With the benefit of image guidance, radiologists are well suited to place central venous catheters. This article reviews the techniques of catheter insertion in the interventional radiology suite and describes common pitfalls.
Radiologists continue to expand their role in the initiation and
maintenanceof central venous access.1 At the University of Chicago,
we began insertingcentral venous catheters (CVC) in 1995, and have
placed over 1000 devices inthe past two years. With the benefit of
image guidance, radiologists are wellsuited to place CVCs.
Ultrasound-guided veni- puncture is associated with lowrates of
puncture-related complications such as pneumothorax,2,3 and
cathetersare rarely malpositioned because they are inserted with
the aid of fluoroscopicvisualization. With strict adherence to
sterile technique, infection rates forCVC insertion are similar to
those reported in surgical series..2-4
A tunneled CVC courses through the subcutaneous tissues on the
chest wallprior to entry in the subclavian, external jugular, or
internal jugular vein(figure l). The catheter may be used for
hemodialysis, plasmapheresis, orlong-term central venous access. A
cuff attached to the outer surface of thecatheter is positioned
within the subcutaneous tunnel and helps to hold thecatheter in
place after tissue healing. This article reviews the techniques
ofcatheter insertion in the interventional radiology suite and
The procedure and possible complications are discussed with the
patient, andinformed consent is obtained. Risks include, but are
not limited to,hemorrhage, pneumothorax, and infection. Infection
is the most commoncomplication. Sterility is essential during
catheter placement. To minimize therisks of infection, radiologists
are fully masked and gowned for the entireprocedure and should
observe standard surgical scrub protocol.
The patient's platelet count should be 50,000/mm3 or greater and
theprothrombin time international normalization ratio should
be=1.3. Although nodata exists to support pre-procedural
antibiotics, it is the standard ofpractice at many institutions
including our own. We administer l gm ofintravenous ceftizoxime
approximately one hour before the procedure, unless thepatient is
already taking antibiotics. Patients in other institutions
commonlyreceive a first generation cephalosporin such as 1 gm of
cephazolin sodium,administered intravenously. Although we
occasionally administer a small amountof intravenous contrast, we
do not routinely check serum creatinine levels. Atour institution,
most tunneled catheters are placed for hemodialysis andpatients
undergo dialysis immediately after the procedure. Conscious
sedationwith fentanyl citrate and midazolam hydrochloride may be
provided perradiologist preference.
The patient is placed in the supine position and instructed to
turn his orher head away from the physician (i.e., if the right
internal jugular vein isused as access, the patient should look to
the left). Patients may wear a capand mask per radiologist's
preference. For obese patients, it is helpful to usea pillow to
slightly extend the neck. The area from the angle of the mandibleto
the nipple line on the side of the puncture is scrubbed with a
standardsurgical skin preparation kit that contains iodophor
detergent soap andpovidone-iodine. The skin is first scrubbed with
detergent for several minutesthen cleansed with povidone-iodine
three times consecutively. Liberal use ofcleansing solution is
suggested. After the application of betadine, towels aredraped
around the puncture site and chest. It is best to partially cover
thepatient's face as it avoids increased anxiety during the
procedure and improvespatient monitoring. The patient should be
covered from head-to-feet, and allequipment near the patient,
including the image intensifier, should be properlycovered with
Inadequate draping-It is very important to cover as much of the
patient andsurrounding area as possible because the wires and
catheters commonly come incontact with anything near the field.
Patient with history of previous CVCs-Since indwelling catheters
may incitethrombosis or occlusion of the central veins, a cursory
ultrasound examinationof the neck is performed to ensure vein
patency before prepping and draping thepatient.
Women with a abundant subcutaneous tissue or large breasts-It is
helpful totape the subcutaneous tissue and breast as caudal as
possible. This minimizesretraction of the central venous line when
the patient assumes an uprightposition.
Venous puncture techniques
If possible, all central venous lines for hemodialysis should be
placed viathe right or left internal jugular vein (IJV). The
subclavian veins should beavoided in patients requiring
hemodialysis because catheter placement in asubclavian vein may
cause stenosis, which will render the ipsilateral
extremityunsuitable for a synthetic graft or arteriovenous
fistula.5 In addition,subclavian vein catheters have a higher
incidence of symptomatic central venousthrombosis than internal
jugular vein catheters.6
The right IJV is the best access site because it is nearly
always orientedin a straight cephalocaudal course toward the heart.
We perform all punctureswith a Micropuncture Introducer Set (Cook,
Bloomington, IN) using ultrasoundguidance. The set contains a
21-gauge needle, .018=guidewire, and co-axial 3Fr. and 5 Fr.
IJV puncture: Freehand real-time needle insertion-The IJV is
easilylocalized adjacent to the carotid artery and thyroid gland.
Although puncturethrough the muscle usually can be performed
without complications, it is bestto puncture the vein between the
medial and lateral heads of thesternocleidomastoid muscle
approximately 1 cm above the clavicle. Puncturingtoo low in the
neck increases the risk of pneumothorax, while puncturing toohigh
may cause the catheter to kink at the vein entry site (figure
A 5 or 7 MHz probe is covered with a sterile drape containing a
small amountof acoustic gel, and betadine is applied to the neck as
a coupling agent. For aright internal jugular vein puncture, the
radiologist should stand at the headof the bed with the probe held
in transverse orientation (figure 3). The probeis positioned so
that the IJV is in the center of the image. After
subcutaneousadministration of 1% lidocaine, the puncture needle is
attached to a 10-ccsyringe and the bevel of the needle is placed on
the skin directly beneath thecenter of the probe (figure 4). The
needle should be oriented slightly mediallyand caudally, and
advanced toward the vein under ultrasound guidance withsimultaneous
The walls of the IJV tend to coapt with constant pressure so
that the needlefrequently passes through both the front and back
walls of the veinsimultaneously. If the bevel is passed deep to the
vein, the needle is slowlywithdrawn with gentle suction on the
syringe until venous blood is aspirated.The .018=guidewire is then
advanced through the needle into the superior venacava (SVC). The
needle is removed, and the 3 Fr. and 5 Fr. dilator combinationis
advanced over the wire. This combination may be left in place while
thesubcutaneous tunnel is formed. Alternatively, a .035=J-tipped
guidewire may bepassed through the 5 Fr. dilator and the dilator
exchanged for an end-holecatheter. The wire-catheter combination
can then be advanced into the inferiorvena cava (IVC).
Subclavian vein puncture-For subclavian punctures, an injection
of iohexolis performed through an intravenous catheter in the wrist
or hand to opacifythe subclavian vein. Using fluoroscopic guidance,
the vein is punctured usingthe micropuncture set at the lateral
aspect of the first rib.7 As the needle isadvanced, the vein will
"tent," and the needle will then"pop" into the vein. It is helpful
to direct the needle mediallytoward the lateral aspect of the first
rib so that if the needle passes throughthe vein, it will stop at
the rib instead of transgressing the pleural spaceand possibly
causing a pneumothorax. It is important not to puncture medial
tothe first rib because this can cause catheter compression by the
subclaviusmuscle or costoclavicular ligament-subclavius tendon
complex, causing"pinch-off" syndrome.8
Not advancing the needle deep enough during IJV puncture-This is
a verycommon and frustrating error resulting from inexperience.
Occasionally, it maybe difficult to visualize the needle tip.
Although the IJV usually is only 1 to2 cm below the skin surface,
the needle frequently needs to be advanced deeperbecause the walls
of the vein coapt with constant pressure. It is important
tovisualize the needle tip at all times to prevent inadvertent
vertebral arteryor carotid artery puncture or pneumothorax.
Withdrawing too forcefully after venipuncture-Only gentle
suction isrequired to aspirate blood from the vein. If too much
suction is used, softtissue may be drawn into the needle,
subsequently occluding the needle's lumen.
Inability to pass wire-The needle may be against the back wall
of the vein.After pulling the wire into the needle, withdraw the
needle 1 to 2 mm and thenadvance the wire. If the wire enters the
IJV or SVC but will not pass distallyinto the more central veins,
exchange the needle for the 3 Fr. dilator includedin the
micropuncture set and inject contrast to evaluate patency of the
Uncertainty of which vessel the needle has entered-This may be a
problemwhen the carotid is immediately posterior to the IJV. If
either the subclavianartery or common carotid artery is
inadvertently punctured, brisk back bleedingusually is evident
through the micropuncture needle. The operator should removethe
needle and compress the area for five minutes. If the operator is
unsure ofwhich vessel the needle has entered, the wire should be
advanced into thevessel and the needle exchanged for a 3 Fr. inner
dilator. A small amount ofcontrast can then be injected using
fluoroscopic guidance to identify thevessel. If an artery has been
punctured, the dilator is removed, and the areais compressed for
Technique of forming a subcutaneous tunnel
Lidocaine is infiltrated along the expected subcutaneous course
of thetunnel. We prefer to use lidocaine without epinephrine,
although lidocaine withepinephrine may decrease bleeding along the
catheter tract. Lidocaine withepinephrine should be avoided in
patients with sickle cell anemia because ofthe risk of skin
necrosis. After lidocaine infiltration, a subcutaneous tractis
fashioned with a blunt-tipped metal or plastic trocar. The tunnel
should beat least 8 to 10 cm in length. We usually enter the skin
on the anterior chestwall approximately 4 to 5 cm caudal to the
lateral third of the clavicle. Ascalpel (#11 blade) is used to make
a 5 mm incision at the desired entry siteon the chest. In women,
the exit site should be placed in a position that doesnot affect
bra straps. The trocar is advanced to the lateral aspect of
thepuncture site, and a scalpel and hemostat are used to dissect to
theblunt-tipped trocar. The tip of the trocar is then pulled
through the skin(figure 5). The incision site at the neck should be
5 to 10 mm in length.
Previous tunneled catheters--In patients with previous
catheters, there maybe abundant scar tissue on the chest wall which
makes tunneling difficult.Using a metal trocar may be especially
helpful in this situation.
Difficulty passing a trocar--If the skin dimples with passage,
the trocar istoo shallow. Redirect the tip more subdermally.
Inadequate neck incision site--The incision site in the neck
should beapproximately 1 cm in length; if the site is too short, it
will be difficult topass the trocar through the skin, and later it
will be nearly impossible toadvance the dilator-peel-away sheath
combination into the vein. This problem isexacerbated by patients
who have scar tissue in the neck from previouscatheters. If the
incision is not deep enough, a tag of skin will tend to kinkthe
catheter when it runs from the tunnel into the vein (figure 6A). If
it isnot wide enough, it will kink the peel-away sheath when the
dilator is removed,precluding the catheter from advancing into the
vein, or it may crimp thepeel-away sheath on the dilator, making it
difficult to pass into the vein(figure 6B).
Inability to pass the trocar through a puncture site in the
neck--In obesepatients and patients who have scar tissue in the
neck, it may be difficult topass the trocar out of the skin.
Frequently, the trocar is able to be advancedvery close to the IJV
but not out through the tissue adjacent to the puncturesite.
Plastic trocars are especially susceptible to this complication.
Ratherthan push vigorously, it is best to carefully dissect to the
tip of the trocarwith a scalpel and hemostat and then pull it
through the skin with a hemostat.
Catheters that are pre-cut to appropriate lengths are utilized
forhemodialysis or plasmapheresis. In general, a 36-cm catheter is
placed from theright and a 40-cm catheter is placed from the left.
Note that differentmanufacturers measure catheters differently;
some measure the length from thetip to the hub while others measure
the length from the tip to the cuff. Theoperator should be familiar
with the catheter used in his or her institutionand that particular
manufacturer's terminology. After the tunnel is made, thecatheter
is attached to the trocar and pulled through the tunnel. For
pre-cutcatheters, it is best to pull the subcutaneous cuff 5 to 6
cm into the tunnelinitially. It is much easier to pull back a
catheter that is placed too deep inthe right atrium than to advance
a catheter that is not in deep enough.
For catheters which need to be cut, appropriate length
determination can bedone in several ways. The .018=micropuncture
set wire used for initial accessmay be positioned so that its tip
is at the junction of the superior vena cavaand right atrium. It is
then bent at the hub of the dilator, removed, and theportion from
tip to kink is measured. Alternatively, the catheter can be
placedon the chest on its projected course. The correct length is
then determinedwith fluoroscopy. After trimming the catheter (if
necessary), a .035=guidewireis advanced through the 5 Fr. dilator
in the neck into the IVC. As the wire isadvanced from the SVC
through the right atrium into the IVC, an ECG is used tomonitor for
ectopy. The wire is then directed away from the right ventricle
andtricuspid valve. If an IVC filter has been previously placed,
the wire shouldbe positioned cephalad to the filter.
The .035=wire is exchanged for a .035=superstiff wire through an
end-holecatheter (the stiffer guidewire allows better support when
placing thesheath-dilator combination). Under fluoroscopic
guidance, the catheter isexchanged for the dilator-peel-away sheath
combination provided in the kit(figure 7). A slight twist at the
skin with firm, constant pressure will helpthe dilator-sheath
combination pass into the vein. The guidewire is thenremoved, and
the opening of the dilator is covered with a finger. The patientis
instructed to hold his breath (to prevent air embolism), the
dilator is thenremoved, and the sheath is crimped between two
fingers to prevent excess backbleeding. If the angiographic table
tilts, the patient should be placed inslight reverse
Trendelenburg's position. The tip of the catheter is
advancedthrough the peel-away sheath as far as possible into the
SVC, and the sheath ispeeled away while the catheter is held in
position (figure 8). After the sheathis completely peeled apart,
the catheter is pushed under the skin with a fingeror hemostat. The
hub of the catheter is retracted slightly to clear any kinksfrom
the catheter (figure 9).
The catheter tip should be slightly distal to the junction of
the SVC andright atrium. It will migrate cephalad 1 to 2 cm when
the patient stands. Theports of the catheter are flushed with 3 to
5 ml of 1000U/ml heparin. The skinedges at the puncture site are
approximated with a single 3-0 Ethilon suture(Ethicon, Somerville,
NJ) or a subcuticular resorbable suture. The catheter isthen
secured to the skin on the chest with an 0 suture. Antibacterial
ointmentis then placed at the skin incision sites, and the sutures
are removed at 3weeks. Iodophor antibacterial ointment also can be
applied, as it is at ourinstitution, though some epidemiologists
and infectious disease consultants donot feel that the ointment
provides additional benefit. A completion radiographis then
obtained on the angiography table (figure 10). We do not
routinelyperform post procedure chest radiographs unless
venipuncture was difficult.
The catheter won't advance through the sheath-This is not
uncommon when aleft-sided approach is used or when abundant scar
tissue in the neck crimps thepeel-away sheath. Place a stiff-shaft
hydrophilic guidewire into the catheter.In most situations, the
reinforced catheter should advance easily.
The catheter kinked at puncture site (figure 11)-This usually
occurs if theincision site in the neck is too short, too
superficial, or if the skinincision is not connected to the
puncture site. To resolve this, take ahemostat and bluntly dissect
the tag between the catheter and vein.
Persistent bleeding-Sit the patient upright and compress both
the puncturesite and the tract. If bleeding persists after 30
minutes, obtain bloodcoagulation parameters. Persistent bleeding
can be caused by venoushypertension or fluid overload. If too much
heparin was inadvertently givenwhen the catheter was flushed,
consider administering 10 mg of protaminesulfate intravenously.
Avoid giving protamine to diabetics taking neutralprotamine
hagedorn insulin, as protamine insulin use may
immunologicallysensitize patients to protamine, leading to
Inability to pass the wire into the IVC-If this occurs, use a 5
Fr. catheterto direct the wire through the right atrium into the
Presence of a previously placed IVC filter-A guidewire passed
into the IVCcan become entangled with an IVC filter. Use
fluoroscopy to quickly scan theabdomen before placement of the
central venous catheter. The tip of theguidewire should be
positioned in the IVC above the filter.
Presence of a left-sided catheter- Due to the tortuous course of
theleft-sided veins, the guidewire should be located in the lower
IVC beforeattempting to pass the dilator-sheath combination. If the
wire is left in theSVC or right heart, these structures could be
punctured by thelaterally-directed force of the dilator. It is best
not to advance thedilator-sheath combination distal to the junction
of brachiocephalic veins.
Presence of the catheter tip high in the SVC-First, place a
guidewire intothe catheter, and then advance the wire.
With the aid of image guidance, tunneled central venous
catheters can beplaced by radiologists with excellent safety and
success. In the interventionalradiology suite, ultrasound is used
to guide internal jugular vein puncture,and fluoroscopy enables
visualization of both the course and position of thecatheter as it
is being deployed. Image guidance virtually eliminates the riskof
several complications reported with unguided placement, such
aspneumothorax, hematoma due to arterial puncture, and catheter
malposition. Amajor impetus for radiologic placement of central
venous catheters in ourhospital is ease of scheduling. In most
cases, same day or next day placementis possible in the radiology
department, whereas scheduling delays often areencountered for
insertion in our busy operating rooms. In our opinion,conventional
insertion techniques using anatomic landmarks offer no
definiteadvantages when compared with image-guided insertion.
1. Mauro MA, Jaques PF: Insertion of long-term hemodialysis
catheters byinterventional radiologists: The trend continues.
Radiology 198:316-317, 1996.
2. Funaki B, Szymski GX, Hackworth CA, et al: Radiologic
placement ofsubcutaneous implantable chest ports for long-term
central venous access. AJR1997 169:1435-1437, 1997.
3. Trerotola SO, Johnson MS, Harris VJ, et al: Outcome of
tunneledhemodialysis catheters placed via the right internal
jugular vein byinterventional radiologists. Radiology 203:489-495,
4. Lund GB, Trerotola SO, Scheel PF Jr, et al: Outcome of
tunneledhemodialysis catheters placed by radiologists. Radiology
5. Cimochowski GE, Worley E, Rutherford WE, et al: Superiority
of theinternal jugular over the subclavian access for temporary
dialysis. Nephron54:154-161, 1990.
6. Henriques HF III, Karmy-Jones R, Knoll SM, Copes WS:
Avoidingcomplications of long-term venous access. Am J Surg
7. Jacques PF, Campbell WE, Dumbleton S, Mauro MA: The first rib
as afluoroscopic marker for subclavian vein access. J Vasc Interv
8. Krutchen AE, Bjarnason H, Stackhouse DJ, et al: The
mechanisms ofpositional dysfunction of subclavian venous catheters.
Radiology 200:159- 163,1996.
9. Vincent GM, Janovski M, Menlove R: Protamine allergy
reactions duringcardiac catheterization and cardiac surgery: Risk
in patients takingprotamine-insulin preparations. Cath and
Cardiovasc Diag 23:164-168, 1991.