption that it is the result of fluid accumulation-the use of diuretics is frequent in clinical practice; however, it is possible that the main pathophysiological mechanisms could be related to fluid re-distribution, rather than accumulation, and high doses of furosemide may be detrimental to renal function. For example, in heart failure patients with diastolic dysfunction and pulmonary edema, high systolic blood pressure can be treated with vasodilators (nitrates), thus avoiding the use of diuretics²⁶. Therefore, it is essential to interpret the BIVA results in the context of the patient's diagnosis, physical examination, and biomarkers (e.g., hemoglobin, albumin, sodium, and creatinine).

Finally, in order to illustrate how BIVA can be used, in a previous report, we found that the patients classified as having fluid overload, according to BIVA at admission to emergency department-even with an accumulated balance of 1212 mL fluids, a value that is considered normal-showed statistically significant higher disease severity with respect to SOFA, and presented higher mortality, compared to those with normal fluid status, thus demonstrating the usefulness of BIVA in critically ill patients²⁷.

Materials

Name	Company	Catalog Number	Comments
Alcohol 70% swabs	NA	NA	Any brand can be used
BIVA software 2002	NA	NA	Is a sofware created for academic use, can be download in http://www.renalgate.it/formule_calcolatori/bioimpedenza.htm in "LE FORMULE DEL Prof. Piccoli" section
Chlorhexidine Wipes	NA	NA	Any brand can be used
Examination table	NA	NA	Any brand can be used
Leadwires square socket	BodyStat	SQ-WIRES
Long Bodystat 0525 electrodes	BodyStat	BS-EL4000
Quadscan 4000 equipment	BodyStat	BS-4000	Impedance measuring range: 20 - 1300 Ω ohms Test Current: 620 μA Frequency: 5, 50, 100, 200 kHz Accuracy: Impedance 5 kHz: +/- 2 Ω Impedance 50 kHz: +/- 2 Ω Impedance 100 kHz: +/- 3 Ω Impedance 200 kHz: +/- 3 Ω Resistance 50 kHz: +/- 2 Ω Reactance 50 kHz: +/- 1 Ω Phase Angle 50 kHz: +/- 0.2° Calibration: A resistor is supplied for independent verification from time to time. The impedance value should read between 496 and 503 Ω.