ORIGINAL ARTICLE

Incidence of intravenous contrast extravasation: increased riskfor patients with deep brachial catheter placementfrom the emergency department

Andrew D. Hardie & Borko Kereshi

Received: 5 November 2013 /Accepted: 12 December 2013# Am Soc Emergency Radiol 2014

Abstract Deep brachial intravenous catheter (IV) placementcan be performed in emergency department patients withdifficult vascular access, but the safety of deep brachial IVfor iodinated contrast administration has not been assessed.This study compares the relative risk for extravasation of deepbrachial IV compared with antecubital IV during powerinjected computed tomography (CT) examinations. A depart-mental practice quality improvement was performed to assessthe rate of IV extravasation for all CT examinations during a1 year period. De-identified data was analyzed with a waiverof informed consent to identify the rate and relative risk ofiodinated contrast extravasation by catheter type. A total of10,750 injections were performed, with 82 extravasationevents (0.8 %). There were 51 extravasations of antecubitalIV from approximately 8,599 placed (0.6 %). For 123 deepbrachial IV placed, there were eight extravasations (6.5 %).The relative risk of a deep brachial IV extravasation was 9.4compared to 0.4 for antecubital placement. Deep brachial IVdemonstrated a markedly higher rate of contrast extravasationthan antecubital IV. For power injected iodinated contrastadministration, it is recommended to avoid the use of deepbrachial IV whenever possible.

Keywords Intravenous . Contrast . Extravasation . CT

Introduction

Computed tomography (CT) utilization in the emergencydepartment setting has been steadily increasing [1] and

frequently requires an intravenous catheter (IV) for iodinatedcontrast administration. The incidence of iodinated contrastIVextravasation during CT imaging has been reported rangingfrom 0.6 to 1.2 % [2–5]. Although serious adverse eventsassociated with IV extravasation are uncommon, there is thepotential for significant morbidity or even loss of limb [5, 6].In order to minimize the rate of IV extravasation, it is impor-tant to identify risk factors. In the emergency department (ED)setting, patients often require intravenous access for numerousreasons including for fluid administration, drug delivery,and also radiologic imaging. For these purposes, deepbrachial IV placement under ultrasound guidance has beenpublished as a reasonable option for intravenous access [7].However, iodinated contrast administration for CT imagingis often more rapid than other fluids and, under highpressure, potentially increasing the rate of extravasation.Further, the morbidity of iodinated contrast extravasationis potentially higher than that of many other administeredagents [6]. As such, the type and location of placement ofthe IV for CT imaging may be essential in preventingextravasation events.

The aim of this study is to compare the incidence ofIV extravasation among IV type, the location of place-ment on the patient, and the location within the hospitalof IV placement. The knowledge gained may be able toalter practice and potentially reduce the rate of IV extravasa-tion and potential morbidity.

Methods and materials

This project was originally performed as a prospective depart-mental practice quality improvement (PQI) project and wasnot required to have institutional review board (IRB) reviewor consent, although it was compliant with the Health Infor-mation Portability and Privacy Act. Once the PQI project was

A. D. Hardie (*) : B. KereshiDepartment of Radiology and Radiological Science, MedicalUniversity of South Carolina, 169 Ashley Avenue, Charleston,SC 29425, USAe-mail: andrewdhardie@gmail.com

Emerg RadiolDOI 10.1007/s10140-013-1185-x

completed, IRB approval was obtained for review and publi-cation of the de-identified data which is presented in thismanuscript. The original project collected data on all intrave-nous contrast administrations for CTexaminations during twoseparate 6-month periods, February 1, 2011 to July 31, 2011andMarch 5, 2012 to September 1, 2012. The total number ofintravenous contrast administrations during these dates as wellas all intravenous contrast extravasation events was recorded.For all extravasation events, the location in the medical facilitywhere the IV was placed (emergency department (ED), inpa-tient, and outpatient) as well as the location of the IV place-ment on the patient's body was recorded.

During the study period, the type and location of all cath-eters which did not have an extravasation event were not ableto be reliably recorded prospectively. However, a significantamount of this data was able to be reconstructed retrospec-tively. First, data for the number of deep brachial IV placed inthe emergency department during the same time periods wasavailable based on de-identified billing codes. Although notall of the patients with a deep brachial IV underwent acontrasted CT scan, the ED coordinator attested that thesignificant majority of these did. As such, the totalnumber of deep brachial IV placed during the studyperiod was used in the study, although the actual num-ber may have been slightly lower. The total number ofantecubital IV placed had to be estimated based on thegeneral experience of the CT technologists. Sinceantecubital is the preferred site of IV in our department,80 % of IV were assumed to have been antecubital. Forother IV placement sites, there were likely a very smallnumber of IV placed at each of the individual othersites (which included forearm, femoral, hand, wrist,etc.). Therefore, the remaining total number of IVplacements was designated as at sites other thanantecubital or deep brachial for this study. Separately,the site of IV placement in the hospital (ED, inpatient,and outpatient) was similarly estimated from de-identified billing codes during the study period. Assuch, an estimated number of IV used during powerinjected CT at each hospital setting was used for thecalculation of data in this study.

Intravenous catheters

All IV used during the study period were of a single type(BD Insyte Autoguard, Becton Dickinson and Company,Utah, USA). The IV sizes ranged from 18 to 24 gauge(g). In the cases of IV extravasation, the number ofextravasation events for each catheter size was as fol-lows: 18 g (25), 20 g (41), 22 g (14), and 24 g (2). Dataon the catheter sizes for all CT examinations was notavailable.

Statistical analysis

All statistical data was recorded in Microsoft Excel 2010(Microsoft, USA). Using basic algebraic calculations, therelative risk of IV extravasation for each catheter type wasassessed.

Results

A total of 10,750 intravenous iodinated contrast administra-tions were performed during the study period, and there were82 extravasation events. The total IV extravasation rate was0.76 %. The mean age of patients was 51.3 years (range 1–95 years) with 39 females. Please see Table 1.

Based on the location within the hospital where thepatient had come to the radiology department, therewere 36 extravasation events from patients from theED, 33 from the inpatient setting, and 13 from theoutpatient setting. The staff which placed the IV wasalso recorded, with all of the ED patients and inpatientshaving had the IV placed within the originating locationprior to transport to the radiology department. Amongoutpatients, 10 of the 13 had the IV placed by radiologynursing staff, while three had a previously existing IVused for the CT examination. The estimated total num-ber of CT examinations for each hospital location dur-ing the study period was 1,290 patients from the ED,2,473 inpatients, and 6,987 outpatients. Hence, the rel-ative risk of IV extravasation for a patient from the EDwas 5.7, while it was 2.3 for an inpatient and 0.4 for anoutpatient. Specifically, for outpatients, there was insuf-ficient data to separate the IV placed by the radiologynursing staff from other sources, although likely thesignificant majority of IV for outpatients were placedby the radiology nursing staff.

Based on the location of the IV placement on the patient,there were 51 extravasations from IV placed at the antecubitalfossa. Eight (8) were from deep brachial IV, which had allbeen placed under ultrasound guidance by ED personnel.Additionally, there were 19 IV extravasations from all otherrecorded sites, which were neither antecubital nor deep brachi-al. In four (4) cases, the site of the IV had not been recorded;hence, these cases were removed from the IV placement siteanalysis. The total number of deep brachial IV placed duringthe study period was 123, while the total number ofantecubital IV was not known exactly but was estimated forthe study period at 8,600. The remaining 2,027 IV wereassumed to represent sites other than antecubital or deepbrachial. Therefore, the rate of extravasation for antecubitalIV, deep brachial IV, and all other IV sites was 0.6, 6.5, and0.9 percent, respectively. The relative risk of extravasation

Emerg Radiol

for antecubital IV, deep brachial IV, and all other IV siteswas 0.4, 9.4, and 1.7, respectively.

Discussion

This study clearly demonstrates the potential for a markedlyincreased rate of IVextravasation with the use of deep brachialIV for power injected CT over antecubital IV. Despite the factthat the deep brachial IV were placed under ultrasound guid-ance by experienced personnel, extravasation events still oc-curred at a high rate. This finding is not unexpected given thereported literature on the potential rate of extravasation withdeep brachial IV placed under ultrasound, which report ratesof at least 5 % even when only used for saline fluid adminis-tration [7]. Given the high pressure and rapid bolus techniqueusually for CT contrast administration, it is not unexpected tofind an even higher rate of extravasation than other reports.However, the clinical significance of a higher rate of

extravasation is likely even greater, given the potential morbid-ity of iodinated contrast extravasation over saline-based fluid.

Although numerous factors may play a role in IV extrava-sation, the findings of this study are unlikely to be related tocatheter type, as only a single catheter type and brand wasused for all examinations. The risk related to catheter size,however, could not be directly evaluated, as only the data forthe extravasation events was available and not the total num-ber of each size catheter placed. Specifically in regard to deepbrachial IV, all personnel placing the catheters had been pre-viously trained in the procedure. However, specific informa-tion on which personnel and their level of training was notavailable. Also, information regarding the patient includingage, weight, and other patient specific data was not known.

Further limitations of the study include the nature of thedata collection, as we were forced to estimate the total numberof antecubital IV placed. Although this number may be notperfectly accurate, it is likely reasonably close to the actualnumber and is overall similar to the rates of prior reports [3, 4].However, the data on the total number of deep brachial placedwas available. In fact, the rate of deep brachial IV extravasa-tion may have been underestimated, as likely not all deepbrachial IV were used for CT. Overall, there was a higher rateof extravasation from IV placed in the ED than that in otherhospital locations. Patients with IV placed in the ED may bemore often acutely ill, dehydrated, or uncooperative, leadingto a higher rate of extravasation. However, this is also likelytrue for inpatients, which had nearly half the rate of IVextravasation. The difference in the rate of extravasation washeavily influenced by the extravasation events for the deepbrachial IV placed in the ED. Following the study, an attemptwas made to ascertain why deep brachial IV is utilized in theED at our institution. While deep brachial placement is chosenfor some patients based on the inability to obtain an IV atanother site, the procedure was sometimes chosen based onperceived speed and ease of placement. As such, it cannot beassumed that deep brachial IV was restricted to patients with-out other available IV access. Regardless of the potentialdifferences in the patient populations, the high overall riskfor extravasation with deep brachial IV suggests that alterna-tive forms of IV access including central line placementshould be considered. Further, this study highlights the ex-tremely low rate of IV extravasation for IV placed by radiol-ogy nursing personnel.

In summary, the rate of IVextravasation during a contrast-enhanced CT exam was found to be significantly lower withantecubital placement than that with deep brachial IV place-ment. As such, for use with CT, deep brachial access shouldremain a last resort option for IV access.

Conflict of Interest The authors declare that they have no conflict ofinterest.

Table 1 Data recorded for all extravasation events including relativepercentages

Total Percent

Extravasations 82 0.76

Location

Emergency (ED) 36 44

Inpatient 33 40

Outpatient 13 16

Total 82 100

Catheter size (g)

18 25 30

20 41 50

22 14 17

24 2 2

Total 82 100

Catheter location

Antecubital Fossa 51 62

Brachial 8 10

Forearm 6 7

Hand 5 6

Other not specified 4 5

Upper arm (biceps, humerus) 4 5

Wrist 3 4

Femoral 1 1

Total 82 100

Catheter placed by

Radiology 10 12

Other service 72 88

Total 82 100

Emerg Radiol

References

1. Berdahl CT, VermeulenMJ, Larson DB, Schull MJ (2013) Emergencydepartment computed tomography utilization in the United States andCanada. Ann Emerg Med. May 14

2. Federle MP, Chang PJ, Confer S, Ozgun B (1998)Frequency and effects of extravasation of ionic and nonionicCT contrast media during rapid bolus injection. Radiology206(3):637–640

3. Jacobs JE, Birnbaum BA, Langlotz CP (1998) Contrast media reac-tions and extravasation: relationship to intravenous injection rates.Radiology 209(2):411–416

4. Wienbeck S, Fischbach R, Kloska SP (2010) Prospective study ofaccess site complications of automated contrast injection with periph-eral venous access in MDCT. AJR Am J Roentgenol 195(4):825–829

5. Wang CL, Cohan RH, Ellis JH, Adusumilli S, Dunnick NR(2007) Frequency, management, and outcome of extravasationof nonionic iodinated contrast medium in 69 657 intravenousinjections. Radiology 243(1):80–87

6. Namasivayam S, Kalra MK, Torres WE, Small WC (2006) Adversereactions to intravenous iodinated contrast media: a primer for radiol-ogists. Emerg Radiol 12:210–215

7. Keyes LE, Frazee BW, Snoey ER, Simon BC, Christy D (1999)Ultrasound-guided brachial and basilic vein cannulation in emergencydepartment patients with difficult intravenous access. Ann EmergMed34(6):711–714

Emerg Radiol