antibiotic guide for ir

7/31/2019 Antibiotic Guide for Ir

1/20

Practice Guideline for Adult AntibioticProphylaxis during Vascularand Interventional Radiology ProceduresAradhana M. Venkatesan, MD, Sanjoy Kundu, MD, David Sacks, MD, Michael J. Wallace, MD,

Joan C. Wojak, MD, Steven C. Rose, MD, Timothy W.I. Clark, MD, MSc, B. Janne dOthee, MD, MPH,Maxim Itkin, MD, Robert S. Jones, DO, MSc, FACP, Donald L. Miller, MD, Charles A. Owens, MD,Dheeraj K. Rajan, MD, LeAnn S. Stokes, MD, Timothy L. Swan, MD, Richard B. Towbin, MD, and

John F. Cardella, MD

J Vasc Interv Radiol 2010; 21:16111630Abbreviations: AHA American Heart Association, GI gastrointestinal, GU genitourinary, IE infective endocarditis, IR interventional radiology,IV intravenous, IVC inferior vena cava, RF radiofrequency, SIR Society of Interventional Radiology, TIPS transjugular intrahepatic portosystemicshunt, UAE uterine artery embolization

PREAMBLE

There is a need for current, formal rec-ommendations in the interventional ra-diology (IR) literature concerning theappropriate use of prophylactic antibi-otics for IR procedures. This is particu-larly important given the increasing in-

cidence of antibiotic resistance and

complications from nosocomial infec-tion. This document summarizes thefindings from the available surgical andIR literature on this topic. Anticipatedpathogens and corresponding antibioticcoverage (dose and duration) are enu-merated for common vascular and non-vascular interventions in adults. Note

that this document is intended to pro-

vide recommendations concerning onlyantibiotic prophylaxis, not treatment ofinfectious complications. It is beyondthe scope of this article to discuss themanagement of infectious complica-tions sustained during vascular and in-terventional radiology procedures.

The membership of the Society of In-

terventional Radiology (SIR) Standardsof Practice Committee represents ex-perts in a broad spectrum of interven-tional procedures from the private andacademic sectors of medicine. Gener-ally, Standards of Practice Committeemembers dedicate the vast majority oftheir professional time to performing in-terventional procedures; as such, theyrepresent a valid broad expert constitu-ency of the subject matter under consid-eration for standards production.

Technical documents specifyingthe exact consensus and literature re-

view methodologies as well as theinstitutional affiliations and profes-sional credentials of the authors ofthis document are available upon re-quest from SIR, 3975 Fair Ridge Dr.,Suite 400 N., Fairfax, VA 22033.

METHODOLOGY

SIR produces its Standards ofPractice documents using the follow-ing process. Standards documents ofrelevance and timeliness are concep-

From the Center for Interventional Oncology, Radi-ology and Imaging Sciences (A.M.V.), National In-stitutes of Health Clinical Center; Department ofRadiology and Radiologic Sciences (D.L.M.), Uni-formed Services University of the Health Sciences;Department of Radiology (D.L.M.), National NavalMedical Center, Bethesda; Department of Radiol-ogy, Division of Interventional Radiology (B.J.D.),University of Maryland Medical Center, Baltimore,Maryland; Department of Medical Imaging (S.K.),Scarborough General Hospital; Division of Vascular

and Interventional Radiology, Department of Med-ical Imaging (D.K.R.), University of Toronto, Uni-versity Health Network, Toronto, Ontario, Canada;Department of Interventional Radiology (D.S.) andSection of Infectious Diseases (R.S.J.), Reading Hos-pital and Medical Center, West Reading; Depart-ment of Radiology, Division of Interventional Radi-ology (M.I.), University of Pennsylvania MedicalCenter, Philadelphia; Department of Radiology(J.F.C.), Geisinger Health System, Danville, Pennsyl-vania; Department of Interventional Radiology(M.J.W.), The University of Texas M.D. AndersonCancer Center, Houston, Texas; Department of Ra-diology (J.C.W.), Our Lady of Lourdes Medical Cen-ter, Lafayette, Louisiana; Department of Radiology(S.C.R.), University of California San Diego MedicalCenter, San Diego, California; Department of Inter-

ventional Radiology (T.W.I.C.), New York Univer-sity School of Medicine, NYU Medical Center, NewYork, New York; Departments of Radiology andSurgery (C.A.O.), University of Illinois Medical Cen-ter, Chicago, Illinois; Department of Radiology andRadiological Sciences (L.S.S.), Vanderbilt UniversityMedical Center, Nashville, Tennessee; Departmentof Radiology (T.L.S.), Marshfield Clinic, Marshfield,Wisconsin; and Department of Radiology (R.B.T.),Phoenix Childrens Hospital, Phoenix, Arizona. Re-ceived May 21, 2010; final revision received July 3,

2010; accepted July 23, 2010. Address correspon-dence to A.M.V., c/o Debbie Katsarelis, 3975 FairRidge Dr., Suite 400 N., Fairfax, VA 22033; E-mail:[email protected]

This document was written by the Standards ofPractice Committee for the Society of InterventionalRadiology and endorsed by the Cardiovascular In-terventional Radiological Society of Europe and Ca-nadian Interventional Radiology Association.

None of the authors have identified a conflict ofinterest.

SIR, 2010

DOI: 10.1016/j.jvir.2010.07.018

Standards of Practice

1611
mailto:[email protected]:[email protected]


2/20

tualized by the Standards of PracticeCommittee members. A recognizedexpert is identified to serve as theprincipal author for the standard.Additional authors may be assigned

dependent upon the magnitude ofthe project.

An in-depth literature search isperformed using electronic medicalliterature databases. Then a criticalreview of peer-reviewed articles isperformed with regard to the studymethodology, results, and conclu-sions. The qualitative weight of thesearticles is assembled into an evi-dence table, which is used to writethe document such that it containsevidence-based data with respect tocontent, rates, and thresholds.

When the evidence of literature isweak, conflicting, or contradictory,co nse nsus fo r the paramete r isreached by a minimum of 12 Stan-dards of Practice Committee mem-

bers using a Modified Delphi Con-sensus Method (Appendix A) (13).For the purposes of these documents,consensus is defined as 80% Delphiparticipant agreement on a value orparameter.

The draft document is criticallyreviewed by the Standards of Prac-

tice Committee members, either bytelephone conference calling or face-to-face meeting. The finalized draftfrom the Committee is sent to the SIRmembership for further input/criti-cism during a 30-day comment pe-riod. These comments are discussed

by the Standards of Practice Com-mittee, and appropriate revisionsmade to create the finished stan-dards document. Before its publica-tion the document is endorsed by theSIR Executive Council.

The current guidelines were writ-ten to help determine appropriateantibiotic prophylaxis for vascularand IR procedures. The elements ofcare necessitate knowledge of thefollowing: (i) the anticipated patho-gens associated with the given inter-vention and (ii) pretreatment evalu-ation and patient-specific factors thatmay be associated with higher likeli-hood of periprocedural infection.The outcome measures or indicatorsfor this process include (i) how oftenantibiotics are given when they

should be given; (ii) how often anti-

biotics are given when they shouldnot be given; and (iii) how often isantibiotic t iming within re co m-mended guidelines. Although prac-ticing physicians should strive to

achieve perfect outcomes, in practiceall physicians will fall short of idealoutcomes to a variable extent. There-fore, in addition to quality improve-ment case reviews conducted afterindividual complications, outcomemeasure thresholds should be usedto assess treatment safety and effi-cacy in ongoing quality improve-ment programs. For the purpose ofthese guidelines, a threshold is a spe-cific level of an indicator that, whenreached or exceeded, should prompta review of departmental policiesand procedures to determine causesand to implement changes, if neces-sary. Thresholds may vary fromthose listed here; for example, pa-tient referral patterns and selectionf ac to rs m ay d ic ta te a d if fe re ntthreshold value for a particular indi-cator at a particular institution.Therefore, setting universal thresh-olds is difficult, and each departmentis urged to adjust its thresholds asneeded to meet its specific qualityimprovement program situation.

SIR is committed to the basic prin-ciples of outcome-focused, evidence-

based medicine. Ideally, every Stan-d ar ds o f P ra ct ic e C om mi tt eerecommendation would be based onevidence derived from multiple pro-spective randomized trials of ade-quate statistical power. Unfortu-n at el y, c ur r en tl y t h er e a r e n opublished multicenter randomizedtrials that evaluate the clinical effi-cacy and indications for antibioticprophylaxis during IR procedures.Within the existing publications, sev-eral major limitations are evident: (i)lack of randomized controlled trialsconcerning the efficacy of IR antibi-otic prophylaxis and reliance on ex-isting case series and retrospectiveanalyses; (ii) extreme variation in theexisting literature concerning patientselection parameters, definitions ofshort-term efficacy, and definitionsof infectious complications; and(iii) absence of systematic assess-ment of long-term efficacy. SIR rec-ognizes the potential pitfalls of de-

veloping evidence-based standards

for antibiotic prophylaxis and ofmaking recommendations regardingthe use of drugs based on studies ofsuboptimal design. However, thesedifficulties are far outweighed by the

potential improvements in safetyand treatment efficacy that may begained by implementing the key les-sons learned from the existing peer-reviewed scientific literature that hasevaluated antibiotic prophylaxisduring surgical and IR procedures.Given the limited scientific founda-tion, most of the recommendationspresented in this document are in-tended to guide clinical practicerather than to mandate the use ofspecific algorithms. The authors

fully anticipate that these guidelineswill be appropriately revised whenfuture studies of greater scientificrigor are available. Levels of evi-dence have been assigned to the cur-rent recommendations within thisclinical practice guideline that ad-here to definitions created by theAmerican College of Cardiology andAmerican Heart Association (AHA)Guidelines Task Force (4,5). A sum-mary of these definitions are pro-vided in Appendix B.

INTRODUCTION

During the past two decades, sev-eral authors have reported on the useof prophylactic antibiotics for IR pro-cedures (6 9). Spies et al (6) were thefirst to provide a critical review onantibiotic prophylaxis in IR. Theirstudy underscored the lack of ran-domized controlled trials of antibi-otic prophylaxis during IR proce-dures, indicating that the selection ofantibiotics for IR prophylaxis waslargely guided by the existing surgi-cal literature (6,9). There has been alack of randomized controlled trialsin the radiology literature to providescientific validation for the role andeffectiveness of antibiotic prophy-laxis during IR procedures. Never-theless, antibiotic prophylaxis has

become the perceived standard ofcare for selected procedures, makingit possible that randomized con-trolled trials may never be per-formed (6,9). Extensive clinical data

on the use of antimicrobial agents in

1612 Guidelines for Adult Antibiotic Prophylaxis during IR Procedures November 2010 JVIR


3/20

surgery exist. As noted by Spies et al(6), the analogous clinical consider-ations in surgery serve as a basis forde ve loping an appro ach to theproper use of antibiotics in IR. Clin-

ical case series and retrospectiveanalyses in the IR literature concern-ing the occurrence, prevention, andmanagement of infectious complica-tions following specific IR proce-dures are also available to guide clin-ical practice (1014). This documentoffers a logical basis for prophylacticuse of antibiotics during IR proce-dures based on the literature avail-able. Many current clinical and envi-ronmental factors mandate currentrecommendations for prophylaxisduring IR procedures. These include

the widening breadth of IR proce-dures and recent revisions to existingsurgical practice guidelines fromwhich many IR prophylaxis guide-lines are derived (6,9,1517). Addi-tional factors include increasing an-t ib i ot i c r es i st a nc e, m or b id i tyassociated with nosocomial infec-tion, and increasing health care costs(9,1517).

The specific antibiotic agent dosesprovided herein are recommenda-tions for normal adult patients. Al-though the doses and agents named

here are suggestions meant to facili-tate the care of IR patients, it is theultimate responsibility of the indi-vidual interventional radiologist topay close attention to individual pa-tient factors, including advanced pa-tient age, existing medications, re-duced renal or hepatic function,medication allergy history, the in-tended procedure and its l ikelypathogens, and the dose and timingof the most recent antibiotic agentdose. Antibiotic choices have beenlisted as first choice if a specific

antibiotic agent has been consistentlydescribed in the published literaturein association with prophylaxis for agiven procedure. When no singleagent has been uniformly reported asthe agent of choice for prophylaxis ofa given procedure, no first-choice an-tibiotic is listed. Instead, a list ofcommon antibiotic choices are pro-vided, which are described in the lit-erature. It is important to note thatthis document should not be consid-ered a replacement for the close andcareful evaluation of individual pa-

tients by their IR practitioners. The

responsibility ultimately lies withthe individual IR physician, whomust determine for him- or herselfthe appropriateness of a specific an-tibiotic regimen for a specific patient

and clinical scenario. As the antibi-otic sensitivities of different patho-gens vary based on time and region,the appropriate prophylactic antibi-otic selection may differ dependingon the institution at which the phy-sician practices. The IR physicianshould therefore periodically consultwith his or her local hospital-basedinfectious disease committee, con-sulting service, microbiology labora-tory, and/or pharmacy for recom-mended medications.

DEFINITIONS

General Definitions

This document uses definitions asoutlined by Spies et al (6) for coloni-zation, infection, clinical infection,and prophylaxis, as follows:

Colonization is the presence of amicroorganism without hostresponse.

Infection is the presence of an or-ganism with host response.

Clinical infection represents infec-tions that produce local signsor symptoms (eg, fever, pain,leukocytosis).

Prophylaxis is the administration ofan antimicrobial in the periproce-dural period in the absence ofclinical infection to prevent an in-fectious complication.

Bacteremia is the presence of bac-teria within the bloodstreamwithout clinical signs or symp-toms of infection (eg, fever,pain, leukocytosis).

Septicemia i s t h e p re se nc e o f pathogenic organisms or theirtoxins in the bloodstream withconcomitant systemic signs andsymptoms of sepsis (6).

Procedure Classification

The National Academy of Sciences/National Research Council has dividedsurgical wounds into four classesclean, clean-contaminated, contami-nated, and dirtyeach of which is as-sociated with a different risk of infection

(7,18). This classification may be used to

guide antibiotic prophylaxis for IR pro-cedures (6).

Clean.A procedure is regarded asclean if the gastrointestinal (GI) tract,genitourinary (GU) tract, or respiratory

tract is not entered, if inflammation isnot evident, and if there is no break inaseptic technique (18). An example of aclean procedure is routine diagnosticangiography (7,18).

Clean-contaminated.A procedureis regarded as clean-contaminated ifthe GI, biliary, or GU tract is entered;if inflammation is not evident; andthere is no break in aseptic tech-nique. An example of a clean-con-taminated procedure is nephrostomytube placement in a patient with ster-ile urine (7,18).

Contaminated .A procedure is re-garded as contaminated if there is entryinto an inflamed or colonized GI or GUtract without frank pus, orif a major breakin aseptic technique occurs (18).

Dirty.A procedure is regardedas dirty if it involves entering an in-fected purulent site such as an ab-scess, a clinically infected biliary orGU site, or perforated viscus (18).

As noted later, antibiotic prophylaxisfor IR procedures is generally recom-mended for procedures that are notclean, or for procedures that are consid-

ered clean but, as a result of which, apotentially significant volume of ne-crotic tissue is generated in potentiallycontaminated areas, eg, embolizationwith intent to create infarction. Al-though National Academy of Sciences/National Research Council surgicalwound classification may be used as aguide, it should be noted that the infec-tious risks of open surgical proceduresdiffer from those of IR procedures (19).As noted by Zarrinpar et al (19), surgicalprophylaxis is typically directed at pre-venting infection of the wound by in-

fected fluid or skin organisms. Thesmall incision made during vascularand nonvascular IR procedures rarelyserves as a site for clinically importantinfection (19). Prophylaxis in IR is morecommonly done to prevent infection re-sulting from the communication that iscreated by a needle or catheter (19) be-tween an infected space and the blood-stream. Data from studies of surgicalwound prophylaxis may therefore notalways be directly applicable to IR pro-cedures performed on the same organ,and the interventional radiologist is en-

couraged to consider the appropriate-

Venkatesan et al 1613Volume 21 Number 11


4/20

ness of prophylaxis in accordance withthe clinical circumstances of the individ-ual patient and site-specific flora (19).

COMPLICATIONS

Complications can be grouped onthe basis of outcome. Major compli-cations from antibiotic therapy resultin admission to a hospital for ther-apy (for outpatient procedures), anunplanned increase in the level ofcare, prolonged hospitalization, per-manent adverse sequelae, or death.Minor complications result in no se-quelae; they may require nominaltherapy or a short hospital stay forobservation (generally overnight;Appendix C). The complication rates

described in the present documentrefer to major complications.

TIMING ANDADMINISTRATION OFPROPHYLAXIS

The basis for timing and duration ofantibiotic prophylaxis of IR proceduresis derived from the surgical literature.The rationale for current surgical antibi-otic prophylaxis emerged from animalexperiments performed in the 1960s byBurke (20), who demonstrated the great-

est suppression of infection when anti-microbial agents were administered be-fore inoculation. The likelihood ofpostprocedure infection increased thelonger the delay between initiation ofantibiotic prophylaxis and the proce-dure; an initial dose given 3 hours aftera procedure resulted in no difference inthe likelihood of infection in treated an-imals versus untreated controls (20).More recent clinical studies corroboratethese findings, demonstrating a fivefoldincreased rate of infectious events whenantibiotic administration occurs more

than 3 hours after a procedure (7,21). Ina randomized double-blinded study ofpatients undergoing colonic surgery, a16% frequency of wound infection oc-curred when no prophylaxis was used;the incidences were 15% when prophy-laxis was given after the procedure andonly 6% when prophylaxis was given 1hour before surgery (6,22). Overall, analmost fourfold greater incidence of in-cisional infection was demonstratedwhen antibiotic agent was withheld ver-sus when it was given preoperatively (18).Initiation of antibiotic therapy postopera-

tively yielded similar results to those in

which parenteral antibiotic treatment wascompletely withheld (6,22).

With respect to duration of ad-ministration, several studies haveevaluated the efficacy of single ver-

sus multiple dose protocols for pro-phylaxis, demonstrating that a singleappropriately timed dose is as effec-tive as a multiple-dose protocol(9,2326). Recommended IR practicefor most procedures warranting pro-phylaxis involves administration of asingle preprocedural dose of an an-timicrobial agent just before com-mencement of a procedure. Excep-tions include procedures when apatient undergoes instrumentationof an obstructed viscus, such as bili-ary or kidney obstruction, in which

the risk of postprocedural bactere-mia caused by intravasation of or-ganisms into the bloodstream re-mains present until the organ isadequately drained (9). The antibi-otic agent used in this setting admit-tedly spans the boundary betweenprophylaxis and treatment, and insuch patients, treatment should becontinued until satisfactory drainageof the viscus is achieved (9). In per-cutaneous drainage procedures, therisk of bacteremia is attributed tocommunication of the fluid cavity

with the arterial, venous, and lym-phatic systems (6,27). In the absenceof data on this topic, Spies et al ( 6)have advocated continued antibiotictherapy for 48 hours after such a pro-cedure to reduce the risk of bactere-mia and its consequences. Othercases that may warrant prolongedprophylaxis include those that in-volve earlier surgical manipulation(eg, bilioenteric anastomosis beforechemoembolization).

Additional factors to consider in-clude patient condition and proce-

dure duration (9). Investigators fromthe Centers for Disease Control (7,28)have described three risk factors as-sociated with an increased risk ofpostoperative surgical wound infec-tio n, in additio n to the level o f wound contamination. These factorswere derived from information col-lected on more than 58,000 operativepatients. A simple multivariate riskindex was created, which prospec-tively predicted surgical wound in-fection risk (28). Factors including aprocedure lasting more than 2 hours,

an abdominal operation, and the

presence of multiple-organ disease,when included in a predictive modelalong with the level of wound con-tamination, allowed identification ofa subgroup of patients who would

sustain 90% of wound infections (28).Prophylactic antibiotic agents shouldbe administered immediately beforetransfer to the interventional suite, or bythe IR nurse when the patient arrives inthe interventional suite. The latter isoptimal, particularly if there is a delay inthe patients arrival, as the time of anti-

biotic agent administration should notbe altered (9). In keeping with the cur-rent Joint Commission on the Accredi-tation of Healthcare Organization 2009Hospital Patient Safety Goals (29),which recapitulate the existing surgical

literature, intravenous (IV) antimicro-bial prophylaxis should be administeredwithin 1 hour before incision, with 2hours allowed for the administration ofvancomycin and fluoroquinolones, andwith discontinuation of the prophylacticantimicrobial agent within 24 hours af-ter intervention, unless clinical circum-stances (eg, obstructed viscus) warrantcontinuation of a prophylactic regimenas therapy. An accurate list of the pa-tients current medications and knownallergies should also be reviewed tosafely prescribe any setting-specific an-

tibiotics and to assess for the potentialallergic or adverse drug reactions (2931). Appropriate antibiotic coverageshould be selected by an experienced IRpractitioner. When a patient is alreadyreceiving antibiotic drugs, the timing ofthe most recent dose of antibiotic agentshould be reviewed, as well as the ap-propriateness of the antibiotic relative tothe procedure and the likely pathogens(9). If indicated, an additional dose ordifferent agent may be administered be-fore the procedure. If a procedure islikely to be prolonged (ie, 2 h), a

supplemental dose of antibioticagent should be considered, depend-ing on the half-life of the agent beingadministered (9).

Antibiotic Resistance

Patient morbidity and mortality re-sulting from infectious complicationshave been increasing during the pasttwo decades (9). This, in large part, isrelated to the increasing frequency andspectrum of antibiotic resistant infec-tions (9,32). Factors facilitating selection

of antibiotic resistance include the over-



5/20

use of antibiotic agents by health carepractitioners, the increased use of inva-sive devices, a greater number of sus-ceptible hosts (ie, those in an immuno-compromised condition), and lapses in

appropriate infection control (9). Whenchoosing an antibiotic agent for prophy-laxis, one needs to consider the likelysource and pathogen, and the prophy-laxis should be specifically directedagainst these organisms. The routineuse of broad spectrum antibiotics is notcost effective and promotes further an-tibiotic resistance (9,33).

PROPHYLAXIS FOR SPECIFICIR PROCEDURES

Vascular Interventions: Angiogra-

phy, Angioplasty, Thrombolysis,Arterial Closure Device Placement,Stent Placement

Diagnostic angiography, routine an-gioplasty, and thrombolysis are consid-ered clean procedures, and with carefultechnique, antibiotic prophylaxis is un-necessary (34). Bacteremia occurs after4%8% of angiographic procedures andis typically asymptomatic (6,34,35).When it does occur, the source is usu-ally an angiographic catheter that wascontaminated before use (6,34). Re-

peated puncture of a site or repeatedcatheterization of an indwelling sheathis believed to increase the risk ofperiprocedural infection (7,36).

Percutaneous vascular closure de-vices, including collagen plug de-vices and suture-mediated closuredevices, are increasingly used duringcardiac and peripheral vascular in-terventions (37,38). Infectious com-plications have been reported afterthe use of these devices; these in-clude groin cellulitis and femoral ar-tery endarteritis (3739). Comorbid

factors described in patients with in-fectious complications include diabe-tes mellitus, obesity, and placementof a percutaneous suture closure de-vice within the previous 6 months,warranting caution when consider-ing the use of a percutaneous closuredevice in such patients (38). How-ever, at the present time, there areinsufficient data to suggest prophy-laxis before placement of these de-vices.

Stent infection is an uncommonbut serious complication of endovas-

cular treatment (9). This is likely sec-

ondary to several factors, includingthe presence of a foreign body andthe presence of inflammation at thesite of stent implantation (7,40). Stentinfection has been reported in the

aorta, iliac, renal, coronary, and sub-clavian arteries (9,4148). When re-ported, complications are major,usually involving arteritis with pseu-doaneurysm formation (9,49). Riskfactors have included repeat punc-ture of the same vessel during ashort time interval or use of a vascu-lar sheath that has been in place formore than 24 hours (9,49). Recom-mendations have been made in theliterature for prophylactic antibioticagent use if an arterial sheath is leftin overnight after stent placement, or

for patients undergoing multiple en-dovascular interventions (46). Oneadvocated approach involves admin-istration of prophylactic antibioticagents before a second procedure topatients in whom repeat interventionis performed within 7 days (38). Al-though infection represents a statis-tically low-risk event, serious out-comes have been associated with thiscomplication. At the present time,prophylactic antibiotic agents are notrecommended for routine arterialstent placement. However, for thosepatients deemed to be at high risk(ie, cases of repeat interventionwithin 7 d, prolonged indwelling ar-terial sheath, or prolonged durationof procedure), prophylaxis may begiven (9,46).

Procedure: cleanOrganisms encountered: Staphylo-

coccus aureus, Staphylococcusepidermidis

Routine prophylaxis recommended: noSpecial considerations: 1 g cefazolin

IV if there is a high risk of stent

infection; or (ii) if the patient isallergic to penicillin, vancomy-cin or clindamycin

Level of evidence: 5, 8 (Appendix B)

Endograft Placement

Prophylactic antibiotic agents aregiven routinely for aortic endografttherapy at many institutions, al-though there is limited scientific ev-idence to support this approach(9,49). Prosthetic graft infection is anuncommon event, but when it oc-

curs, it carries a high mortality rate

(9,50,51). Prophylaxis is similarlyrecommended for peripheral en-dografts, including those for superfi-cial femoral artery recanalizationand dialysis acce ss e ndo grafts,

which are increasingly used in IRpractice.

Procedure classification: cleanOrganisms encountered: S aureus, S

epidermidisRoutine prophylaxis recommended: yesFirst-choice antibiotic agent: 1 g ce-

fazolin IVAlternate choices: (ii) if penicillin-aller-

gic, vancomycin or clindamycinLevel of evidence: 5, 8

Lower-extremity Superficial VenousInsufficiency Treatment

Current therapies for treatment oflower-extremity superficial venous in-sufficiencyincluding varicose veinsinclude endovascular thermal ablation,sclerotherapy, and ambulatory phlebec-tomy (52). Although routine practice in-cludes the use of sterile technique beforethese procedures, there are insufficientdata in the literature to warrant routineantibiotic prophylaxis (52). Adverseevents related to these therapies are typ-ically minor, and include paresthesiasand superficial phlebitis, with more se-rious adverse eventsprimarily per-taining to endovascular thermal abla-tionincluding deep vein thrombosis,neurologic injury, and skin burns (52).


epidermidisRoutine prophylaxis recommended: noLevel of evidence: 8

Inferior Vena Cava Filter Placement

Infection after inferior vena cava

(IVC) filter placement has not been asignificant problem in clinical prac-tice. One case of fatal septicemia af-ter placement of an IVC filter has

been reported in the IR literature (7).In this instance, the IVC filter (LGMVenaTech IVC Filter; L-G Medical,E va ns t on , I ll in oi s) w as p la ce dthrough the site of an indwellingcentral venous catheter (53). Whenplacing an IVC filter, a fresh ve-nous access site is recommended(7,53). However, routine filter place-ment in an uninfected patient does

not warrant prophylaxis (7,54).



6/20

Procedure classification: cleanRoutine prophylaxis recommended: noLevel of evidence: 5, 6

Central Venous Access

The use of antibiotic prophylaxisfor central venous catheter place-ment is controversial. Central ve-nous catheter placement is classifiedas a clean procedure. Therefore, pro-phylaxis is not routinely recom-mended, although meticulous steriletechnique and appropriate preproce-dural preparation is mandatory (9).When catheter-related infections dooccur, these are most commonlycaused by coagulase negative Staph-

ylococcus species (9). Migration ofskin organisms into the catheter tractis the most common route of infec-tion, with contamination of the cath-eter hub contributing to colonizationof long-term catheters (9). Most in-fections in patients with long-termvenous access are caused by Gram-positive bacteria (9).

A recent version of the Centers forDisease Control Guidelines for thePrevention of Intravascular Cathe-ter-related Infections (9,55) under-scored the importance of sterile tech-nique, the use of 2% chlorhexidine

skin preparation, avoidance of rou-tine catheter change, and the use ofantiseptic or antibiotic agentim-pregnated catheters only if previousinfection rates are high. van de We-tering et al (56) evaluated the effi-cacy of administration of antibioticagents before insertion of a centralvenous catheter with or without van-comycin/heparin flush technique inthe first 45 days after catheter place-ment to minimize Gram-positivecatheter-related infections in oncol-ogy patients (9), and antibiotic agent

administration before catheter inser-tion followed by vancomycin/hepa-rin flush was associated with a de-creased incidence of Gram-positiveinfections (56). Of note, this strategywas not advocated for general prac-tice, but in the setting of recurrentcatheter infection (9,5658). Loo et al(59) reported a survey of 196 consec-utive central venous catheters placedin the intensive care unit in 151 pa-tients, in which an antiseptic agentimpregnated catheter was alternatedon a bimonthly basis with a standard

triple-lumen central venous catheter.

There was no difference in the cath-eter-related bacteremia rates be-tween the two groups, and althoughthe impregnated catheter group hada lower cumulative infection rate for

dwell times less than 5 days, the dif-ference between the cumulative in-fection rates was not statistically sig-nificant for dwell times of 6, 7, or 8days (59).

Some authors have suggested thatroutine antibiotic prophylaxis may

be warranted before place ment oftunneled catheters with implantableports, given that these patients areoften in an immunocompromisedstate, and given the severity of com-plications and difficulty in treating

infected ports (9). The benefit of an-tibiotic prophylaxis for central ve-nous access remains unproven, andtherefore the routine use of antibioticprophylaxis remains an area of con-troversy. Prevention is critical tominimize the likelihood of infectionsassociated with implantable ports, asthese are difficult to treat and poten-tially fatal (9). Preventive strategiesinclude meticulous attention to ster-ile technique. The use of catheterswith the fewest number of lumensnecessary for management of the pa-

tient reduces portals for colonization(9). For patients who require inter-mittent long-term venous access, im-plantable devices are recommended, asthese are more resistant to contamina-tion (9). Such devices are now in use fordialysis access management in an at-tempt to reduce infection rates. Al-though the use of antibiotic prophylaxismay be considered in specific clinicalscenarios (eg, immunocompromised pa-tients who require catheter placement

before chemotherapy and those with a

history of catheter infection), routineprophylaxis for these procedures is notrecommended.


epidermidisRoutine prophylaxis recommended: no

consensusSpecial considerations (eg, immuno-

compromised patients who re-quire catheter placement beforechemotherapy and those with ahistory of catheter infection):

1 g cefazolin IV; or in the case

of penicillin allergy, vancomy-cin or clindamycin

Level of evidence: 8

Embolization and

ChemoembolizationHistoric IR data have reported a

higher incidence of transient bactere-mia after embolization comparedwith diagnostic angiography (60). Ina prospective study of 45 patientsundergoing embolization, 32% of pa-tients who received no antibiotictreatment developed bacteremia, al-though none developed clinical sep-sis. The organisms isolated includedS epidermidis, Streptococcus species,and Corynebacterium species, all ofwhich are normal in mucosal or skinflora (60). In contrast, no patientswho received antibiotic prophylaxishad positive blood cultures (60).Findings from this study suggestthat bacteremia is especially likelyfollowing embolization. Findingsfrom additional studies support theuse of antibiotic prophylaxis beforeembolization. Reed et al (61) re-ported the results of 494 hepatic che-moembolization procedures, amongwhich 14 were performed withoutprophylactic antibiotics. One of these

14 cases resulted in fatal sepsiswithin 24 hours of intervention(7,61). In a patient-by-patient analy-sis of these data, of the 226 patientswho received prophylaxis, six devel-oped hepatic abscesses but no casesof fatal sepsis occurred (61). In a sep-arate clinical series of 410 emboliza-tion cases (62), eight deaths were re-ported, of which seven were caused

by infectious complications in pa-tients who had not received antibi-otic prophylaxis. Prophylaxis tar-geted against skin pathogens is

recommended before performing tu-mor and/or solid organ emboliza-tion, including the liver, kidney, andspleen, when there is an intent tocreate infarction or a high likelihoodof infarction, as this may result in apotentially significant volume of ne-crotic tissue in potentially contami-nated areas (9,19,60). Routine pro-phylaxis remains controversial in thesetting of embolization for the pur-poses of controlling bleeding from aviscus or solid organ, such as in thesetting of trauma. Depending on the

clinical setting and treatment goals,



7/20

the target organ, and likelihood ofadditional pathogens, the antibioticregimen should be adjusted accord-ingly (9).

The IR literature regarding the use

of antibiotic prophylaxis for chemo-embolization is small. In the absenceof randomized clinical trials, the ef-fectiveness of prophylaxis in this set-ting is unproven, although severalclinical series (9,11,12,6366) havesuggested that major infectious com-plications may be sustained in thispopulation. Many operators rou-tinely administer antibiotic prophy-laxis for this procedure, includingcoverage for skin flora and for Gram-negative enteric organisms, eventhough this practice has not been

prospectively proven to be of benefitfor all patients (9,11,12,61,6266).Some practitioners recommend con-tinued administration of antibioticsfor 37 days after chemoemboliza-tion to cover Gram-negative entericpathogens, although this too lacksprospective validation (9,67) . Pa-tients without an intact sphincter ofOddi as a result of earlier surgery orsphincterotomy or biliary drainageare at increased risk for subsequentabscess formation (67). The risk of

postembolization infection appearsto be reduced by the performance ofa bowel preparation the night beforetreatment and by ensuring coverageof Gram-positive and Gram-negativeaerobic and anaerobic organisms (eg,tazobactam/piperacillin). However,it remains to be seen how high theinfectious risks are, despite adoptionof an aggressive regimen (12).

Even fewer reports exist concern-ing infectious complications of ra-dioembolization. There have been in-dividual reports of hepatic abscess

after yttrium-90 radioembolization,but ongoing research is needed inthis population to better assess therisks of infectious complications andto determine the efficacy of antibioticprophylaxis (6769).

Procedure classification: clean; clean-contaminated (bilioenteric surgery)

Organisms encountered: S aureus,Streptococcus species, Corynebac-terium species with or without en-teric flora (in cases of earliersphincter of Oddi manipulation/

bilioenteric surgery)

Routine prophylaxis recommended:yes (if intent to create infarc-tion or high likelihood of solidorgan infarction)

First-choice antibiotic agent: no con-

sensusCommon antibiotic choices: (i) 1.53 g

ampicillin/sulbactam IV (hepaticchemoembolization); (ii) 1 g cefa-zolin and 500 mg metronidazoleIV (hepatic chemoembolization);(iii) 2 g ampicillin IV and 1.5mg/kg gentamicin (hepatic che-moembolization); (iv) 1 g ceftriax-one IV (hepatic chemoembolizationor renal or splenic emboliza-tion); (iv) if penicillin-allergic,vancomycin or clindamycin

plus aminoglycosideSpecial considerations: in patientsfor hepatic chemoembolizationwithout an intact sphincter ofOddi (previous sphincterot-omy, biliary drainage, historyof bilioenteric anastomosis),consider tazobactam/piperacil-lin; also consider bowel prepa-ration in this population

Level of evidence: 4, 7, 8

Uterine Artery Embolization

Uterine artery embolization (UAE)has become a desirable alternative tohysterectomy or myomectomy for thetreatment of symptomatic leiomyomas(13,70). The risk of infection has beenlargely described in the literature as

being low, and is reported to occur in0.2%1% of patients (9,13,71,72). How-ever, several of the reported cases offatal sepsis occurring after UAE haveoccurred in patients who did not re-ceive prophylaxis (9,7375). The mostcommon organisms causing infectionafter UAE are common skin flora(Staphylococcus or Streptococcus spe-cies), as is the case for other solid or-gan embolization. At least one fatalcase of sepsis after UAE has been re-ported in a patient who developed anEscherichia coli urinary tract infectionafter the procedure and fatal subse-quent E coli sepsis (9,74).

The role of prophylactic antibioticsfor patients undergoing UAE has beendebated in the literature (13). The jointworking party for the Royal College ofRadiologists and the Royal College of

Obstetricians and Gynecologists in the

United Kingdom has recommendedthat prophylactic antibiotic treatmentshould not be given at the time of theembolization as infectious complica-tions are generally delayed (as long as

23 weeks after the procedure) (13,76,77). In the Ontario Uterine FibroidEmbolization Trial (77), which in-cluded 555 patients, antibiotic prophy-laxis (1 g cefazolin IV) was routinelyadministered at four hospitals and re-served for patients at increased risk ofinfection at an additional four hospi-tals. This trial reported only two infec-tion-related hysterectomies in 570UAE procedures, with one occurringin each cohort (13,77). A retrospectivereview by Rajan et al (78) of 410 pa-tients undergoing UAE was performed toidentify risk factors for the developmentof intrauterine infection after UAE (13,78).Multiple variables were analyzed as pre-dictors for intrauterine infectious compli-cations requiring medical and/or sur-gical therapy, including the use ofpreprocedural antibiotics, embolicagent used, quantity of embolic mate-rial, location of tumors (eg, submuco-sal, nonsubmucosal), and size and lo-cation of the dominant tumor (78).Intrauterine infectious complicationsrequiring IV antibiotic therapy and/or

surgery occurred in five patients(1.2%), with no specific risk factor forintrauterine infection after UAE iden-tified (78). Rajan et al (78) concludedthat infection after embolization is rare(13,78). In contrast to these studies,there has been a single clinical seriesfrom the United Kingdom (9,79) thatdescribed a 17% readmission rate forinfection after UAE (seven of 42 pa-tients). There was one infection-re-lated hysterectomy. Three of the pa-tients had urinary tract infection, mostlikely secondary to bladder catheter-ization; in the three remaining pa-tients, no source of infection was iden-tified (9,79). The antibiotic prophylaxisregimen used in this clinical series in-cluded amoxicillin/clavulanate threetimes daily and metronidazole twicedaily on the day of the procedure, fol-lowed by a further 48-hour adminis-tration of the same antibiotic agentafter the procedure. Patients were ad-mitted to the hospital for 3 days (79).This multiple-day, multidrug ap-proach to prophylaxis is largely the

exception for UAE prophylaxis in the



8/20

IR literature; as in other clinical set-tings, multidrug regimens and pro-longed antibiotic administration have

been critiqued for eliminating normalGram-positive organisms, thereby al-

lowing Gram-negative organisms toproliferate (13,80). Notwithstandinglittle evidence from randomized con-trolled clinical trials, most clinical se-ries reporting on UAE have routinelyadministered a single dose of prepro-cedural antibiotic drugs, with the cur-rent antibiotic of choice being 1 g ofcefazolin (9). Cefazolin is inexpensiveand has activity against the likelysource of pathogens during solid or-gan embolization, ie, skin pathogens(Staphylococcus or Streptococcus spe-cies), as well as activity against E coli

(9). Other options that have beendescribed include gentamicin plusclindamycin, ampicillin, ampicillin/sulbactam, or vancomycin in the pen-icillin-allergic patient (9,72,80).

Patients with a history of hydrosal-pinx may warrant special consideration

before UAE, although there is no pub-lished consensus regarding whichagents should be used. Practitioners ex-perienced in performing UAE in pa-tients with a history of hydrosalpinx rec-ommend prophylaxis with doxycycline100 mg twice daily for 7 days before theprocedure (J. Spies, MD, personal com-munication, October 2009).

Procedure classification: clean; clean-contaminated

Organisms encountered: S aureus, Sepidermidis, Streptococcus specieswith or without E coli

Routine prophylaxis recommended: yesFirst-choice antibiotic agent: no

consensusCommon antibiotic choices: (i) 1 g ce-

fazolin IV; (ii) 900 mg clindamy-cin IV plus 1.5 mg/kg gentami-

cin; (iii) 2 g ampicillin IV;(iv) 1.53 g ampicillin/sulbac-tam IV; (v) if penicillin-allergic,can use vancomycin

Special considerations: if history ofhydrosalpinx, 100 mg doxycy-cline twice daily for 7 days


Transjugular IntrahepaticPortosystemic Shunt (TIPS) Creation

As noted by Beddy et al (13), tran-sjugular intrahepatic portosystemic

shunt (TIPS) creation was originally

used as a bridge to transplantation inpatients with severe complications ofportal hypertension (13). With the in-troduction of covered stents and theirimproved long-term patency, the em-

ployment of TIPS has expanded(13,81). Two separate infectious com-plications from TIPS have been de-scribed: (i) periprocedural sepsis with-out stent infection and (ii) TIPS stentinfection (9,13).

Periprocedural sepsis has beendescribed since the early days ofTIPS procedures and has been re-ported to occur in as many as 17% ofpatients (13,82). The causative organ-isms are usually skin flora (eg, Staph-

ylococc us or Streptococcus species)

(13). Deibert et al (83) studied theeffectiveness of a single dose of asecond-generation cephalosporin toprevent post-TIPS infection. Patientswho underwent 105 transjugular in-terventions were randomized to re-ceive no antibiotic treatment (46 in-terventions) or 2 g cefotiam (56interventions), which was adminis-tered at the beginning of the proce-dure (9,82). Post-TIPS infection wasdefined by an increase in white

blood cell count (15,000/L), fever(38.5C), or a positive blood cul-ture result (83). Patients who did notreceive cefotiam had an infectionrate of 20%, versus an infection rateof 14% in patients treated with anti-

biotics (83). The difference in infec-tion rate between the two groups didnot reach statistical significance, sug-gesting that routine use of prophy-lactic antibiotics for TIPS creationmay not be of value (9,83). As noted

by Ryan et al (9), the effectiveness ofcefotiam for TIPS prophylaxis mayhave been limited by its limited ac-

tivity against enterococcal species. Amore appropriate agent may have

been one with enhanced coverageagainst Enterococcus species, whichmay have resulted in a different out-come in this study (9). As patientsundergoing TIPS creation typicallyhave multisystem disease and do nottolerate the sequelae of infection,many interventionalists administerantibiotic prophylaxis in this setting,despite a relative lack of evidencefrom randomized controlled clinical

trials (8). The survey of IR prophy-

laxis regimens by Dravid et al (8)reported that a majority of respon-dents (69%) use antibiotic prophy-laxis in patients undergoing TIPScreation; an infection rate of 13% was

reported by respondents in this sur-vey. The authors concluded that pro-phylaxis is indicated for TIPS proce-dures, and suggested a regimen ofcefoxitin 1 g IV every 6 hours for 48hours (8). As noted by Beddy et al(13 ), some interventionalists useceftriaxone 1 g IV once daily for 48hours rather than cefoxitin, given itsenhanced activity against E coli, En-terobacter species, Gram-negative

bacteria, anaerobes, and enterococci,and its once-daily dosing regimen.Ampicillin/sulbactam also has im-proved coverage against Enterococcusspecies and can be used for TIPS pro-phylaxis (13).

Infection of the TIPS stent itselfbeen more recently described, re-portedly occurring in 1.7%5.1% ofcases (9,8286). Sanyal et al (82) werethe first to describe the clinical pic-ture of infection of the TIPS stent.Their diagnosis was based on thepresence of fever with positive bloodcultures and (i) the presence of stentthrombus or vegetations or (ii) per-

sistent bacteremia in a patient with aTIPS and no other detectable sourceof infection (9,82). Clinical signs andsymptoms reported in this setting in-cluded fever, tender hepatomegaly,hypoxemia, septic pulmonary em-

boli, septic shock, neutrophilia, andsubsequent development of necrotiz-ing fasciitis (9,82). The causative or-ganisms included oral and entericaerobic Gram-negative bacteria andCandida species (9). Most patients re-sponded to antibiotic therapy (9,82).DeSimone et al (84) studied 99 TIPSscreated during an 8-year period andidentified five patients with no otheralternative source of bacteremia whowere presumed to have TIPS stent infec-tions. Patients developed bacteremia amedian of 100 days after TIPS place-ment (range, 6732 d), well beyond theeffective period of antibiotic prophy-laxis. Bacteremia resolved in all casesafter treatment with IV antibiotic agents(84). Although acute infection related toTIPS creation appears to be uncommon,and the value of prophylactic antibiotic

agents has not been demonstrated via



9/20

randomized clinical trials, most investi-gators continue to administer antibioticprophylaxis to patients before TIPS pro-cedures (8,9,13).

Procedure classification: clean; clean-contaminated

Organisms encountered: skin flora(S epidermidis, S aureus), Coryne-bacterium species, biliary patho-gens, enteric Gram-negative rods,anaerobes, Enterococcus species

Routine prophylaxis recommended: yesFirst-choice antibiotic agent: no

consensusCommon antibiotic choices: ( i ) 1 g

ceftriaxone IV or (ii) 1.53 g am-picillin/sulbactam IV; (iii) ifpenicillin-allergic, can use van-

comycin or clindamycin andaminoglycoside


NONVASCULARINTERVENTIONS

Fluoroscopically GuidedGastrostomy and GastrojejunostomyTube Placement

Fluoroscopically guided percutane-ous radiologic gastrostomy was first de-

scribed in 1981 (14,85). Since that time,technical modifications have beenmade, including use of larger-bore tubes(2024 F vs initial use of 916 F tubes)and the introduction of gastropexy de-vices in 1986 (13,14,86). At present, thetwo most common methods of gastros-tomy tube placement are the pushmethod, which involves percutaneousintroducer technique through the ante-rior abdominal wall, and the more com-monly performed pull method, whichtraverses the oropharynx. Infectiouscomplications of gastrostomy tubeplacement are most commonly peris-tomal infections. The pull placementtechnique has been associated with ahigh incidence of peristomal infectiouscomplications (4%30%), resulting inrecommendations for routine prophy-lactic antibiotics for this procedure (8789). A prospective, randomized trialof prophylactic antibiotic agents (90)showed that cefazolin (1 g IV) reducedthe rate of infection from 28.6% to 7.4%in the setting of endoscopically placed(ie, pull-type) gastrostomy tubes. Of in-

terest, an increasing incidence of antibi-

otic-resistant peristomal infections hasbeen described, most commonly methi-cillin-resistant S aureus, attributed to na-sopharyngeal colonization, which trav-els along the pull-type percutaneous

endoscopic gastrostomy to the peris-tomal wound during gastrostomyplacement (9194). In principle, the per-cutaneous push technique avoids tubepassage through the oropharynx andthereby prevents the deposition of mi-croorganisms at the peristomal site.However, there are some data to sug-gest that, in head and neck cancer, thereis an increased risk of peristomal infec-tion and susceptibility to gastrostomysite infections irrespective of whether

bacteria are introduced through the

transoral or transesophageal route(14,95). Cantwell et al (14) described a15% rate of peristomal infection in a se-ries of 57 patients with head and neckcancer undergoing percutaneous gas-trostomy (n 53) or gastrojejunostomy(n 4) tube placement (via the push orintroducer technique). All instances ofperistomal infection occurred in patientswho had not received antibiotic prophy-laxis (n 20). Patients receiving pro-phylaxis (n 37) were administered 1 gcefazolin IV and twice-daily cephalexin500 mg for 5 days orally or via gastros-

tomy (n 35), or clindamycin 600 mgIV and 600 mg twice daily orally or viagastrostomy for 5 days (n 2) (14). De-spite these findings, there remains con-troversy with regard to prophylaxis forpush (ie, introducer) gastrostomy place-ment, as additional published data sug-gest no added benefit to antibiotic pro-phylaxis in this setting (95). Shastri et al(96) performed a prospective random-ized double-blind placebo controlledtrial in which 97 patients undergoingintroducer percutaneous endoscopic

gastrostomy placement for malignantoropharyngeal cancers were random-ized to receive placebo or prophylaxis (2g ceftriaxone IV). Infection rates werelow in both groups, with clinically sig-nificant wound infection observed inone patient in each group during theimmediate 7-day postprocedure fol-low-up (96).

Procedure classification: clean-con-taminated.

Organisms encountered: skin flora (S epi-dermidis, S aureus), Corynebacterium

species

Routine prophylaxis recommended: forpull technique, yes; for push tech-nique, no consensus

First-choice antibiotic agent: pull tech-nique, 1 g cefazolin IV

Special considerations: (i) second-genera-tion cephalosporin (head/neck can-cer); (ii) second-generation cephalo-sporin followed by an oral courseof a first-generation cephalosporin(head/neck cancer); (iii) if penicil-lin-allergic, can use vancomycin orclindamycin


LIVER AND BILIARYINTERVENTIONS

Biliary Drainage

The normal unobstructed biliarytree typically contains no bacteria (13).However, in the setting of biliary dis-ease, the biliary tract should be viewedas contaminated (13,97). The incidenceof infectious complication after biliarydrainage procedures has been re-ported to vary between 24% and 46%,with most interventional radiologistsroutinely using antibiotic prophylaxisfor biliary drainage (13). In the surveyof IR prophylaxis regimens by Dravid

et al (8), 89% of respondents indicatedthat they always used prophylaxis for

biliary drainage, with 53% of nonusersreporting infective complications after

biliary drainage. Preprocedural biliarycultures have been reported to be use-ful in planning the antibiotic strategy,although few interventionalists rou-tinely perform them (13). When ob-tained, the most common isolates in-clude Enterococcus species, as well asyeast, Gram-negative aerobic bacilli,and Streptococcus viridans (9,98). Se-

lected strains associated with mortal-ity include E coli and Clostridium spe-cies, which have been isolated in asmany as 75% of cases of fatal biliarysepsis (9,99). Risk factors for bacterialcolonization in patients with biliaryobstruction undergoing percutaneous

biliary drainage include periproce-dural fever, previous biliary instru-mentation, and bilioenteric anastomo-sis (9,98).

Although a majority of intervention-alists report antibiotic prophylaxis be-fore biliary drainage, there is variation

regarding the regimen advocated. Clark



10/20

et al (100) performed a prospectivestudy of 480 biliary procedures, rangingfrom simple tube injection to primary

biliary drainage tube placement, inwhich all patients received 1 g cefotetan

IV for prophylaxis. Forty-two patientsdeveloped an increased white blood cellcount or fever, with seven (2%) devel-oping overt sepsis; greater infection riskwas observed in patients who had un-dergone biliary intervention previously(9,100). As a result, these authors haveadvocated the addition of 4 g IV me-zlocillin to cefotetan in patients with ahistory of biliary intervention (100).Other investigators advocate prophy-laxis with third-generation cephalospo-rins, given these antibiotic agents en-hanced biliary excretion compared with

second-generation agents (9). Ryan et al(9) advocate the use of 1 g ceftriaxoneIV, citing its long duration of activityand its favorable dosing schedule. Am-picillin/sulbactam is considered bymany the optimal antibiotic agent forprophylaxis in the setting of biliary in-tervention, given its activity against En-terococcus species (9).

As noted by Ryan et al (9), patientswith advanced biliary disease, includingthose with hepatolithiasis, are at signif-icant risk of liver abscess formation, sec-ondary biliary cirrhosis, portal hyper-

tension, and death from sepsis orhepatic failure. In these patients, the

boundary between prophylaxis andtherapy is becomes blurred. Effectiveantimicrobial therapy is as crucial acomponent of therapy as relief of bilestasis by IR manipulation or clearance of

biliary stone burden (9). Sheen-Chen etal (101) demonstrated the presence of

bacteria in the bile of all patients withhepatolithiasis, most commonly Gram-negative bacteria such as Klebsiella spe-cies, E coli, and Pseudomonas, Enterococ-cus, and Bacteroides species, the latter

being the most frequently found anaer-obes (9,101). As is the case for other IRinterventions, antibiotic treatmentsshould be adjusted according to the re-sults of bacteriologic cultures, with anti-

biotic therapy continued in the setting ofobstruction, until relief of the obstruc-tion has been achieved (9,13,101).

Procedure classification: clean-con-taminated; contaminated

Organisms encountered: Enterococcusspecies, Candida species, Gram-negative aerobic bacilli, S viri-

dans, E coli, and Clostridium spe-

cies; Klebsiella, Pseudomonas, andBacteroides species, particularlyin cases of advanced biliary dis-ease, including hepatolithiasis

Routine prophylaxis recommended: yes

First-choice antibiotic agent: noconsensusCommon antibiotic choices: (i) 1 g ceftri-

axone IV; (ii)1.53 g ampicillin/sul-bactam IV; (iii) 1 g cefotetan IV plus4 g mezlocillin IV; (iv) 2 g ampicillinIV plus 1.5 mg/kg gentamicin IV;(v) if penicillin-allergic, can usevancomycin or clindamycin andaminoglycoside


GU PROCEDURES

Percutaneous Nephrostomy TubePlacement, Tube Exchange, UreteralStents

Percutaneous nephrostomy andureteric stent implantation are com-mon IR procedures (13,102). In gen-eral, procedures performed in theGU tract are regarded as clean-contaminated when they are per-formed on an unobstructed system,with no previous infection and nohistory of intervention (7). Factorspredisposing to infection include ad-

vanced age, diabetes, bladder dys-function, indwelling catheter, earliermanipulation, ureterointestinal anas-tomosis, bacteriuria, and stones(7,9,13,103,104). In these patients, theGU tract should be managed as con-taminated (7). The GU tract is re-garded as dirty if clinical infection ispresent; in these instances, antibiotictherapy rather than prophylaxis isthe appropriate aim (7). The mostcommon organisms to infect the GUtract are Gram-negative rods (E coli,Proteus species, and Klebsiella spe-

cies) and Enterococcus species (7). Ifculture results are not available, am-picillin combined with gentamicin orsulbactam is appropriate, as is ceftri-axone (7,13). Cronan et al (102) de-scribed an incidence of bacteremia of17% in patients undergoing nephros-tomy tube exchange, with no differ-ence in incidence between patientsreceiving preprocedural antibioticagents compared with those who didnot (102). Serious infectious compli-cations are encountered in patientsundergoing treatment for urinary

obstruction, with septic shock re-

ported in as many as 7% of patientsundergoing nephrostomy drainagefor pyonephrosis (7,103106). Coch-ran et al (104) stratified patients intoa high-risk group (aforementioned

risk factors) and a low-risk group(none of the aforementioned risk fac-tors) (104). In the low-risk group,14% developed evidence of sepsiswhen prophylactic antibiotic drugswere not given; 10% developed evi-dence of sepsis when prophylacticantibiotic drugs were given (no sta-tistical difference). In the high-riskgroup, these figures were 50% and9%, respectively (104). Despite theseresults, many authors recommendthe use of a prophylactic antibioticagent for all patients, irrespective of

risk, with a majority of intervention-alists reporting routine antibioticprophylaxis for GU procedures, andnonusers reporting a 40% incidenceof infective complications (8,13). Anexception to this practice includesthe routine tube change in unin-fected and unobstructed immuno-competent cases, in which routineprophylaxis has not been describedin the literature, and is not war-ranted (7). Patients who have signsof infection, or who are classified

into a high risk group as mentionedearlier, should be treated with an ap-propriate antibiotic agent before in-tervention (9). As in biliary interven-tions, antibiotic treatment should becontinued to treat obstructed sys-tems until relief of obstruction isachieved. The results of urine cul-tures should be used to tailor ongo-ing therapy (9).

Procedure classification: clean-con-taminated; contaminated

Organisms encountered: E coli and

Proteus, Klebsiella, and Enterococ-cus speciesRoutine prophylaxis recommended:

yes (except for routine tubechange in uninfected patients)

First-choice antibiotic: no consensusCommon antibiotic choices: (i) 1 g ce-

fazolin IV; (ii) 1 g ceftriaxone IV;(ii) 1.53 g ampicillin/sulbac-tam IV; (iv) 2 g ampicillin IVand 1.5 mg/kg gentamicin IV;(v) If penicillin-allergic, can usevancomycin or clindamycin andaminoglycoside

Level of evidence: 4, 5



11/20

TUMOR ABLATION

Percutaneous tumor ablation, in-cluding radiofrequency (RF) abla-tion, has effectively treated small (

3 cm) liver lesions, and has shownpromise in the treatment of lung, re-nal, and adrenal tumors (13,107110). The issue of prophylactic anti-

biotic agents for tumor ablation iscontroversial, with some operatorsadministering them universally andothers only in selected cases (13).There have been no randomized con-trolled trials on antibiotic agent usein patients undergoing RF ablation; atpresent, most of the published data per-taining to this topic relate to the per-sonal experience of various groups

(13). Infections after tumor ablationhave been mostly described in thesetting of hepatic RF ablation, withcomplications including cholangitisand liver abscess being rare (re-ported in 1.5% of cases) but poten-tially severe (13,111,112). Instancesin which patients are at an increasedrisk of hepatic abscess warrant care-ful evaluation, and include historiesof bilioenteric anastomosis, biliarystent placement, and sphincterot-omy, all of which lead to retrograde

enteric bacterial communicationwith the biliary tract (13). As notedby Choi et al (111), the mechanism ofabscess formation is thought to re-sult from bacterial colonization andgrowth in the zone of ablation (111).Before RF ablation of hepatic lesionsin routine cases, investigators haverecommended 1.5 g IV ampicillin/sulbactam (13). There remains noconsensus on the effectiveness ofprophylactic antibiotic agents for pa-tients undergoing RF ablation ofliver, lung, adrenal, renal, or othersolid lesions (13). Indeed, the occur-rence of delayed abscess formationweeks after tumor ablation despiteprophylactic antibiotic coveragemakes superinfection of the thermallesion a possibility, given that ther-mally damaged tissue may providean especially favorable environmentfor bacterial growth (109). This phe-nomenon highlights the importanceof postprocedural follow-up and sur-veillance to ensure timely and appro-priate antibiotic therapy, in addition

to up-front prophylaxis in patients at

risk. In the absence of definitive sci-entific evidence, many practitionerscontinue to empirically use prophy-laxis (13,64,112).

Procedure classification (site-dependent):clean; clean-contaminated (eg, bil-ioenteric anastomosis/bypass)

Organisms encountered (organ-depen-dent): generally S aureus, S epider-midis, Streptococcus species withor without E coli; in cases of pre-vious bilioenteric anastomosis,consider organisms similar tothose for liver/biliary interven-tion, eg, E coli, Proteus species,Klebsiella species, and Enterococ-cus species

Routine prophylaxis recommended: no

consensusFirst-choice antibiotic agent: no

consensusCommon antibiotic choices: (i) 1.5 g

ampicillin/sulbactam IV (liver);(ii) 1 g ceftriaxone IV (renal);(iii) 1 g cefazolin IV (bone); alter-nate choices (site-dependent),(iv) if penicillin-allergic, can sub-stitute vancomycin or clindamy-cin for Gram-positive coverage;aminoglycoside for Gram-nega-tive coverage


PERCUTANEOUS ABSCESSDRAINAGE

Patients referred for percutaneousabscess drainage are typically already

being treated with antibiotic drugs (7).When patients are referred for percu-taneous abscess drainage and are notalready being treated with antibioticdrugs, the IR physician should evalu-ate the appropriateness of empiric an-tibiotic therapy before obtaining cul-ture data via catheter drainage. In theabsence of published data on this sub-

ject, it is the opinion of the authors thatempiric antibiotic coverage before ab-scess drainage should be reserved forpatients who present with clinicalsigns and symptoms of infection (eg,fever, leukocytosis) at the time of thedrainage procedure. Admittedly, anti-

biotic agent use in this setting spansthe boundary between prophylaxisand treatment. As abscesses are typi-cally polymicrobial, broad-spectrum

antibiotic agents are warranted in the

absence of existing culture data. In theotherwise asymptomatic patient, thereis potential benefit to avoiding unnec-essary wide-spectrum antibiotic cover-age by awaiting the results of culture

and sensitivity of drainage specimenswhen available. The most common

bacteria found in intraabdominal ab-scesses are Gram-negative rods andanaerobes, particularly E coli, Bacte-roides fragilis, and Enterococcus species(7,8,113). Pyogenic liver abscesses aremost often caused by Enterobacter spe-cies and anaerobes (7,114). Older anti-

biotic regimens used in this settinghave included ampicillin, gentamicin,and metronidazole, but more commoncurrent regimes include second- orthird-generation cephalosporins, suchas cefoxitin 1 g IV every 6 hours, ceftri-axone 1 g IV every 24 hours, or ampi-cillin/sulbactam 3 g IV every 6 hours.A combination of clindamycin andgentamicin may be used if the patienthas a severe penicillin allergy (8).

Procedure classification: dirtyOrganisms encountered: skin flora, S epi-

dermidis, S aureus, Corynebacteriumspecies; intracavitary pathogens,Gram-negative bacteria, Enterococ-cus species, E coli, B fragilis, otheranaerobes

Routine prophylaxis recommended:yes

First-choice antibiotic agent: noconsensus

Common antibiotic choices: (i) 12 gcefoxitin IV every 6 hours;(ii) 12 g cefotetan IV every 12 h;(iii) 1 g ceftriaxone IV every 24 h;(iv) 3 g ampicillin/sulbactam IVevery 6 h; (v) if penicillin-aller-gic, can use vancomycin or clin-damycin for Gram-positivecoverage; aminoglycoside for

Gram-negative coverageLevel of evidence: 8

PERCUTANEOUS BIOPSY

As noted by McDermott et al (7),image-guided biopsies do not requireantibiotic prophylaxis unless per-formed via the transrectal route. Themost common indication for the tran-srectal approach is prostate biopsy (7).Suggested prophylaxis for this proce-dure includes preprocedural gentami-

cin 80 mg intramuscularly (IM) 30



12/20

minutes before the procedure, fol-lowed by a 5-day course of oral cipro-floxacin 250 mg twice daily (7). Sieberet al (113) advocate the use of oralciprofloxacin 500 mg twice daily for 4

days, commencing the day before bi-opsy. In their retrospective review of4,439 biopsies performed with the useof this antibiotic regimen (113), infec-tive complications (ie, urinary tract in-fection) occurred in a minority of pa-tients (n 5).

Procedure classification: nontransrectal,clean; transrectal, contaminated

Organisms encountered: transrectal,Gram-negative bacteria, Entero-coccus species, E coli, B fragilis,other anaerobes

Routine prophylaxis recommended:yes (transrectal route); no (non-transrectal route)

First-choice antibiotic agent: transrec-tal, no consensus

Common antibiotic choices: transrec-tal, (i) 80 mg gentamicin IM plus250 mg ciprofloxacin twice dailyorally for 5 days; (ii) 500 mg cip-rofloxacin twice daily orally for 4days, commencing the day be-fore biopsy


PERCUTANEOUSVERTEBROPLASTY

Vertebroplasty is employed to treatthe pain associated with osteoporoticcompression fractures and pathologicvertebral compression (13). Postproce-dure infection is rare in this setting,and clinical data from randomized tri-als are currently lacking. However,when infection does occur in this set-ting, surgical debridement of the in-fected vertebral body can be techni-

cally very difficult (13,115,116). Thecommon pathogens are from skinflora. Prophylactic antibiotics aretherefore commonly administered tothese patients, with some investigatorsadding antibiotics to the polymethylmethacrylate before intraosseous in-

jection, although the efficacy of the lat-ter has not been proven (13,117). Anappropriate prophylactic regimenagainst the involved pathogens in-volves 1 g cefazolin IV 30 minutes be-fore the procedure (13). If the patient isallergic to penicillin or cephalosporins,

vancomycin may be administered (13).

Procedure classification: cleanOrganisms encountered: skin flora (S

epidermidis, S aureus, Corynebacte-rium species)

Routine prophylaxis recommended:

yesFirst-choice antibiotic agent: ( i ) 1 gcefazolin IV

Alternate choices: (ii) if penicillin-al-lergic, can use vancomycin orclindamycin


MANAGEMENT OFANTIBIOTICHYPERSENSITIVITY

Patients frequently state that they

have a penicillin allergy (9). However,penicillin allergies are only uncom-monly manifested as an anaphylacticreaction. Macpherson et al (118) stud-ied 1,260 patients attending a pread-mission clinic before routine surgery,among whom 22% described an anti-

biotic allergic reaction. After review ofthe patient medical charts and on fur-ther follow-up questioning, the major-ity of allergies were known side ef-fects of the drugs (118). In all patientswho had been administered an alter-native agent as a result of a history of

previous penicillin allergy, nonewere found to have had a previousallergic reaction (13,118). Patients whohave a questionable penicillin allergy,or have had only fever or rash, may besafely given -lactam antibiotic agentswithout fear of anaphylaxis (9,13). If apatient has a penicillin allergy, the cli-nician should determine whether it isof an anaphylactic or nonanaphylacticvariety (9). Two percent of the popu-lation has some degree of penicillinhypersensitivity, but most reactions to

-lactam agents are nonanaphylacticand usually manifest clinically as amild maculopapular rash or fever (9).Such considerations are not insignifi-cant, as the cost of prescribing alterna-tive agents to a patient with a docu-mented penicillin allergy can be morethan double the cost of an appropriatepenicillin and generally results in theuse of more broad-spectrum agents,which are more likely to confer antibi-otic resistance (13,119). Penicillin aller-gies are only uncommonly caused byan anaphylactic reaction (eg, broncho-

spasm, laryngospasm, hypotension, or

hives) (13). However, if a patient hashad an anaphylactic reaction to peni-cillins, they should never receive pen-icillin. In addition, use of cephalospo-rins in these patients should be

approached with caution because ofthe 15% cross-hypersensitivity withthese agents (9). However, it is advis-able in this setting to query hospitalrecords to determine if the patient hasever received a cephalosporin in thepast without allergic reaction. If so, acephalosporin can be used in the fu-ture. Although monobactams (eg, az-treonam) are structurally related to-lactam agents, they are unrelated interms of allergic potential (9). There isno cross-reactivity between monobac-tam and -lactam agents, and these

drugs may be used safely in patientswith anaphylactic reactions to -lac-tam agents (9,120). However, it should

be noted that there is cross-reactivitybetween carbapenem (eg, imipenem,meropenem) and -lactam agents.

MANAGEMENT OFVALVULAR HEART DISEASE

Current AHA recommendations forantibiotic prophylaxis against infec-tive endocarditis (IE) are for those car-

diac conditions associated with thehighest risk of adverse outcomes fromendocarditis (121). These include casesof prosthetic cardiac valve or pros-thetic material used for cardiac valverepair; previous IE; unrepaired cya-notic congenital heart disease, includ-ing palliative shunts and conduits;completely repaired congenital heartdefect with prosthetic material or de-vice, whether placed by surgery or bycatheter intervention during the first 6months after the procedure; repaired

congenital heart disease with residualdefects at the site or adjacent to the siteof a prosthetic patch or prosthetic de-vice (which inhibit endothelialization);and cardiac transplantation recipientswho develop cardiac valvulopathy(121). Notably, except for these condi-tions, IE prophylaxis is not recom-mended for any other form of congen-ital heart disease, including mitralvalve prolapse (121).

For patients at high risk who un-dergo an invasive respiratory tractprocedure to treat an established in-

fection, such as drainage of an ab



13/20

scess or empyema, the antibiotic reg-imen administered to these patientsshould co ntain an agent activeagainst S viridans (121). Suggestedregimens include 2 g amoxicillin

orally, 2 g ampicillin IM or IV, or 1 gcefazolin IM or IV or 1 g ceftriaxoneIM or IV (121). In patients allergic topenicillin or ampicillin and able totake oral medications, 2 g cephalexinorally or 600 mg clindamycin orallyor 500 mg azithromycin orally or 500mg clarithromycin orally may be ad-ministered (121). In patients allergicto penicillins or ampicillin and un-able to take oral medication, 1g cefa-zolin IM or IV or 1 g ceftriaxone IMor IV or 600 mg clindamycin IM or IVmay be administered (121). If the in-

fection is known or suspected to becaused by S aureus, the regimen shouldcontain an agent active against S aureus,such as an antistaphylococcal penicillinor cephalosporin, or vancomycin in pa-tients unable to tolerate a-lactam agent(121). Vancomycin should be adminis-tered if the infection is known or sus-pected to be caused by a methicillin re-sistant strain of S aureus (121).

In a departure from previous rec-ommendations, the AHA indicatedin their 2007 guidelines (121) that ad-ministration of prophylactic antibi-

otic agents solely to prevent endocar-ditis should not be recommended forpatients who undergo GU or GI tractprocedures. Reasons cited for thisdeparture include the lack of studiesshowing a possible association betweenGI o r GU tract procedures and IE(122,123). As indicated by Wilson et al(121), the cases of IE temporally associ-ated with a GI or GU tract procedurehave been largely anecdotal, with a sin-gle or very small number of cases re-ported. No published data demon-strate a conclusive link between

procedures of the GI or GU tract andthe development of IE (121). Also citedis the increase in Enterococcus speciesresistant to penicillins, vancomycin,and aminoglycosides (122124). Thesewere antibiotics recommended for IEprophylaxis in previous AHA guide-lines (125). Therefore, only in patients

at high risk who have an establishedGI or GU tract infection or for thosewho receive antibiotic therapy to pre-vent wound infection or sepsis associ-ated with a GI or GU tract procedure is

antienterococcal coverage consideredreasonable, such as with penicillin,ampicillin, piperacillin, or vancomycin(121). However, as noted by the cur-rent AHA guideline authors (121), nopublished studies demonstrate thatsuch therapy would prevent entero-coccal IE. For patients at high riskscheduled for an elective ureteral stentplacement or other urinary tract ma-nipulation who have an enterococcalurinary tract infection or colonization,antibiotic therapy to eradicate Entero-coccus species from the urine before

the procedure is considered reason-able (121). Amoxicillin or ampicillin isthe preferred agent for enterococcalcoverage for these patients. Vancomy-cin may be administered to patientsunable to tolerate ampicillin (121).

For patients at high risk who un-dergo an invasive procedure that in-volves infected skin, skin structure,or musculoskeletal tissue, the thera-peutic regimen administered fortreatment of the infection shouldcontain an agent active against Staph-

yloco ccu s species and -hemolytic

Streptococcus species, such as an an-tistaphylococcal penicillin or a ceph-alosporin (121). Vancomycin or clin-damycin may be administered topatients unable to tolerate a -lactamagent or who are known or sus-pected to have an infection caused

by a methicillin-resistant strain ofStaphylococcus (121).

CONCLUSIONS

Effective antibiotic prophylaxisfor vascular and IR requires a thor-

ough knowledge of likely pathogens,procedure-specific infection risks,and appropriate antibiotic coverage(7). Choice of medication should alsotake into account patient-specific fac-tors, including renal and hepaticfunction and antibiotic allergy his-tory. There continues to be few ran-

domized controlled data to help for-mulate the prophylactic regimens forinterventional procedures. Clinicalpractice is large ly info rme d, atpresent, via existing surgical data

and interventional radiology cohortstudies and clinical series. Given theability of antibiotic resistance to ob-viate historically effective regimens,an ongoing review of current pri-mary surgical and IR literature onthis topic by interventional radiolo-gists remains crucial. It is equally im-portant to remain abreast of currenthospital and/or area practice. Thesemeasures underscore a timely andevidence-based approach to antibi-otic prophylaxis for interventionalprocedures.

Acknowledgments: Dr. Aradhana M.Venkatesan authored the first draft ofthis document and served as topic leaderduring the revisions of the draft. Dr. San-joy Kundu is chair o f the SIR Standards ofPractice Committee. Dr. John F. Cardellais Councilor of the SIR Standards Divi-sion. All other authors are listed alpha-betically. Other members of the Stan-dards of Practice Committee and SIR whoparticipated in the development of thisclinical practice guideline are as follows(listed alphabetically): John Fritz An-gle, MD, Ganesh Annamalai, MD, Ste-

phen Balter, PhD, Daniel B. Brown, MD,Danny Chan, MD, Christine P. Chao,MD, Horacio DAgostino, MD, Sean R.Dariushia, MD, Brian D. Davison, MD,Debra Ann Gervais, MD, S. Nahum Gold-berg, MD, Alan T. Hirsch, MD, Eric J.Hohenwalter, MD, Sanjeeva P. Kalva,MD, Arshad Ahmed Khan, MD, Venkat-aramu Krishnamurthy, MD, Ashish Ma-hajan, MD, Patrick C. Malloy, MD, GloriaM. Martinez-Salazar, MD, J . KevinMcGraw, MD, Philip M. Meyers, MD,Steven F. Millward, MD, Charles A.Owens, MD, Darren Postoak, MD, AnneC. Roberts, MD, Wael A. Saad, MD, Ta-

run Sabharwal, MD, Marc R. Sapoval,MD, Cindy Kaiser Saiter, NP, Marc S.Schwartzberg, MD, Samir S. Shah, MD,Nasir H. Siddiqi, MD, Raymond H.Thornton, MD, Patricia E. Thorpe, MD, T.Gregory Walker, MD, Ronald R. Ward-rope, RN, Darryl A. Zuckerman, MD,Curtis W. Bakal, MD, Curtis A. Lewis,MD, MBA, JD, Kenneth S. Rholl, MD.

APPENDIX A: CONSENSUS METHODOLOGY

Reported complication-specific rates in some cases reflect the aggregate of major and minor complications. Thresholdsare derived from critical evaluation of the literature, evaluation of empirical data from Standards of Practice Committee

members practices, and, when available, the SIR HI-IQ System national database.



14/20

Consensus on statements in this document was obtained with a modified Delphi technique (13).

APPENDIX B

Levels of Evidence and Classes of Recommendation to Adhere to Definitions Established by the American College of

Cardiology/American Heart Association Guidelines Task Force (4,5)

Level of Evidence Definition

1 Randomized clinical trials or metaanalyses of multiple clinical trials with substantial treatmenteffects

2 Randomized clinical trials with smaller or less significant treatment effects3 Prospective, controlled, nonrandomized, cohort studies4 Historic, nonrandomized, cohort, or case-control studies5 Case series: patients compiled in serial fashion, lacking a control group6 Animal studies or mechanical model studies7 Extrapolations from existing data collected for other purposes, theoretical analyses8 Rational conjecture (common sense); common practices accepted before evidence-based guidelines

APPENDIX C: SIR STANDARDS OF PRACTICE COMMITTEE CLASSIFICATION OFCOMPLICATIONS BY OUTCOME

Minor ComplicationsA. No therapy, no consequenceB. Nominal therapy, no consequence; includes overnight admission (up to 23 hours) for observation only.

Major ComplicationsC. Require therapy, minor hospitalization ( or to 24 hrs, but 48 hours)D. Require major therapy, unplanned increase in level of care, prolonged hospitalization ( 48 hours).E. Permanent adverse sequelaeF. Death



15/20

Practice Guidelines for Antibiotic Prophylaxis During Vascular and Interventional Radiology Procedures

Procedure

PotentialOrganisms

Encountered

RoutineProphylaxis

RecommendedFirst-choiceAntibiotic

Angiography, angioplasty, thrombolysis,arterial closure device placement, stentplacement

S aureus, S epidermidis No None

Endograft placement S aureus, S epidermidis Yes Cefazolin 1 g IVSuperficial venous insufficiency treatment S aureus, S epidermidis No NoneIVC filter placement S aureus, S epidermidis No NoneTunneled central venous access S aureus, S epidermidis No consensus None

Embolization and chemoembolization (ifintent to create infarction or highlikelihood of infarction)

S aureus, S epidermidis, Streptococcusspp, Corynebacterium spp, and/orenteric flora (if prior sphincter ofOddi manipulation orbilioenteric surgery)

Yes No consensus

UAE S aureus, S epidermidis, Streptococcusspp, and/or E coli

Yes No consensus

TIPS creation S aureus, S epidermidis,Corynebacterium spp, biliarypathogens, enteric Gram-negative rods, anaerobes,Enterococcus spp

Yes No consensus

Fluoroscopically guided gastrostomy andgastrojejunostomy tube placement

S aureus, S epidermidis,Corynebacterium spp

No consensus (ifintroducerpushtechnique); Yes(if pullgastrostomy)

1g cefazolin IV (pullgastrostomy)

Liver and biliary interventions Enterococcus spp, Streptococcus spp,aerobic Gram-negativeorganisms (E coli, Klebsiella spp,etc) Clostridium spp, Candida spp,and anaerobes

Yes No consensus

GU procedures E coli, Proteus, Klebsiella,Enterococcus

Yes No consensus

Tumor ablation S aureus, S epidermidis, Streptococcusspp, and/or E coli

No consensus No consensus

Percutaneous abscess drainage S aureus, S epidermidis,Corynebacterium spp; aerobicGram-negative bacteria andanaerobes

Yes No consensus

Percutaneous biopsy Transrectal: bowel flora, mostlyanaerobes and aerobic Gram-negative, Streptococcus spp

No(nontransrectal);yes (transrectal)

Nontransrectal, none;transrectal, 80 mggentamicin IV/IMplus 250 mgciprofloxacin twicedaily orally for 5 d

Percutaneous vertebroplasty S aureus, S epidermidis,

Corynebacterium spp

Yes 1 g cefazolin IV



16/20

Practice Guidelines for Antibiotic Prophylaxis During Vascular and Interventional Radiology Procedures

Common Antibiotic Choices CommentsLevel of

Evidence

1 g cefazolin IV (if high risk stent infection); if penicillin-allergic, can use vancomycin or clindamycin

Procedure classification: clean 5, 8

If penicillin-allergic, can use vancomycin or clindamycin Procedure classification: clean 5, 8None Procedure classification: clean 8None Procedure classification: clean 5, 61g cefazolin IV (eg, immunocompromised patients before

chemotherapy; history of catheter infection); ifpenicillin-allergic, can use vancomycin or clindamycin

Procedure classification: clean (nontunneled catheter:no prophylaxis)

8

1.53 g ampicillin/sulbactam IV (hepaticchemoembolization); 1 g cefazolin and 500 mgmetronidazole IV (hepatic chemoembolization); 2 gampicillin IV and 1.5 mg/kg gentamicin (hepaticchemoembolization); 1 g ceftriaxone IV (hepaticchemoembolization or renal, splenic embolization); if

penicillin-allergic, use vancomycin or clindamycin andaminoglycoside

Special considerations: if patient without intactsphincter of Oddi, consider tazobactam/piperacillinand bowel preparation; procedure classification:clean-contaminated; contaminated (bilioentericsurgery)

4, 7, 8

(i) 1 g cefazolin IV; (ii) 900 mg clindamycin IV plus 1.5mg/kg gentamicin IV; (iii) 2 g ampicillin IV; (iv) 1.53g ampicillin/sulbactam IV; if penicillin-allergic, canuse vancomycin or clindamycin

If history of hydrosalpinx, doxycycline 100 mg twicedaily for 7 d; procedure classification, clean; clean-contaminated

4, 5, 8

(i) 1 g ceftriaxone IV; (ii) 1.5-3 g ampicillin/sulbactam IV;if penicillin-allergic, can use vancomycin andaminoglycoside

Procedure classification: clean; clean-contaminated 2, 4, 5

Procedure classification: clean-contaminated; specialconsiderations: if head and neck cancer, consider(i) a second-generation cephalosporin; (ii) second-generation cephalosporin followed by oral course offirst-generation cephalosporin, or (iii) clindamycin orvancomycin in penicillin-allergic patients

3, 5

(i) 1 g ceftriaxone IV; (ii) 1.53 g ampicillin/sulbactamIV; (iii) 1 g cefotetan IV and 4 g mezlocillin IV; (iv)2 g ampicillin IV and 1.5 mg/kg gentamicin IV; (v) ifpenicillin-allergic, can use vancomycin or clindamycinand aminoglycoside

Procedure classification: clean-contaminated,contaminated

5

(i) 1 g cefazolin IV; (ii) 1 g ceftriaxone IV; (iii) 1.53 gampicillin/sulbactam IV; (iv) 2 g ampicillin IV and1.5 mg/kg gentamicin IV; (v) if penicillin-allergic, canuse vancomycin or clindamycin and anaminoglycoside

Procedur

antibiotic guide for ir

Documents