| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 All authors: Department of Surgery, University of Massachusetts Medical Center, 55 Lake Ave., N., Worcester, MA 01655.
Received October 23, 1998;
accepted after revision July 2, 1999.
Address correspondence to F. A. Anderson.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Study subjects had an extravasation of 5% dextrose in water (nonionic contrast equivalent) in one arm and 0.9% sodium chloride solution (ionic contrast equivalent) in the other. An EDA was placed over the site of infusion and connected to a power injector. Injections were performed at 0.25 ml/sec (n = 40), 2.5 ml/sec (n = 62), or 5 ml/sec (n = 20).
RESULTS. At infusion rates of 2.5 and 5 ml/sec, the device halted the injector in every subject after an average volume of 12.5 ± 1.6 ml was delivered. At 0.25 ml/sec, the device failed to halt the injector in 11 of 20 events. After reprogramming the algorithm, 10 more subjects were tested at the lowest injection rate. The device halted 18 of 20 extravasation events with an average volume of 3.7 ± 0.5 ml.
CONCLUSION. In our model of contrast extravasation, the EDA halted a power injector with reliability and reproducibility before a large volume of contrast material was delivered. The sensitivity of the device approached, but did not reach, 100%. This device may serve to diminish the morbidity of extravasation events.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although IV infusions of iodinated contrast material do not cause the measured skin impedance to vary greatly, in animal studies extravasation of ionic contrast material into soft tissue leads to measurable changes in skin impedance, with ionic extravasations decreasing impedance and nonionic extravasations increasing impedance [3]. The extravasation detection accessory (EDA; E-Z-EM, Westbury, NY) is a small biocompatible nonreactive tetrapolar electrode patch that contains conductors embedded in a hydrogel substrate to measure skin impedance at the site of IV infusion of iodinated contrast material. Through an electronic feedback loop with a power injector (Percu-Pump Touchscreen with EDA, E-Z-EM), the EDA halts the infusion of contrast material on detection of changes in impedance suggestive of extravasation.
The EDA has been proven effective in measuring impedance changes that occur with the extravasation of ionic and nonionic contrast material in animals. Furthermore, a model of extravasation that mimics the impedance changes that occur with the subcutaneous infiltration of these contrast agents has been developed and tested in human volunteers. The ionic analog used was 0.9% sodium chloride solution; 5% dextrose in water was used as a nonionic analog [3].
A clinical trial of the EDA involving 185 patients undergoing contrast-enhanced CT with a power injector was recently completed [3]. Although able to measure soft-tissue impedance in humans with the EDA, the investigators were unable to adequately evaluate device sensitivity because of the absence of extravasation events. The purpose of our study was to determine the sensitivity of this device by using a model of ionic and nonionic contrast extravasation in a group of healthy volunteers.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Extravasation of 0.9% sodium chloride solution causes changes in tissue impedance measurements similar to changes caused by ionic contrast material, and extravasation of 5% dextrose in water causes impedance changes similar to nonionic contrast material [3]. In this study, 5% dextrose in water solution ("dextrose") was injected into the left arm of the subject; the procedure was repeated with 0.9% sodium chloride solution ("saline") injected into the right arm of the subject.
A PercuPump disposable syringe set (E-Z-EM) was placed in a PercuPump injector and filled with either 5% dextrose or 0.9% saline. The injector line was flushed with infusate and connected to the angiocatheter in the subject's forearm. Once the PercuPump was primed for injection, an electrode patch was placed on the subject's forearm so that the axis of the angiocatheter was parallel to the axis of the patch electrode with the tip of the angiocatheter centered beneath the electrode. The patch was then connected to the power injector with the supplied insulated electric cable.
Reading from the electrode patch, we took a baseline impedance measurement from a sensor in the power injector. On verification of a stable reading from the sensor, we began the injection. On the basis of a randomization table created at the begining of the study, the rate of injection of contrast analog was set at 0.25 ml/sec (n = 40), 2.5 ml/sec (n = 76), or 5 ml/sec (n = 40). The same rate of injection was used in both arms of each patient. The power injector was set to shut off after 25 ml was infused in the event of failure of the EDA to halt the injection. At the completion of each injection, the site of the extravasation was inspected and the size of the wheal was measured to confirm the occurrence of a subcutaneous extravasation event. The angiocatheter was removed and a small sterile dressing was applied. All study participants were advised to follow-up with the investigators for any injection-site problems persisting beyond 24 hr. All data were collected on a floppy disk for later analysis. The Student's t test was used to compare measured variables obtained among individuals within each group and between each study group. A p value of less than 0.05 was chosen as statistically significant.
Because of problems with the algorithm as originally programmed for detecting impedance changes at the slowest rate of injection, an additional 10 subjects were enrolled at the end of the initial study to retest the EDA after software modifications were completed by the manufacturer. The outlined protocol with the injection rate of 0.25 ml/sec was used to study these subjects. Because the purpose of this additional study was simply to evaluate in vivo the ability of the device to detect extravasation and halt the power injector, impedance measurements were not recorded., Therefore, these results are reported separately below.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Women had higher baseline impedance values than men (p <
0.0002). Right arm impedance in women was 113.0 ± 19.1 
and left
arm impedance was 113.8 ± 21.4 (p = not significant).
In men, right arm impedance was 86.6 ± 15.7 and left arm
impedance was 82.0 ± 14.0 (p = not significant).
Seven women and three men had injection rates of 0.25 ml/sec. Twenty-two women and nine men had injection rates of 2.5 ml/sec. Six women and four men had injection rates of 5 ml/sec. Whereas infusion (extravasation) of 0.9% saline caused a significant decrease in impedance regardless of the rate of injection, the increase in measured impedance after infusion of 5% dextrose was not statistically significant at the infusion rates of 0.25 ml/sec in men (p = 0.24) or 5 ml/sec in either men or women (p = 0.12). Women had a greater absolute change in impedance than men during the infusion of 0.9% saline at 0.25 and 2.5 ml/sec. All other changes in impedance were similar in men and women, regardless of the rate of infusion.
The EDA failed in 11 of 20 extravasation events (six with 0.9% saline, five with 5% dextrose) at the injection rate of 0.25 ml/sec. In each of the 82 extravasation events at the more rapid rates of injection, the EDA halted the power injector before the default 25 ml of solution was delivered. Excluding the failures at the slowest rate of injection, the average volume of fluid extravasated was 12.5 ± 1.6 ml. A significantly larger volume of dextrose than saline (13.5 ± 1.7 ml versus 11.6 ± 0.6 ml, p < 0.0001) was infused into the soft tissue before extravasation was detected and the infusion was halted. No adverse events were reported from the soft-tissue swelling that resulted from the extravasation of either solution.
Because of the sensitivity problems observed at the lowest rate of injection, the manufacturers reprogrammed the algorithm for halting the power injection to adjust for the slower rate of change of impedance in this group. With this new algorithm, the manufacturers evaluated the stored data for the slow injection group (0.25 ml/sec) on a simulator and noted that the reprogrammed EDA was capable of identifying extravasation and halting the power injector before the default limit of 25 ml was reached in 19 of 20 injections.
To better validate these in vivo observations, we tested an additional 10 individuals at an infusion rate of 0.25 ml/sec. Six women and four men, 24-54 years old (mean, 36 ± 9 years) participated. Of the 20 infusions, the EDA did not halt the power injector before the delivery of the preset 25-ml injection limit in two events (one 0.9% saline and one 5% dextrose). Excluding the two failures, the average volume infused was again significantly greater for 5% dextrose than for the 0.9% saline (4.3 ± 0.7 ml versus 3.1 ± 0.3 ml, p < 0.01), with both volumes significantly smaller than those observed before the modification of the algorithm.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
For the measurement of impedance, the EDA uses the standard tetrapolar electrode configuration [9], based on the passage of a weak high-frequency alternating current between two outer electrodes with voltage changes across the field measured through two separate inner electrodes. It provides real-time values. We have shown, as have other [3], that impedance values are higher in women than in men, presumably because of a higher subcutaneous fat content [10] that acts as an electric insulator. Although the age of the women was greater than that of the men in our study, others have noted no correlation between age and mean skin impedance [3].
Although no consensus on the volume of extravasated contrast material representing a dangerous threshold has been found, surgical debridement after any extravasation of ionic contrast material greater than 20 ml has been recommended to minimize subsequent soft-tissue destruction [11]. Extravasation of nonionic contrast agents appears to result in less significant soft-tissue injury; volumes of 150 ml are tolerated without significant long-term sequelae [12, 13, 14]. In the absence of the ability to completely prevent contrast extravasation, minimizing the volume of contrast extravasation is the next best prophylactic measure.
Because of the low incidence of contrast extravasation events during power injection and the obvious inappropriateness of intentional extravasation of iodinated contrast material into human subjects, we used a model that simulated the electric changes that occur with the extravasation of ionic and nonionic contrast material. Our series of 122 extravasation events would require an evaluation of 30,500 power-injection contrast studies assuming an extravasation rate of 0.4%. This number increases to 348,571 contrast studies if the extravasation rate of 0.035% is used.
The EDA tracked impedance changes with reliability and reproducibility. Extravasation of 0.9% saline decreased impedance and 5% dextrose increased impedance. The failure to achieve statistical significance greater than baseline for both men and women receiving dextrose at 0.25 ml/sec does not explain the initial failure of the device at the slowest rate of injection; six of 11 initial failures occurred with 0.9% saline.
Because the baseline impedance values for men and women were different, relative changes in impedance values were interpreted for each group separately. The relative change in impedance was the same (p = not significant) for 5% dextrose at all rates of injection and for 0.9% saline at 5 ml/sec. It is unclear why 0.9% saline at the slower rates of injection caused a greater change in impedance in women than in men.
Regardless of the actual change in impedance from baseline, at injection rates of 2.5 and 5 ml/sec the EDA detected a change in measured impedance sufficient to limit extravasation. The device was 100% sensitive at these rates in our model. Furthermore, the EDA halted the power injector after an average of only 12.5 ml was injected into the subcutis with no large-volume (> 20 ml [12]) events occurring. Although a significantly larger volume of dextrose than saline was injected before termination of the injection, both volumes were small by clinical standards. None of the study participants reported a delayed problem with the extravasation site.
Although many subjects in our study complained of stinging discomfort at the completion of the infusion of saline, most noted no discomfort during or after the infusion of dextrose. This report is consistent with the findings of others who have noted an alert patient who had large-volume extravasation events that were discovered by the technologist rather than the patient at the completion of a contrast study [15]. Furthermore, unconscious patients or those at either extreme of age may not be able to communicate a problem with a contrast infusion to a technologist working in a separate room. The benefits of the EDA are obvious in such situations.
The EDA has other clinical potential. Three different types of agents are responsible for soft-tissue destruction on extravasation: hyperosmolar agents (e.g., iodinated contrast material, parenteral nutrition), vasopressor agents, and cytotoxic agents (i.e., chemotherapeutic agents) [16]. Extravasation of the latter two types of agents should theoretically also lead to measurable changes in impedance. Although the EDA is designed to halt the infusion of contrast material by power injection, the device can be configured to work at rates of injection slower than those used in this study (Anderson FA, Zimmit A, Williams R, unpublished data). The applicability of impedance-change measurements to the attenuation of extravasation of other agents remains open for investigation.
The modified algorithm for detecting extravasation at the lowest tested rate of injection improved the sensitivity of the device from 45% to greater than 90% and significantly reduced the volume of extravasate. Although that was an obvious improvement over the initial algorithm, the ideal device would have a 100% sensitivity; these failures serve as a reminder that technologic advances are not a replacement for careful focused observation of the patients in our care.
Although the EDA cannot be used in every situation that power injection of iodinated contrast material is used (e.g., contrast agents delivered through a central vein catheter), its use in appropriate clinical situations should dramatically reduce the morbidity (and associated cost and liability) of extravasation events, regardless of the type of contrast agent used. More significantly, it will improve patient safety.
Our results are based on a model of extravasation that mimics the impedance changes occurring with the extravasation of iodinated contrast material into soft tissue. Our model tested only the most extreme form of extravasation and not a partial extravasation of a vein. Our data at the lowest rate of injection, however, are more representative of a partial extravasation because power injectors are not routinely used at 0.25 ml/sec. This study was designed to test the sensitivity of a new method to limit extravasation during contrast-enhanced CT. It was not designed to provide data regarding the specificity or false-positive rate of this method. Additional studies are underway that will address the issue of the rate of false-positive results in clinical practice.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
