Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Birnbaum, B. A.
Right arrow Articles by Glick, S. N.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Birnbaum, B. A.
Right arrow Articles by Glick, S. N.
(Radiology. 1999;212:431-438.)
© RSNA, 1999

Gastrointestinal Imaging

Extravasation Detection Accessory: Clinical Evaluation in 500 Patients1

Bernard A. Birnbaum, MD, Rendon C. Nelson, MD, Judith L. Chezmar, MD and Seth N. Glick, MD

1 From the Departments of Radiology, Hospital of the University of Pennsylvania, 3400 Spruce St, Philadelphia, PA 19104 (B.A.B.); Duke University Medical Center, Durham, NC (R.C.N.); Emory University Hospital, Atlanta, Ga (J.L.C.); and Hahnemann University Medical Center, Philadelphia, Pa (S.N.G.). Received July 30, 1998; revision requested September 4; revision received October 29; accepted February 15, 1999. Address reprint requests to B.A.B. (e-mail: birnbaum@oasis.rad.upenn.edu).


   Abstract
TOP
 Abstract
Introduction
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
PURPOSE: To assess the ability of an extravasation detection accessory (EDA) to detect clinically important (>=10 mL) extravascular injection of iodinated contrast material delivered with an automated power injector.

MATERIALS AND METHODS: Five hundred patients referred for contrast material–enhanced body computed tomography (CT) participated in a prospective, multiinstitutional, observational study in which the EDA was used to identify and interrupt any injection associated with clinically important extravasation. The presence or absence of extravasation was definitively established with helical CT at the injection site (injection rate, from 1.0 to 5.0 mL/sec; mean, 2.9 mL/sec; median, 3.0 mL/sec).

RESULTS: There were four true-positive (extravasation volumes, 13–18 mL), 484 true-negative, 12 false-positive, and no false-negative cases. The prevalence of overall and clinically important (>=10 mL) extravasation was 3.6% (18 of 500 cases) and 0.8% (four of 500 cases), respectively. The EDA had a sensitivity of 100% (four of four cases; 95% CI: 51%, 100%) and a specificity of 98% (484 of 496 cases; 95% CI: 96%, 99%) in the detection of clinically important extravasation.

CONCLUSION: The EDA is easy to use, safe, and accurate in the monitoring of intravenous injections for extravasation, which could prove especially useful in high-flow-rate CT applications.

Index terms: Angiography, complications, **.442 • Computed tomography (CT), contrast media • Contrast media, complications


   Introduction
TOP
 Abstract
 Introduction
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Automated power injectors have become standard equipment in radiology departments practicing state-of-the-art body computed tomography (CT) (1). These devices enable contrast material to be delivered as a uniform, high-flow-rate, nonfragmented bolus, which is essential for many contrast material–enhanced CT applications (2,3). This is particularly true for helical CT protocols such as CT angiography, hepatic biphasic CT, and pancreatic biphasic CT, in which contrast material infusion rates of 3.0–5.0 mL/sec have been recommended (4–7). The major risk associated with use of automated power injectors is the well-known complication of contrast material extravasation at the injection site. Previous articles (2,3,8–11) have shown that extravascular injection of contrast material occurs with a prevalence of 0.04%–1.3%. The vast majority of these extravasations involve small volumes of contrast material that result in mild soft-tissue injury and require only supportive care (8–14). Severe skin ulceration and necrosis are extremely rare but can result from extravasation volumes as small as 10 mL (15). Although conventional ionic contrast media are more toxic to the skin than are nonionic agents, extravasation of either type of contrast material may cause direct soft-tissue injury or neurovascular compromise secondary to compartment syndrome (13,16–24). These catastrophic injuries often necessitate plastic surgery or emergent fasciotomy and may place the radiologist at risk for litigation.

Traditionally, efforts to avoid morbidity from extravasation injuries have focused on prevention by identifying patients at risk and following recommended guidelines (13). We recently described an extravasation detection accessory (EDA) that is capable of detecting and interrupting automated injections complicated by extravasation (25). The purpose of this study was to assess the ability of this device to detect clinically important episodes of extravasation.


   MATERIALS AND METHODS
TOP
 Abstract
 Introduction
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 References
 
We conducted a prospective, nonblinded, observational study in which the EDA was used to monitor automated injection of contrast material in 500 patients at four institutions. The research protocol was approved by an institutional review board at each of the participating centers.

Patients
The study population consisted of 500 adult patients who were referred for routine contrast-enhanced CT at the Hospital of the University of Pennsylvania (n = 125), Duke University Medical Center (n = 157), Emory University Hospital (n = 168), and Hahnemann University Medical Center (n = 50) between April 24, 1997, and November 6, 1997. There were 257 men and 243 women with a mean age of 58 years (range, 19–90 years), mean weight of 77 kg (range, 38–155 kg), and mean height of 170 cm (range, 137–206 cm). Patients were considered eligible for study participation if they were older than 18 years, willing and able to provide informed consent, scheduled to undergo routine contrast-enhanced CT (either as an inpatient or outpatient), and if they had at least one healthy, functioning arm without evidence of arterial, venous, or lymphatic insufficiency. Patients were excluded from participation if they were younger than 18 years, unable or unwilling to provide informed consent, unable or unwilling to comply with the protocol requirements, or if they had previously undergone mastectomy on the ipsilateral side of the injection site. The patients included in this study had not participated in any prior research involving the EDA and have not been previously reported on.

Scanning Protocol
All CT examinations (CT HiSpeed Advantage, GE Medical Systems, Milwaukee, Wis; CTI, GE Medical Systems; Somatom Plus 4; Siemens Medical Systems, Erlangen, Germany) were performed according to the standard of care as practiced in each institution. Intravenous access was established by a trained physician, nurse, or technologist. A new plastic venous cannula was placed in the peripheral arm vein in 479 (96%) of the 500 patients, whereas a preexistent intravenous catheter was secured for injection in 21 (4%). Existing catheters were used if they flushed easily with saline solution, demonstrated acceptable blood return, and showed no evidence of phlebitis. Injection sites for the 500 patients were located in the antecubital fossa in 391 (78%), the forearm in 49 (10%), the back of the hand in 37 (7%), and the wrist in 23 (5%). The left arm was used in 261 (52%) patients, and the right arm was used in 239 (48%). All intravenous cannulas were 20 or 22 gauge. Twenty-gauge cannulas, 1.0 inch (2.54 cm) in length, were used in 430 (86%) of the 500 patients, and 22-gauge cannulas, 1.25 inches (3.18 cm) in length, were used in 70 (14%). Metallic butterfly needles were not used in this study.

The appropriate type, concentration, and volume of intravenous contrast material was loaded into the power injector syringe (PercuPump II; E-Z-Em, Westbury, NY). Contrast material selection was based on patient history, the type of CT study to be performed, and individual institutional preference for contrast media class usage. Nonionic, low-osmolality contrast material (Omnipaque 300, Nycomed, Princeton, NJ; Isovue 300, Bristol-Myers Squibb, New Brunswick, NJ) was used in 349 (70%) of the 500 patients, whereas ionic, high-osmolality contrast material (Conray 60, Mallinckrodt Medical, St Louis, Mo; Hypaque 60, Nycomed, New York, NY) was used in 151 (30%). The mean intended volume of injection was 132 mL (range, 50–184 mL; median, 140 mL). The injection flow rate was determined by the supervising physician or nurse on the basis of the size of the venous cannula and the particular CT application. The injection rates used in this study were 1.0 mL/sec (n = 6), 1.2 mL/sec (n = 1), 1.5 mL/sec (n = 21), 1.8 mL/sec (n = 2), 2.0 mL/sec (n = 58), 2.2 mL/sec (n = 3), 2.3 mL/sec (n = 4), 2.5 mL/sec (n = 16), 3.0 mL/sec (n = 319), 3.5 mL/sec (n = 4), 4.0 mL/sec (n = 64), and 5.0 mL/sec (n = 2). The mean flow rate was 2.9 mL/sec (range, 1.0–5.0 mL/sec; median, 3.0 mL/sec). The distribution of flow rates appeared Gaussian (Fig 1). The maximal flow rate was 5.0 mL/sec for 20-gauge cannulas and 3.5 mL/sec for 22-gauge cannulas. A physician or nurse was available to manually palpate and visually monitor the injection site for extravasation during injection initiation (before scanning) in all patients (n = 500). At one institution (n = 157), a nurse remained with the patient to monitor the entire course of the intravenous injection. Radiation exposure to monitoring personnel was minimized in these cases with use of lead aprons.



View larger version (13K):
[in this window]
[in a new window]
  		Figure 1. Frequency distribution of flow rates. A Gaussian distribution is identified, centered near the median injection rate of 3.0 mL/sec.

 
The EDA was used to monitor electrical impedance at the percutaneous injection site in all patients. A detailed description of the appearance of this device and how it functions has been previously reported (25). In brief, the EDA consisted of a pliable, adhesive-backed electrode patch that was placed on the patient's arm directly above the tip of the cannula. The sensor patch was positioned so that the array of the electrode was in direct line with the tip of the cannula, with the axis of the cannula parallel to the axis of the array. The patch was then coupled, via connecting cable, to a dedicated power injector, where the EDA electronic processing module was located. Device setup time for nursing personnel was less than 20 seconds. The EDA determined the baseline impedance of the skin and underlying subcutaneous tissues and then monitored the injection site for a change in this value during delivery of the contrast material. The rate of change or the slope of impedance over time was used to trigger an extravasation event, which automatically discontinued the injection and set the power injector into a pause mode (25). The EDA was calibrated to detect extravasation at a clinically important 10-mL threshold level.

If the injection was interrupted, as occurred in 16 patients, the injection site was evaluated by monitoring personnel to determine whether it was necessary to reposition the venous cannula to a new injection site and attempt another injection (n = 2), abort the injection entirely and continue CT with whatever amount of contrast material may have been successfully administered (n = 6), or reposition the arm and attempt another injection (n = 8). The power injection could not be reinstated unless it was deemed appropriate by the physician or nurse supervising the study. Impedance monitoring was halted and the sensor electrode removed at the conclusion of the injection. The power injectors used in this study were coupled to computer equipment, which collected the impedance data on floppy disks. This enabled ex post facto data analysis of concurrent clinical observations and device recordings and provided the device manufacturer with impedance data that could be used to analyze and improve the software algorithm of the EDA.

Proof of Diagnosis
After each patient underwent conventional contrast-enhanced CT, he or she underwent limited, low-dose (multiple scan average dose to arm, 0.42 cGy), 3-second helical CT of the injection site to definitively determine the presence or absence of extravascular contrast material. This CT examination was prescribed after a scout radiograph of the arm was acquired to precisely locate the cannula tip and was performed with 120 kVp, 120 mAs, 10-mm section collimation, 2:1 helical pitch, and contiguous section reconstruction. If extravasation was identified radiologically, its volume was quantitatively estimated by an investigator (B.A.B., R.C.N., J.L.C., S.N.G.) at each clinical site. This was accomplished by using the trace function software of the CT scanner to manually draw a line around the periphery of the extravasated contrast material on each axial image in which extravasation was identified. The trace function software automatically calculated the area of the region of interest. Data for the regions of interest containing extravasated contrast material were then summed together to provide an estimate of total extravasation volume.

Definitions
Extravasation was defined as the presence of any amount of extravascular contrast material at CT. Extravasation was considered clinically important if the volume of extravasation was at least 10 mL. Study definitions of outcome classification were based on the fact that the EDA was calibrated to detect extravasation at a threshold of 10 mL. True-positive cases were defined as those in which the EDA interrupted the injection and CT demonstrated clinically important extravasation. False-positive cases were defined as those in which the EDA interrupted the injection and CT demonstrated either no extravasation or extravasation of less than 10 mL. True-negative cases were defined as those in which the injection proceeded without interruption and CT demonstrated either no extravasation or extravasation of less than 10 mL. False-negative cases were defined as those in which the injection proceeded without interruption and CT demonstrated clinically important extravasation.

Statistical Analysis
The primary outcomes analyzed for the study population were the prevalence of overall extravasation and the prevalence of clinically important extravasation. In addition, the sensitivity, specificity, positive predictive value, and negative predictive value of the EDA for detecting clinically important extravasation were determined by using the foregoing definitions for outcome classification; 95% CIs were calculated.


   RESULTS
TOP
 Abstract
 Introduction
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 References
 
Four hundred eighty-four (96.8%) of the 500 injections were determined to be true-negative cases. These included 470 cases in which the injection proceeded without interruption, the patient was asymptomatic, and CT demonstrated no radiologic evidence of contrast material extravasation (Fig 2). Also included in this category were 13 cases in which the injection proceeded without interruption and the patient remained asymptomatic; however, CT demonstrated trace extravasation of contrast material at the tip of the venous cannula (Fig 3). The volumes of these "subclinical" extravasations ranged from 0.5 to 3.0 mL. The last true-negative case involved a patient who received an antecubital fossa injection of nonionic contrast material delivered at 3 mL/sec through a newly established, 20-gauge intravenous catheter. This injection was interrupted by the nurse monitoring the injection site, who noted visual evidence of extravasation after 17 mL of contrast material had been administered. CT helped confirm the presence of extravasation, which was estimated to be 8–9 mL. Presumably, 8–9 mL remained within the intravascular space in this patient. Analysis of the impedance data from this injection revealed that the extravasation threshold of the EDA algorithm had not been reached and that the nurse detected the extravasation before the EDA was configured to trigger.



View larger version (114K):
[in this window]
[in a new window]
  		Figure 2a. True-negative case of nonextravasated injection. (a) CT scout view of the injection site shows no evidence of extravasated contrast material near the tip (arrow) of the radiopaque cannula. (b) Multiimage display of axial CT scans obtained at the injection site demonstrates that the injected vein (arrow) has a normal appearance, with no perivenous extravasation.

 


View larger version (160K):
[in this window]
[in a new window]
  		Figure 2b. True-negative case of nonextravasated injection. (a) CT scout view of the injection site shows no evidence of extravasated contrast material near the tip (arrow) of the radiopaque cannula. (b) Multiimage display of axial CT scans obtained at the injection site demonstrates that the injected vein (arrow) has a normal appearance, with no perivenous extravasation.

 


View larger version (141K):
[in this window]
[in a new window]
  		Figure 3a. True-negative case with trace extravasation. (a) CT scout view of the injection site shows a very small amount of extravasated contrast material (arrow) adjacent to the tip of the cannula. (b) Multiimage display of axial CT scans obtained at the injection site demonstrates perivenous extravasation (arrow), which was estimated to have a volume of 1.7 mL.

 


View larger version (129K):
[in this window]
[in a new window]
  		Figure 3b. True-negative case with trace extravasation. (a) CT scout view of the injection site shows a very small amount of extravasated contrast material (arrow) adjacent to the tip of the cannula. (b) Multiimage display of axial CT scans obtained at the injection site demonstrates perivenous extravasation (arrow), which was estimated to have a volume of 1.7 mL.

 
Four (0.8%) of the 500 injections were true-positive cases in which the EDA interrupted the injection and CT documented clinically important extravasation (Fig 4) (Table). These extravasations occurred at initiation of injection in one patient and during the course of the injection in three patients. This determination was made by comparing the volumes of extravasation (range, 13–18 mL) with the volumes of contrast material actually delivered with the power injector (range, 18–77 mL). Visual signs of extravasation (swelling or discoloration at the percutaneous injection site) were noted in all four cases, although this was very subtle in one asymptomatic patient and was identified only in retrospect after the EDA paused the injection. Most extravasations occurred in patients with newly established intravenous catheters (n = 3), 20-gauge cannulas injected at 3.0 mL/sec (n = 3), and nonionic contrast material (n = 3). Extravasation sites included the wrist (n = 2), antecubital fossa (n = 1), and forearm (n = 1).



View larger version (117K):
[in this window]
[in a new window]
  		Patient 1. True-positive case of clinically important extravasation. (a) CT scout view of the injection site shows extravasated contrast material (arrow) within the subcutaneous tissue surrounding the cannula. (b) Multiimage display of axial CT scans obtained at the injection site demonstrates perivenous extravasation (arrow), which was estimated to have a volume of 18 mL.

 


View larger version (141K):
[in this window]
[in a new window]
  		Patient 1. True-positive case of clinically important extravasation. (a) CT scout view of the injection site shows extravasated contrast material (arrow) within the subcutaneous tissue surrounding the cannula. (b) Multiimage display of axial CT scans obtained at the injection site demonstrates perivenous extravasation (arrow), which was estimated to have a volume of 18 mL.

 

View this table:
[in this window]
[in a new window]
  		True-Positive Cases: Clinically Important Extravasation
 
There were no false-negative cases. All injections complicated by extravascular injection of at least 10 mL of contrast material were correctly detected and interrupted by the EDA .

Twelve (2.4%) of the 500 injections were determined to be false-positive cases because the EDA interrupted the injection and CT failed to document extravasation. All false-positive injections occurred at the antecubital fossa in patients with newly established intravenous catheters. These injections were primarily performed by using 20-gauge cannulas (n = 11) and nonionic contrast material (n = 11), with flow rates varying from 3.0 to 5.0 mL/sec (mean, 3.5 mL/sec; median, 3.0 mL/sec). The volume of contrast material delivered at suspension of injection varied from 8.0 to 124 mL (mean, 53 mL). In four (33%) of these 12 cases, the EDA paused the injection after only 8–9 mL of contrast material had been delivered. Nursing personnel monitoring the injection site were believed to have palpated the EDA sensor patch in these four cases as well as in five others during impedance monitoring as contrast material was being infused. One false-positive case was believed to have resulted from vigorous patient tremors during the injection. The cause of the EDA miscue in the remaining two cases is unknown.

The prevalence of overall extravasation and clinically important (>=10 mL) extravasation was 3.6% (18 of 500 cases) and 0.8% (four of 500 cases), respectively. The EDA demonstrated a sensitivity of 100% (four of four cases; 95% CI: 51%, 100%) and a specificity of 98% (484 of 496 cases; 95% CI: 96%, 99%) in the detection of clinically important extravasation. The positive predictive value of an EDA response was 25% (four of 16 cases; 95% CI: 10%, 50%), whereas the negative predictive value was 100% (484 of 484 cases; 95% CI: 99%, 100%).


   DISCUSSION
TOP
 Abstract
 Introduction
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
References
 
Traditionally, efforts to avoid extravasation have focused on prevention by exercising special care in high-risk groups. These groups include patients who are uncommunicative or unlikely to complain of pain (infants and children, elderly, or unconscious) and severely debilitated or chronically ill patients who are at increased risk for extravasation and severe sequelae (12,13). Extravasation is more common when tourniquets are not released (24), contrast material is administered with metallic cannulas (26), multiple attempts are made to secure intravenous access (27), and injections are performed by using indwelling intravenous catheters that have been in place for more than 20 hours (28). In addition, substantial morbidity may result when extravasation occurs in the dorsum of the hand, foot, and wrist (24,29) or in an injected limb compromised by arterial, venous, or lymphatic insufficiency (13). As a result, it is advised that automated injection of contrast material be performed by using newly established intravenous catheters located at acceptable percutaneous injection sites and that trained personnel directly monitor the injection site for signs of extravasation (9,13).

Extravascular injection of contrast material is a well-known complication of contrast-enhanced CT (8–14). In most practices, the reported frequency of extravasation is less than 1% (3,8–11). Although the number of reported extravasations has increased since automated injection of contrast material replaced drip-infusion and hand-bolus injection techniques, the relationship between extravasation frequency and injection rate remains controversial. Sistrom et al (9) detected extravasation in six (0.1%) of 5,598 patients examined with an injection rate of 0.5 mL/sec and noted that the extravasation rate more than doubled (13 [0.25%] of 5,294) when the rate of infusion increased to 1.0–1.5 mL/sec. Recent studies performed with higher injection rates suitable for helical CT, however, have failed to demonstrate such a relationship (10,11). Federle et al (10) retrospectively identified extravasation in 48 (0.9%) of 5,106 patients examined with injection rates of 1.0–5.0 mL/sec and were unable to identify a linear trend between extravasation frequency and injection rate. Similarly, Jacobs et al (11) prospectively detected extravasation in 41 (0.6%) of 6,660 patients examined with injection rates of 0.5–4.0 mL/sec and showed no correlation between flow rate and frequency of extravasation. Neither study demonstrated correlation between volume of extravasation and injection rate.

Although most radiologists exercise diligence and follow recommended guidelines, extravasation of radiologic contrast material continues to be an event that occurs on the order of three to four times per month during contrast-enhanced CT. Moreover, automated power injection of contrast material may result in large-volume (50–150 mL) extravasation because the mechanical injector will continue to infuse the contrast material at a fixed pressure regardless of whether the injection is intra-or extravascular (8,13). Modern injectors, such as the device used in this investigation, have the ability to automatically suspend an injection when a maximum pressure limit is exceeded; however, this technique has never been shown to be an effective means of detecting or preventing clinically important extravasation. Large-volume extravasation may result in severe damage to extravascular tissue. This complication is most likely to occur when the injection site is not monitored throughout injection of the entire bolus and when uncommunicative or asymptomatic patients experience a "silent" extravasation. In the first instance, monitoring personnel may leave the CT room to avoid radiation exposure and are unavailable to detect a delayed extravasation event that occurs after initiation of injection of the bolus of contrast material. In the second instance, monitoring personnel may fail to detect nonsuperficial extravasation if the patient appears to be in no distress. In our experience, this most commonly occurs in patients who experience painless extravasation of nonionic contrast material in the antecubital fossa (14).

Results of recent studies have shown that the overwhelming majority of patients who develop an extravasation tolerate it well and experience only brief and self-limited symptoms (10,14). Severe extravasation injuries are rare, but they can cause substantial morbidity. Severe injuries usually manifest as tissue necrosis and may require extensive débridement with split- or full-thickness skin graft procedures (9,13,14). Subfascial extravasation of large volumes of contrast material may result in compartment syndrome, which may necessitate emergency fasciotomy to relieve neurovascular compromise (10,13). Other reported sequelae of extravasation include hypoesthesia, marked extremity deformity, weakness, persistent pain, decreased range of motion, and difficulty performing activities of daily living (13,14). Many radiologists prefer to inject low-osmolality, nonionic contrast material to reduce the risk of severe extravasation injury. This practice is based, in part, on results in animal studies that have demonstrated nonionic agents are less toxic to the skin and subcutaneous tissues than are conventional ionic contrast media (16–19). Unfortunately, extravasation of moderate to large volumes of nonionic contrast material may still result in clinically important soft-tissue injury (14,21–23). Because the risk of injury often correlates with extravasation volume, every effort should be made to limit the extent of extravasation regardless of the type of contrast material used.

The EDA evaluated in this study is a U.S. Food and Drug Administration–approved, commercially available product designed to detect and interrupt intravenous injections complicated by low-volume extravasation (25). The EDA electrode patch serves to monitor tissue impedance at the percutaneous injection site and is configured to detect the change in impedance that characteristically occurs with extravascular injection of contrast material. The device is calibrated to detect extravasation at a threshold of 10 mL and to suspend the injection before 20 mL of contrast material is extravasated.

We found the EDA to be easy to use, safe, and accurate for monitoring intravenous injections for extravasation. The prevalence of clinically important extravasation in this study was 0.8% (four of 500 cases), which is comparable to extravasation rates recently reported in the literature (10,11). The device demonstrated a sensitivity of 100% (95% CI: 51%, 100%) and a specificity of 98% (95% CI: 96%, 99%) in the detection of clinically important extravasation. All true-positive injections were detected and suspended by the EDA, which limited the volume of extravasation to 13–18 mL. No false-negative cases occurred; therefore, the negative predictive value of an uninterrupted injection was 100% (95% CI: 99%, 100%).

The prevalence of overall extravasation in this study was 3.6% (18 of 500 cases), a finding that exceeds all previously reported extravasation rates (2,3,8–11). This reflects the fact that we used CT as the standard of reference, which enabled detection of 13 (2.6%) cases of trace extravasation (0.5–3.0 mL) of contrast material at the cannula tip in totally asymptomatic patients. These extravasations would have gone undetected if we had used clinical criteria for diagnosis. These cases were classified as true-negative because the extravasation volume did not reach the EDA threshold. The device functioned properly in these patients because the EDA algorithm was specifically designed to prevent clinically unimportant trace extravasation from interrupting the injection of contrast material (25). There was one additional true-negative case in which CT demonstrated radiologic evidence of low-volume extravasation (8.0–9.0 mL). This injection was suspended by a monitoring nurse who identified a very small, superficial extravasation before the EDA was activated. Ex post facto analysis of this patient's impedance data revealed that the EDA algorithm was functioning properly and that the nurse interrupted the injection before the EDA reached the required threshold for triggering interruption of the bolus injection. This case emphasizes the fact that diligent monitoring personnel will occasionally detect a superficial extravasation before the EDA. This will result in a slight reduction in extravasation volume, which may theoretically benefit the patient.

The posttest probability or predictive value of a positive test in this study was only 25%, which reflects the presence of the four true-positive and 12 false-positive cases. The low positive predictive value was not unexpected given the fact that we applied a very sensitive (100%) and specific (98%) test to a population with a very low (0.8%) pretest probability of disease (extravasation). It is not unusual for most positive test results to be false-positive when an excellent screening test is applied to a population with a low pretest probability of disease (30). Nevertheless, it is important to understand why these false-positive cases occurred because unnecessary interruption of uncomplicated injections can adversely affect scan quality and patient throughput.

Nine (75%) of the 12 false-positive cases were believed to have resulted from inadvertent manipulation of the EDA electrode patch during infusion of contrast material. This occurred when nursing personnel vigorously palpated the sensor patch within seconds of bolus initiation in four patients and while monitoring the course of the injection in five. Motion at the percutaneous injection site may have contributed to a 10th false-positive case in which a patient experienced multiple tremors during the injection. Ex post facto analysis of the impedance data in these 10 cases revealed that the false-positive findings caused by motion at the injection site may be averted by modifying the EDA software algorithm. As a result, the device manufacturer has since implemented this improvement to the EDA system. The cause of the last two false-positive readings is unknown. Further research is necessary to determine whether additional physiologic factors contribute to the impedance characteristics of intravenous injections. If this is the case, it may be possible to filter out additional signals that cause false-positive readings.

Several limitations of our study must be addressed. First, although the 0.8% prevalence of clinically important extravasation in this study is comparable to recently reported extravasation rates (10,11), we identified only four true-positive cases. It would have been optimal if a greater number of true-positive injections occurred, since this would have helped verify the ability of the EDA to detect all cases of clinically important extravasation. The small number of true-positive cases also precluded us from attempting to correlate injection variables with extravasation. Second, we determined the volume of extravasation by drawing area tracings around extravascular contrast material on axial CT scans and then summing data from these regions of interest together. This is not an exact measurement technique and may have resulted in slight under- or overestimation of extravasation volume. We believe that this technique allowed us to reasonably approximate extravasation volume to within 1.0–1.5 mL in those cases in which extravasation exceeded 10 mL, and it may have been more accurate for smaller extravasations in which it was easier to identify and trace the entire periphery of the extravasation. Finally, our study population was biased toward patients who received new intravenous catheters. Only 4% of our study subjects received injections via preexisting catheters. It is possible that our results may not be directly applicable to an inpatient population preferentially presenting with preexisting intravenous access. Nevertheless, it is encouraging that the EDA functioned properly and suspended an intravenous injection complicated by delayed extravasation in a patient studied with an existing intravenous catheter.

In summary, the EDA evaluated in this study demonstrated a sensitivity of 100% (95% CI: 51%, 100%) and a specificity of 98% (95% CI: 96%, 99%) in the detection of clinically important extravasation. Extravasation volume was limited to a maximum of 18 mL, and use of CT as the standard of reference helped confirm that the bolus of contrast material was not interrupted in patients who experienced inconsequential extravasation at the cannula tip. We believe that this device represents a technologic breakthrough in dynamic contrast-enhanced CT and has the potential to eliminate nearly all clinically important contrast material extravasation events, whether with ionic or nonionic contrast media. The device should prove especially useful for high-flow-rate CT applications because the EDA algorithm is configured to suspend extravasated injections at a threshold of 10 mL, regardless of injection rate. Its use may be particularly efficacious in patients who are at increased risk for extravasation and severe sequelae, including patients with preexisting intravenous catheters.

Further study is needed to assess whether successful implementation of the device will obviate support personnel to monitor the injection site. If so, the device has the potential to reduce radiation exposure to support personnel and simultaneously reduce costs and improve efficiency by permitting these individuals to be reassigned to other tasks (25). Although the EDA detected all cases of clinically important extravasation in this study, the number of extravasations was small. Additional investigation in a patient population with a higher prevalence of extravasation is necessary to confirm the high sensitivity that we report. A large-scale multicenter clinical trial is currently under way to determine the sensitivity and specificity of the modified EDA algorithm. This software enhancement should reduce the false-positive rate of the EDA while maintaining the high sensitivity and specificity shown in this study. If this is achieved, radiologists will be able to power inject contrast material with a higher level of safety and confidence.


   Footnotes
 
**. Multiple body systems Back

Abbreviation: EDA = extravasation detection accessory

Author contributions: Guarantors of integrity of entire study, all authors; study concepts, B.A.B.; study design, B.A.B.; definition of intellectual content, B.A.B., R.C.N., J.L.C.; literature research, B.A.B.; clinical studies, all authors; data acquisition, all authors; data analysis, B.A.B.; statistical analysis, B.A.B.; manuscript preparation, B.A.B.; manuscript editing, B.A.B., R.C.N., J.L.C.; manuscript review, B.A.B., R.C.N., J.L.C.


   References
TOP
 Abstract
 Introduction
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 

  1. Silverman PM, Kohan L, Ducic I, et al. Imaging of the liver with helical CT: a survey of scanning techniques. AJR 1998; 170:149-152.[Abstract/Free Full Text]
  2. Shuman WP, Adam JJ, Shoenecker SA, Tazioli PR, Moss AA. Use of a power injector during dynamic computed tomography. J Comput Assist Tomogr 1986; 10:1000-1002.[Medline]
  3. Miles SG, Rasmussen JF, Litwiller T. Safe use of an intravenous power injector for CT: experience and protocol. Radiology 1990; 176:69-70.[Abstract/Free Full Text]
  4. Rubin GD, Dake MD, Napel S, et al. Spiral CT of renal artery stenosis: comparison of three-dimensional rendering techniques. Radiology 1994; 190:181-189.[Abstract/Free Full Text]
  5. Hollett MD, Jeffrey RB, Jr, Nino-Murcia M, et al. Dual-phase helical CT of the liver: value of arterial phase scans in the detection of small (<=1.5 cm) malignant hepatic neoplasms. AJR 1995; 164:879-884.[Abstract/Free Full Text]
  6. Oliver JH, III, Baron RL. Helical biphasic contrast-enhanced CT of the liver: technique, indications, interpretation and pitfalls. Radiology 1996; 201:1-14.[Abstract/Free Full Text]
  7. Lu DSK, Vedantham S, Krasny RM, Kadell B, Berger WL, Reber HA. Two-phase helical CT for pancreatic versus hepatic phase enhancement of tumor, pancreas and vascular structures. Radiology 1996; 199:697-701.[Abstract/Free Full Text]
  8. Cohan RH, Dunnick NR, Leder RA, Baker ME. Extravasation of nonionic radiologic contrast media: efficacy of conservative treatment. Radiology 1990; 176:65-67.[Abstract/Free Full Text]
  9. Sistrom CL, Gay SB, Peffley L. Extravasation of iopamidol and iohexol during contrast-enhanced CT: report of 28 cases. Radiology 1991; 180:707-710.[Abstract/Free Full Text]
  10. Federle MP, Chang PJ, Confer S, Ozgun B. Frequency and effects of extravasation of ionic and nonionic contrast media during rapid bolus injection. Radiology 1998; 206:637-640.[Abstract/Free Full Text]
  11. Jacobs JE, Birnbaum BA, Langlotz CP. Contrast media reactions and extravasation: relationship to intravenous injection rates. Radiology 1998; 209:411-416.[Abstract/Free Full Text]
  12. Burd DAR, Santis G, Milward TM. Severe extravasation injury: an avoidable iatrogenic disaster. BMJ 1985; 290:1579-1580.
  13. Cohan RH, Ellis JH, Garner WL. Extravasation of radiologic contrast material: recognition, prevention, and treatment. Radiology 1996; 200:593-604.[Free Full Text]
  14. Cohan RH, Bullard MA, Ellis JH, et al. Local reactions after injection of iodinated contrast material: detection, management and outcome. Acad Radiol 1997; 4:711-718.[Medline]
  15. Ayre-Smith G. Tissue necrosis following extravasation of contrast media (letter). J Can Assoc Radiol 1982; 33:104.[Medline]
  16. Kim SH, Park JH, Kim YI, Kim CW, Han MC. Experimental tissue damage after subcutaneous injection of water-soluble contrast media. Invest Radiol 1990; 25:678-685.[Medline]
  17. McAlister WH, Kissane JM. Comparison of soft-tissue effects of conventional ionic, low osmolar ionic and nonionic iodine containing contrast material in experimental animals. Pediatr Radiol 1990; 20:170-174.[Medline]
  18. Cohan RJ, Leder RA, Bolick D, et al. Extravascular extravasation of radiographic contrast media: effects of conventional and low-osmolar agents in the rat thigh. Invest Radiol 1990; 25:504-510.[Medline]
  19. Elam EA, Dorr RT, Lagel KE, Pond GE. Cutaneous ulceration due to contrast extravasations: experimental assessment of injury and potential antidotes. Invest Radiol 1991; 26:13-16.[Medline]
  20. Leung PC, Cheng CY. Extensive local necrosis following the intravenous use of x-ray contrast medium in the upper extremity. Br J Radiol 1980; 53:361-364.[Medline]
  21. Pond GD, Dorr TR, McAleese KA. Skin ulceration from extravasation of low-osmolality contrast medium: a complication of automation. AJR 1992; 158:915-916.[Free Full Text]
  22. Memolo M, Dyer R, Zagoria RJ. Extravasation injury with nonionic contrast material (letter). AJR 1993; 160:203-204.[Free Full Text]
  23. Young RA. Injury due to extravasation of nonionic contrast material (letter). AJR 1994; 162:1499.[Medline]
  24. Upton J, Mulliken JB, Murray JE. Major intravenous extravasation injuries. Am J Surg 1979; 137:497-506.[Medline]
  25. Nelson RC, Anderson FA, Jr, Birnbaum BA, Chezmar JC, Glick SN. Contrast media extravasation during dynamic CT: detection with an extravasation detection accessory. Radiology 1998; 209:837-843.[Abstract/Free Full Text]
  26. Gothlin J. The comparative frequency of extravasal injection at phlebography with steel and plastic cannulas. Clin Radiol 1972; 23:183-184.[Medline]
  27. Stanley MD, Meister E, Fuschuber K. Infiltration during intravenous therapy in neonates: comparison of Teflon and Vialon catheters. South Med J 1992; 85:883-886.[Medline]
  28. Lewis GBH, Heckler JF. Radiological examination of failure of intravenous infusions. Br J Surg 1991; 78:500-501.[Medline]
  29. Gault DT. Extravasation injuries. Br J Plast Surg 1993; 46:91-96.[Medline]
  30. Riegelman RK, Hirsch RP. Studying a study and testing a test: how to read the health science literature 3rd ed. New York, NY: Little, Brown, 1996; 183-191.




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Birnbaum, B. A.
Right arrow Articles by Glick, S. N.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Birnbaum, B. A.
Right arrow Articles by Glick, S. N.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE