CONGEN

ITALHEA

RT

Implementing a Standardized PerioperativeAntibiotic Prophylaxis Protocol for NeonatesUndergoing Cardiac SurgeryMeghan T. Murray, MPH, Rozelle Corda, FNP, Rebecca Turcotte, MD, Emile Bacha, MD,Lisa Saiman, MD, MPH, and Ganga Krishnamurthy, MBBSSchool of Nursing, Department of Surgery, and Department of Pediatrics, Columbia University Medical Center;and Department of Infection Prevention and Control, New York-Presbyterian Hospital, New York, New York

Background. A lack of perioperative antibiotic pro-phylaxis guidelines for neonates undergoing cardiacsurgery has resulted in a wide variation in practice. Wesought to do the following: (1) Determine the safety of aperioperative antibiotic prophylaxis protocol for neonatalcardiac surgery as measured by surgical site infections(SSIs) rates before and after implementation of the pro-tocol; and (2) evaluate compliance with selected processmeasures for perioperative antibiotic prophylaxis.

Methods. This quasi-experimental study included allcardiacprocedures performedonneonates fromJuly 2009 toJune 2012 at a single center. An interdisciplinary task forcedeveloped a standardized perioperative antibiotic prophy-laxis protocol in the fourth quarter of 2010. The SSI rateswere compared in the preintervention (July 2009 toDecember 2010) versus the postintervention periods(January 2011 to June 2012). Compliance with process mea-sures (appropriate drug, dose, timing, and discontinuation

Accepted for publication April 21, 2014.

Address correspondence to Dr Krishnamurthy, Columbia UniversityMedical Center, 3959 Broadway BH12 North, #1211, New York,NY 10032; e-mail: gk2008@columbia.edu.

� 2014 by The Society of Thoracic SurgeonsPublished by Elsevier Inc

of perioperative antibiotic prophylaxis) was compared inthe 2 periods.Results. During the study period, 283 cardiac pro-

cedures were performed. The SSI rates were similar in thepreintervention and postintervention periods (6.21 vs 5.80per 100 procedures, respectively). Compliance with the 4process measures significantly improved postintervention.Conclusions. Restricting the duration of perioperative

antibiotic prophylaxis after neonatal cardiac surgery to 48hours in neonates with a closed sternum and to 24 hoursafter sternal closurewas safe anddid not increase the rate ofSSIs. Compliancewith selectedprocessmeasures improvedin the postintervention period. Additional multicenterstudies are needed to develop national guidelines for peri-operative prophylaxis for this population.

(Ann Thorac Surg 2014;98:927–33)� 2014 by The Society of Thoracic Surgeons

urgical site infections (SSIs) are a serious health

Scomplication after operative procedures and areassociated with increased mortality and health care costs[1]. The SSI rates after pediatric cardiac surgery rangefrom 2.3% to 9.9% [2–7]. Thus, prevention of SSIs is aquality improvement metric for health care organizations,and state and federal regulatory bodies [8–14]. The JointCommission has made the reduction of SSIs and imple-mentation of evidence-based practices for prevention ofSSIs, including appropriate use of perioperative prophy-laxis, a national patient safety goal for 2013 [8]. Severalnational organizations including the Society of ThoracicSurgeons, the Centers for Disease Control andPrevention, American Society of Health-System Phar-macists, the Surgical Infection Society, the InfectiousDiseases Society of America, and the Society for HealthCare Epidemiology of America have developed guide-lines for perioperative antibiotic prophylaxis after cardiacsurgery [9–14].

However, current guidelines are largely restricted toadult populations as few studies describe the efficacy ofperioperative antibiotic prophylaxis for children [11].Thus, pediatric recommendations are largely based onexpert opinion and from data extrapolated from adultpatients [11]. As a result, there is wide variation inpractice across pediatric cardiac centers; severaldifferent prophylactic antibiotic regimens are followedand for varying durations [15–21]. Despite lack of evi-dence of clinical benefit, extended antibiotic prophylaxisuntil removal of thoracotomy tubes or invasivelines remain common practices in pediatric populations[3, 15–21].The Surgical Infection Prevention Project (SIP) was

introduced in 2002 as a quality improvement initiativeto promote optimal use of perioperative antibiotic pro-phylaxis [10]. The SIP adopted 3 processes of caremeasures to reduce SSIs; selection of appropriateantibiotic consistent with published guidelines,administration of parenteral antimicrobial prophylaxiswithin 60 minutes prior to skin incision, anddiscontinuation of antimicrobial prophylaxis within 24hours of surgery and within 48 hours for cardiacsurgery to reduce SSI. These evidence-based recom-mendations have been incorporated into the broader

0003-4975/$36.00http://dx.doi.org/10.1016/j.athoracsur.2014.04.090

928 MURRAY ET AL Ann Thorac SurgNEONATAL ANTIBIOTIC PROPHYLAXIS 2014;98:927–33

CONGEN

ITALHEA

RT

Surgical Care Improvement Project, and adopted by theJoint Commission and other national quality improve-ment forums [10].

To date, evidence-based recommendations are un-available for perioperative antibiotic prophylaxis forneonates who undergo cardiac surgery. We hypothe-sized that limiting prophylaxis to 48 hours as recom-mended for adult cardiac surgical patients would be safein neonates. We present a protocol, developed at ourinstitution by an interdisciplinary task force, for peri-operative antibiotic prophylaxis for neonatal cardiacsurgery. A quasi-experimental study of all neonates whounderwent cardiac surgery was conducted with thefollowing objectives: (1) To determine the safety of aperioperative antibiotic prophylaxis protocol forneonatal cardiac surgery as measured by SSI ratesbefore and after implementation of the protocol; and (2)valuate compliance with selected process measures forperioperative antibiotic prophylaxis.

Material and Methods

Study Design, Site, and SubjectsWe performed a quasi-experimental study of neonateswho underwent cardiac surgery from July 1, 2009 to June30, 2012 at our tertiary care academic center in New YorkCity. Some subjects were previously included in a studyof risk factors for SSIs [22]. The SSIs were compared inthe preintervention (July 1, 2009 to December 31, 2010)and postintervention (January 1, 2011 to June 30, 2012)periods. Included subjects were neonates 30 or less daysof age who underwent cardiac surgery and receivedpreoperative and postoperative care in the cardiacsection of the neonatal intensive care unit (NICU).Excluded neonates were premature (< 37 completedweeks of gestation) infants who underwent ligationof the ductus arteriosus only, in whom insertion ofa permanent pacemaker was the only procedureperformed (as the perioperative prophylaxis regimen isdifferent), and who died within 24 hours after surgery.The Institutional Review Board for Columbia UniversityMedical Center approved this study with a waiver ofinformed consent.

Perioperative Antimicrobial Prophylaxis RegimenIn the fourth quarter of 2010, after reviewing existingevidence, an interdisciplinary task force developed aconsensus-based protocol for perioperative antibioticprophylaxis for neonatal cardiac surgery. The task forceincluded representatives from Departments of PediatricCardiothoracic surgery, Pediatric Cardiology, Neona-tology, Pediatric Infectious Disease, Infection Preventionand Control, and Anesthesiology. The perioperativeprophylaxis recommendations included the use of theappropriate antimicrobial agent, dose, timing of preop-erative dose, intraoperative re-dosing (if relevant), anddiscontinuation (Table 1). After receiving institutionalapproval, the protocol was implemented on January 1,2011. During the implementation period, the task force

met quarterly and SSI rates and compliance with theprophylaxis recommendations data were shared.During the preintervention period, the perioperative

antimicrobial prophylaxis regimen was similar to thatfollowed during the postintervention period with 3 ex-ceptions. (1) For neonates who underwent lateral thora-cotomy or median sternotomy whose sternal wound wasclosed in the operating room, cefazolin was continueduntil all chest tubes were removed. (2) For neonateswhose sternal wound was left open, vancomycin andgentamicin were provided until sternal closure afterwhich cefazolin was provided until all chest tubes wereremoved. (3) Prescriber audit and feedback of perioper-ative prophylaxis parameters were not performed.

Process of Care MeasuresThe interdisciplinary task force adopted the following 4processes of care measures to assess compliance with thestandardized protocol in the areas of correct drug, dose,timing, and discontinuation.

(1) Correct drug: as per protocol (Table 1).(2) Correct dose for cefazolin or vancomycin: � 10% of

recommended dose (Table 1).(3) Correct timing: Preoperative, as per protocol

(Table 1). Intraoperative re-dosing; cefazolin re-dosing should be initiated 4 hours � 30 minutesfrom previous dose or 8 hours � 30 minutes fromprevious dose if vancomycin is used. Postoperative,as per protocol (Table 1).

(4) Correct discontinuation: as per protocol (Table 1).

Case DefinitionsThe task force performed active surveillance for SSIspreintervention and postintervention. The Center forDisease Control and Prevention National HealthcareSurveillance Network definition for SSI was used [23].Patients were followed for SSI for 90 days after thesurgical procedure. The Pediatric Infection Preventionand Control, NICU, and cardiothoracic surgery teamsdetected potential SSIs. A member of the cardiacsurgical team adjudicated cases (R.C.). If a patient hadmore than 1 surgical procedure, the SSI was attributedto the most recent procedure unless the procedure wasperformed secondary to infection (eg, debridement).

Data CollectionThe following data were collected from each patient’selectronic medical record by a single investigator (M.M.):age at time of initial surgery; sex; weight; gestational age;risk adjustment in congenital heart surgery (RACHS-1)category; compliance with the 4 process measures forappropriate perioperative antibiotic prophylaxis; and SSImicrobiology results, when available. The RACHS-1groups cardiac surgical procedures with similar opera-tive mortality rates into 6 categories with group 1 pro-cedures carrying the lowest risk and group 6 procedures,the highest risk of mortality [24].

Table1.

Protocol

forPeriop

erativeAntibiotic

Prop

hylaxisforNeona

talCardiac

Surgeryan

dProcessof

CareMeasures

Surgical

Approach,O

pen

orClosedSternum

Pre

andIntrao

perativeAntibiotic

Prophylax

isPostoperativeAntibiotic

Prophylax

isDuration

ofAntibiotic

Prophylax

isCardiacsu

rgerythrough

lateralthoracotom

yor med

iansternotom

y,closed

sternum

Cefazolin:3

0mg/kg

/dose

Given

intheORwithin

60minutesprior

toskin

incision

.Rep

eatdosein

4hou

rsifsu

rgical

procedure

exceed

s4hou

rs.

Cefazolin:3

0mg/kg

/dose

�7day

sold:q

12hou

rs>7day

sold:q

8hou

rs

Disco

ntinuean

tibiotics48

hou

rsaftersu

rgeryen

dtime.

Use

vanco

mycin

15mg/kg

/dose

(Administerupto

120minutesprior

toskin

incision

)ifkn

ownco

lonization/infectionwithMRSA

Use

vanco

mycin

15mg/kg

/dose

�7day

sold:q

12hou

rs>7day

sold:q

8hou

rsifkn

ownco

lonization/

infectionwithMRSA

.Cardiacsu

rgery,

open

sternum

Cefazolin

30mg/kg

/dose

Given

intheORwithin

60minutesprior

toskin

incision

Rep

eatdosein

4hou

rsifsu

rgical

procedure

exceed

s4hou

rs.

Van

comycin:

15mg/kg

/dose

�7day

sold:q

12hou

rs>7day

sold:q

8hou

rsan

d

Disco

ntinuean

tibiotics24

hou

rsaftersternal

closure

Use

vanco

mycin

15mg/kg

/dose

(Administerupto

120minutesprior

toskin

incision

)ifkn

ownco

lonization/infectionwithMRSA

.

Gen

tamicin

Allag

es:4

mgq24

hou

rs

MRSA

¼methicillin

resistan

tStap

hylococcus

aureus;

OR

¼op

eratingroom

;q¼

every.

929Ann Thorac Surg MURRAY ET AL2014;98:927–33 NEONATAL ANTIBIOTIC PROPHYLAXIS

CONGEN

ITALHEA

RT

Statistical MethodsDescriptive analyses were completed to determine theincidence of SSI and compliance with the perioperativeantibiotic prophylaxis guidelines. To ensure the compa-rability of the patient population in the preinterventionand postintervention periods, characteristics of includedneonates undergoing surgery during these 2 time periodswere compared. For continuous and categoric character-istics, t tests and c2 tests were used to test for statisticalsignificance, respectively. The proportions of appropriateperioperative antibiotic prophylaxis and the rate of SSIbefore and after protocol implementation were comparedusing c2tests. Differences were considered significant fora p value 0.05 or less. Statistical analyses were conductedusing SAS (9.2 for Windows; SAS Institute Inc, Cary, NC).

Results

During the study period, 305 neonates underwent 317cardiac surgical procedures, of which 34 were excluded(Fig 1). In the final analysis, 283 procedures wereincluded; 145 preintervention and 138 postintervention.Overall, the study population was similar during thepreintervention and postintervention periods (Table 2).However, significantly more complex surgeries wereperformed during the postintervention period (p ¼ 0.02).

Incidence of SSI During Preintervention andPostintervention PeriodsDuring the study period, 17 SSIs (6.00 SSI per 100 pro-cedures) were diagnosed; 14 were superficial and 3 weredeep incisional SSIs. The rate of SSI was similar in thepreintervention and postintervention periods (6.21 vs 5.80per 100 procedures, respectively, p ¼ 0.88). The mediannumber of days to SSI was 20.0 (range 4 to 49 days) anddid not differ significantly between the preinterventionand postintervention periods (p ¼ 0.53).

Pathogens Causing SSIMethicillin-susceptible Staphylococcus aureus (MSSA) wasthe most common pathogen during both study periods.The MSSA caused 67% of SSI in the preinterventionperiod and 38% of SSI during the postintervention period(p ¼ 0.35).

Compliance With Perioperative Antibiotic ProphylaxisProcess MeasuresThe most common deficiencies noted during the pre-intervention period were the following (Table 3): (1) 20%of patients did not receive antibiotics within 60 minutes ofskin incision; (2) 84% of patients had inadequatepostoperative dosing; and (3) 27% of eligible patientsdid not receive intraoperative re-dosing. As the durationof antibiotic prophylaxis was governed by the presence ofchest tubes in the preintervention period, as expected,only 38% of subjects had antibiotics discontinued within48 hours of surgery or within 24 hours of chest closure.During the postoperative period, compliance with

the 4 process of care indicators significantly improved

Fig 1. Figure describes a flow diagram ofpatient population and procedures.

930 MURRAY ET AL Ann Thorac SurgNEONATAL ANTIBIOTIC PROPHYLAXIS 2014;98:927–33

CONGEN

ITALHEA

RT

and generally exceeded 90% (Table 3). However,postoperative drug dosing remained suboptimal; only65% of subjects received the recommended dose andunderdosing of cefazolin was the most commondeficiency (range, 5 to 26 mg/kg). Compliance withappropriate antibiotic prophylaxis was similar amongsubjects with and without SSIs (Table 4).

Comment

To our knowledge, this is the first study to investigate theimpact of a standardized perioperative antibiotic pro-phylaxis protocol for neonatal cardiac surgery. Wefocused on neonates because they are at greater risk forSSIs than older children and adults [2, 4, 7]. Furthermore,no recommendations exist for perioperative antibioticprophylaxis after neonatal cardiac surgery. The AmericanSociety of Health-System Pharmacists report offers rec-ommendations for cardiac surgical antibiotic prophylaxisfor adults and children older than 1 year of age, but donot include neonates and younger infants [11].

Table 2. Characteristics of Neonates Who Underwent Cardiac Sur

CharacteristicPreintervention Period

n ¼ 145 (%)

Male 80 (55)Mean age (days) at surgery, std dev. 13.7 � 23.3Mean birth weight (kg) SD 3.16 � 0.6Premature, <37 weeks 21 (14)White race 62 (43)RACHS-1 category >3 56 (39)

RACHS-1 ¼ risk adjustment in congenital heart surgery.

We confirmed our hypothesis that adult guidelineswere safe to use in neonates. Limiting the duration ofperioperative antibiotic prophylaxis to 48 hours in neo-nates with a closed sternum did not increase SSI rates inour study. The 6% rate of SSI in our study population iswithin the range reported in other studies; however,those studies also included older children [2–7]. Our re-sults are similar to studies of adult and pediatric cardiacsurgical patients wherein no incremental benefit wasnoted for extending prophylaxis beyond 48 hours or untilchest tubes or invasive catheters and vascular lines wereremoved [18, 25]. However, our findings differed fromthose reported by Maher and colleagues [20] whereinlimiting prophylaxis to 48 hours increased the rate ofSSI in pediatric cardiac surgical patients compared withwhen guided by the presence of thoracotomy tubes andvascular catheters.Recommendations for perioperative prophylaxis have

been proposed by several national organizations,including the Society of Thoracic Surgeons, andall discourage the extended use of antibiotic prophylaxis

gery in the Preintervention and Postintervention Periods

Postintervention Periodn ¼ 138 (%) c2/t Tests p Value

76 (55) 0.0003 0.9911.6 � 15.3 �0.89 0.383.07 � 0.62 �1.20 0.2312 (9) 2.30 0.1359 (43) 0.00 1.0073 (53) 5.81 0.02

Table 3. Comparison of Perioperative Antibiotic Prophylaxis Compliance Process Measures in the Preintervention andPostintervention Periods

Performance MeasurePreintervention Period

n ¼ 145 (%)Postintervention Period

n ¼ 138 (%) p Value OR 95% CI

Perioperative prophylaxis givenPreoperative 138 (95) 137 (99) 0.07 0.14 0.02–1.19Intraoperative re-dosing,n ¼ 131

44/60 (73) 70/71 (99) <0.01 0.04 0.01–0.31

Postoperative 141 (97) 136 (99) 0.68 0.52 0.09–2.88Correct drug

Preoperative 132 (96) 134 (98) 0.50 0.49 0.12–2.01Postoperative 124 (88) 133 (98) <0.01 0.16 0.05–0.58

Correct dosePreoperative 115 (87) 124 (93) 0.14 0.55 0.24–1.24Postoperative 20 (16) 86 (65) <0.01 0.11 0.06–0.19

Correct timingPreoperative 106 (80) 132 (99) <0.01 0.06 0.01–0.27Intraoperative 39 (89) 66 (94) 0.30 0.47 0.12–1.87

Correct discontinuation 47 (38) 126 (95) <0.01 0.03 0.01–0.08

CI ¼ confidence interval; OR ¼ odds ratio.

931Ann Thorac Surg MURRAY ET AL2014;98:927–33 NEONATAL ANTIBIOTIC PROPHYLAXIS

CONGEN

ITALHEA

RT

[11–13]. Prolonged antibiotic administration carries thehazard of promoting growth of resistant bacterial strainsor yeast, or developing Clostridium difficile colitis or othertoxicities without substantially decreasing infectious risk[11, 13, 18, 25]. However, surveys of practice haverevealed that prolonged antibiotic prophylaxis is commonpractice in pediatric cardiac surgery; 40% of surgeonscontinued antibiotic prophylaxis beyond 24 to 48 hoursafter surgery and 30% of centers continued prophylaxisuntil chest tube removal [15–17]. We speculate severalreasons why caregivers are reluctant to limit periopera-tive antibiotic prophylaxis duration. Most neonatal car-diac surgeries are performed on cardiopulmonary bypass,which has detrimental effects on cellular and humoralresponse [26]. Furthermore, the mortality rate is greatly

Table 4. Perioperative Antibiotic Prophylaxis Process Measures an

ParameterSSI

n ¼ 17 (%)No S

n ¼ 26

Antibiotics givenPreoperative 16 (94) 259Intraoperative, n ¼ 131 7/9 (78) 107/122Postoperative 17 (100) 260

Correct drugPreoperative 16 (100) 250Postoperative 16 (94) 241

Correct dosePreoperative 15 (94) 224Postoperative 6 (38) 100

Correct timingPreoperative 13 (81) 225Intraoperative 7 (100) 98

Correct discontinuation 12 (75) 161

CI ¼ confidence interval; OR ¼ odds ratio; SSI ¼ surgical site infectio

increased in patients who develop organ space SSIs[27]. Invasive devices do increase the risk of health careassociated infections but there is no evidence thatperioperative prophylaxis reduces this risk.No guidelines exist for antibiotic prophylaxis for pa-

tients who have an open sternum postoperatively. Hence,different antibiotic regimens are used and often forextended durations [15–20]. Our study did not find adifference in SSI rates in neonates in the preinterventionand postintervention periods.To date, most studies in children have focused on the

duration of antibiotic prophylaxis after cardiac surgery,rather than the selection, timing, and dosing of prophy-laxis. We chose to expand the process measures becauseadherence to appropriate timing and agent has reduced

d SSI

SI6 (%) p Value OR 95% CI

(97) 0.39 0.43 0.05–3.73(88) 0.24 0.49 0.09–2.59(98) 0.69 – –

(97) 0.58 – –

(93) 0.38 1.26 0.16–10.03

(90) 0.33 0.57 0.07–4.53(41) 0.20 0.85 0.30–2.40

(90) 0.23 2.08 0.55–7.79(92) 0.55 – –

(67) 0.18 1.49 0.47–4.77

n.

932 MURRAY ET AL Ann Thorac SurgNEONATAL ANTIBIOTIC PROPHYLAXIS 2014;98:927–33

CONGEN

ITALHEA

RT

SSIs in adult populations [28–30]. During the pre-intervention period only 80% of neonates who underwentsurgery received antibiotic prophylaxis within 60 minutesof skin incision. Furthermore, about 25% of neonates whowere candidates for re-dosing did not receive an addi-tional dose of antibiotics in the operating room. We alsofound that despite quarterly audits and feedback to keystakeholders including the cardiothoracic surgical team,neonatal cardiac care team, and anesthesiologists, post-operative dosing of cefazolin remained suboptimal in thepostintervention period, as infants were frequentlyunderdosed. We speculate that a default dosing recom-mendation in our electronic medical record or the use ofan incorrect weight for dose calculations by providerslikely contributed to suboptimal dosing.

Several methods including hospital collaboratives,electronic medical records, and order sets have beenemployed to improve SIP performance measures in adultsurgical patients [29, 30]. Our study confirms the impor-tance of antibiotic stewardship programs and dedicatedinterdisciplinary task forces in improving adherence tothese process measures. Periodic audits, feedback, andeducation to prescribers and key stakeholders signifi-cantly improved the selection, dosing and timing, anddiscontinuation of prophylactic antibiotics in our study.

Despite improved adherence to perioperative antibioticprophylaxis measures in the postintervention period, astatistically significant decline in SSI rate was not noted inour study, although a higher proportion of complex caseswere performed. Studies evaluating the effectivenessof SIP process measures on reducing SSI rates inadult surgical patients have generated conflicting results[30–32]. While some studies have shown a decline in SSIrates with improved adherence to process measures,others have not demonstrated a similar benefit [30–32].Factors other than appropriate delivery of perioperativeantibiotic prophylaxis may also influence SSI ratesincluding hospital case volume, case-mix index, surgicalskill, working environment, and overall quality of care[31, 32].

Our study did not find significant differences in rates ofcompliance with any of the 4 process measures betweenneonates who developed SSI and those who did notdevelop SSI. These results appear to contradict our earlierfinding that incorrect timing of preoperative antibioticprophylaxis was a risk factor for SSI after cardiac surgery[22]. We speculate that several differences in study designmay account for these divergent findings. The previousstudy was conducted between January 2010 andDecember 2011, included all infants younger than 1 yearof age, and analyzed outcomes after 552 procedures.Thus, the sample size of the current study may havebeen too small to detect an association betweenincorrect preoperative timing of prophylactic antibioticand SSI. It is also possible that factors other thanappropriate perioperative prophylaxis influence SSIrates in neonates.

There are several limitations to this study. This is asingle-center study with a unique, cardiac NICU and thusthe findings may not be generalizable to other pediatric

cardiac ICUs in which care is offered to older children inthe same unit. Although the proportion of patients whoreceived appropriate perioperative antibiotic prophylaxisincreased significantly during the postinterventionperiod, the decline in the incidence of SSI during thisperiod was not statistically significant. The sample sizemay be too small or the study period may be too short tosee a statistically significant decrease in the rate of SSI.Furthermore, SSI may not have been reported if patientspresented to other institutions.In conclusion, this study demonstrates that limiting the

duration of perioperative antibiotic prophylaxis forneonatal cardiac surgery to 48 hours if the sternum isclosed or to 24 hours after sternal closure is safe asdetermined by SSI rates. Additional multicenter studiesare needed to develop national guidelines for perioper-ative prophylaxis for this population. The implementationof the consensus protocol with feedback resulted in sig-nificant improvements in the delivery of perioperativeantibiotic prophylaxis in 4 process measures includingappropriate drug selection and dosing, timely adminis-tration, and discontinuation. However, deficiencies inpostoperative dosing persisted in the postinterventionperiod. Further research is needed to determine the as-sociation between each of the perioperative antibioticprophylaxis process of care measures and the occurrenceof SSI.

This work was supported, in part, by the National Institute forNursing Research – National Institute of Health (R01 NR010821)to L.S.

References

1. de Lissovoy G, Fraeman K, Hutchins V, Murphy D, Song D,Vaughn BB. Surgical site infection: incidence and impact onhospital utilization and treatment costs. Am J Infect Control2009;37:387–97.

2. Levy I, Ovadia B, Erez E, et al. Nosocomial infections aftercardiac surgery in infants and children: incidence and riskfactors. J Hosp Infect 2003;53:111–6.

3. Costello JM, Graham DA, Morrow DF, et al. Risk factors forsurgical site infection after cardiac surgery in children. AnnThorac Surg 2010;89:1833–41.

4. Allpress AL, Rosenthal GL, Goodrich KM, Lupinetti FM,Zerr DM. Risk factors for surgical site infections after pedi-atric cardiovascular surgery. Pediatr Infect Dis J 2004;23:231–4.

5. Nateghian A, Taylor G, Robinson JL. Risk factors for surgicalsite infections following open-heart surgery in a Canadianpediatric population. Am J Infect Control 2004;32:397–401.

6. Ben-Ami E, Levy I, Katz J, Dagan O, Shalit I. Risk factors forsternal wound infection in children undergoing cardiacsurgery: a case-control study. J Hosp Infect 2008;70:335–40.

7. Sohn AH, Schwartz JM, Yang KY, Jarvis WR, Guglielmo BJ,Weintrub PS. Risk factors and risk adjustment for surgicalsite infections in pediatric cardiothoracic surgery patients.Am J Infect Control 2010;38:706–10.

8. The Joint Commission; Oakbrook Terrace, IL. Availablefrom: http://www.jointcommission.org/assets/1/18/NPSG_Chapter_Jan2013_HAP.pdf. Accessed on 9/18/2013.

9. Bratzler DW, Hunt DR. The Surgical Infection Preventionand Surgical Care Improvement Projects: national initiativesto improve outcomes for patients having surgery. Clin InfectDis 2006;43:322–30.

933Ann Thorac Surg MURRAY ET AL2014;98:927–33 NEONATAL ANTIBIOTIC PROPHYLAXIS

CONGEN

ITALHEA

RT

10. Bratzler DW, Houck PM. Antimicrobial prophylaxis for sur-gery: an advisory statement from the National SurgicalInfection Prevention Project. Clin Infect Dis 2004;38:1706–15.

11. Bratzler DW, Dellinger EP, Olsen KM, et al. Clinical practiceguidelines for antimicrobial prophylaxis in surgery. Am JHealth-Syst Pharm 2013;70:195–283.

12. Engelman R, Shahian D, Shemin R, et al. The Society ofThoracic Surgeons practice guideline series: antibiotic pro-phylaxis in cardiac surgery, part II: antibiotic choice. AnnThorac Surg 2007;83:1569–76.

13. Edwards FH, Engelman RM, Houck P, Shahian DM,Bridges CR. The Society of Thoracic Surgeons practiceguideline series: antibiotic prophylaxis in cardiac surgery,part I: duration. Ann Thorac Surg 2006;81:397–404.

14. Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR.Guideline for prevention of surgical site infection, 1999.Hospital Infection Control Practices Advisory Committee.Infect Control Hosp Epidemiol 1999;20:250–78.

15. Woodward CS, Son M, Calhoon J, Michalek J, Husain SA.Sternal wound infections in pediatric congenital cardiacsurgery: a survey of incidence and preventative practice. AnnThorac Surg 2011;91:799–804.

16. Alphonso N, Anagnostopoulos PV, Scarpace S, et al. Peri-operative antibiotic prophylaxis in paediatric cardiac sur-gery. Cardiol Young 2007;17:12–25.

17. Knoderer CA, Anderson EM, Cox EG. Survey of congenitalheart surgeons’ preferences for antimicrobial prophylaxis forpediatric cardiac surgery patients. Am J Health-Syst Pharm2008;65:2008–10.

18. Kato Y, Shime N, Hashimoto S, et al. Effects of controlledperioperative antimicrobial prophylaxis on infectious out-comes in pediatric cardiac surgery. Crit Care Med 2007;35:1763–8.

19. Dagan O, Cox P, Ford-Jones L, Ponsonby J, Bohn D. Nosco-mial infection following cardiovascular surgery: Comparisonof two periods, 1987 vs. 1992. Crit Care Med 1999;27:104–8.

20. Maher KO, VanDerElzen K, Bove EL, Mosca RL,Chenoweth CE, Kulik TJ. A retrospective review of threeantibiotic prophylaxis regimens for pediatric cardiac surgicalpatients. Ann Thorac Surg 2002;74:1195–2000.

21. �Alvarez P, Fuentes C, García N, Modesto V. Evaluation ofthe duration of the antibiotic prophylaxis in paediatric

postoperative heart surgery patients. Pediatr Cardiol 2012;33:735–8.

22. Murray MT, Krishnamurthy G, Corda R, et al. Surgical siteinfections and bloodstream infections in infants after cardiacsurgery. J Thorac Cardiovasc Surg 2013;S0022-5223(13):00968-9. [Epub ahead of print].

23. Horan TC, Andrus M, Dudeck MA. CDC/NHSN surveillancedefinition of health care-associated infection and criteria forspecific types of infections in the acute care setting. Am JInfect Control 2008;36:309–32.

24. Jenkins KJ. Risk adjustment for congenital heart surgery: theRACHS-1 method. Semin Thorac Cardiovasc Surg PediatrSurg Annu 2004;7:180–4.

25. Harbarth S, Samore MH, Lichtenberg D, Carmeli Y. Pro-longed antibiotic prophylaxis after cardiovascular surgeryand its effect on surgical site infections and antimicrobialresistance. Circulation 2000;101:2916–21.

26. Markewitz A, Faist E, Weinhold C, et al. Alterations of cell-mediated immune response following cardiac surgery. EurJ Cardiothorac Surg 1993;7:193–9.

27. Tortoriello TA, Friedman JD, McKenzie DE, et al. Media-stinitis after pediatric cardiac surgery: a 15-year experience ata single institution. Ann Thorac Surg 2003;76:1655–60.

28. Bratzler DW, Houck PM, Richards C, et al. Use of antimi-crobial prophylaxis for major surgery. Baseline results fromthe National Surgical Infection Prevention Project. Arch Surg2005;140:174–82.

29. McCahill LE, Ahern JW, Gruppi LA, et al. Enhancingcompliance with Medicare guidelines for surgical infectionprevention. Experience with a cross-disciplinary qualityimprovement team. Arch Surg 2007;142:355–61.

30. Dellinger EP, Hausmann SM, Bratzler DW, et al. Hospitalscollaborate to decrease surgical site infections. Am J Surg2005;190:9–15.

31. Stulberg JJ, Delaney CP, Neuhauser DV, Aron DC, Fu P,Koroukian SM. Adherence to Surgical Care ImprovementProject measures and the association with postoperative in-fections. JAMA 2010;303:2479–85.

32. Hawn MT, Itani KM, Gray SH, Vick CC, Henderson W,Houston TK. Association of timely administration of pro-phylactic antibiotics for major surgical procedures and sur-gical site infection. J Am Coll Surg 2008;206:814–21.