Journal of Pediatric Surgery 51 (2016) 154158

Contents lists available at ScienceDirect

Journal of Pediatric Surgery

journal homepage: www.elsevier.com/locate/jpedsurg

Risk factors and management of Nuss bar infections in 1717 patients

over 25 years
Robert J. Obermeyer a,b,, Erin Godbout b, Michael J. Goretsky a,b, James F. Paulson c, Frazier W. Frantz a,b,
M. Ann Kuhn a,b, Michele L. Lombardo a,b, E. Stephen Buescher a,b, Ashley Deyerle a, Robert E. Kelly Jr. a,b
a
b
c

Children's Hospital of The King's Daughters, Norfolk, VA, USA

Eastern Virginia Medical School, Norfolk, VA, USA
Old Dominion University, Norfolk, VA, USA

a r t i c l e

i n f o

Article history:
Received 2 October 2015
Accepted 9 October 2015
Key words:
Pectus
Excavatum
Nuss
Infection

a b s t r a c t
Purpose: An increase in postoperative infections after Nuss procedures led us to seek risks and review management. We report potential risk factors and make inferences for prevention of infections.
Methods: An IRB-approved retrospective chart review was used to evaluate demographic, clinical, surgical, and
postoperative variables of patients operated on between 10/1/2005 and 6/30/2013. Those with postoperative infection were evaluated for infection characteristics, management, and outcomes with univariate analyses.
Results: Over this 8-year period (20052013), 3.5% (30) of 854 patients developed cellulitis or infection, signicantly more than 1.5% (13) in our previous report of 863 patients, 19872005 (p = .007). The most frequent organism cultured was methicillin-sensitive Staphylococcus aureus. Patients who were given clindamycin
preoperatively (5 of 26 patients) had higher infection rates than those who received cefazolin (25 of 828)
(19% vs 3%, p b .001). Patients treated with a peri-incisional ON-Q (I-Flow, Kimberly-Clark, Irvine, CA) also had
higher infection rates (8.3% vs 2.4%, p b .001). Of the 30 patients who developed an infection, eighteen (60%)
with cellulitis or supercial infections did not require surgical treatment or early bar removal. The other twelve
patients (40%) with deep hardware infections required an average of 2.2 operations (range 16), with 3 (25%)
requiring removal of their stabilizer and 3 (25%) requiring early bar removal. None of these three patients experienced recurrence of pectus excavatum at 2 to 4 years of follow-up.
Conclusion: Preoperative antibiotic selection and use of ON-Q's may inuence infection rates after Nuss repair.
Nuss bars could be preserved in 90% of all patients with an infection and even 75% of those with a deep hardware
infection. Attempts to retain the bar when an infection occurs may help prevent pectus excavatum recurrence.
Level of Evidence = III.
2016 Elsevier Inc. All rights reserved.

1. Background/Purpose
Pectus excavatum (PE) is the most common chest wall deformity
and has an incidence of 1 in 400 births (0.25%). Minimally invasive repair of pectus excavatum (MIRPE) was initially reported in 1998 by
Nuss et al. [13] and has become the most common method of surgical
treatment. This procedure has undergone several modications over
the last two decades, but all patients ultimately require placement of a

Authors' role: Obermeyer, Godbout, Goretsky, Frantz, Kuhn, Lombardo, Kelly chart
review and acquisition of data for analysis; Obermeyer, Goretsky, Paulson, Lombardo,
Buescher, Kelly drafting and revising manuscript; Obermeyer, Goretsky, Frantz, Kuhn,
Lombardo, Kelly provided and cared for study patients; Paulson statistical analysis.
Corresponding author at: Department of Surgery, Eastern Virginia Medical School,
Children's Hospital of The King's Daughters, 601 Children's Lane, Suite 5B, Norfolk, VA
23507, USA. Tel.: +1 757 668 7703; fax: +1 757 668 8860.
E-mail address: robert.obermeyer@chkd.org (R.J. Obermeyer).
http://dx.doi.org/10.1016/j.jpedsurg.2015.10.036
0022-3468/ 2016 Elsevier Inc. All rights reserved.

substernal metal bar to elevate the sternum as originally described. Infectious complications after any surgical procedure are problematic,
but they are particularly vexing when they occur after orthopedic hardware is implanted. Furthermore, hardware removal often signicantly
compromises the outcome of the intended treatment outcome. The accepted rate of infection in surgical cases with a clean wound class is reported to be 1%5% while more recent reports demonstrate an
incidence of 2.6% [4]. The specic infection rate in the literature after
the Nuss procedure is reported to be 1.5%6.9% [5]. A perceived increase
in postoperative infections after the Nuss procedure at our institution
led us to evaluate our actual rate of infection and seek potential risk factors. In recent literature, preservation of infected orthopedic hardware
has been successful in 68% of these cases [6]. The success rate of Nuss
bar preservation after a deep bar infection is difcult to extract from
other reports. The literature has wide variations in bar preservation
rates ranging from 25% to 80% [5,79] (Table 1). Inquiries from other
surgeons about bar infections over the years led us to review our management and to attempt to provide guidance in these difcult cases.

R.J. Obermeyer et al. / Journal of Pediatric Surgery 51 (2016) 154158

Table 1
Bar preservation after deep bar infection.
Author

Year

Preserved

Calkins
Van Renterghem
Shin
Tanaka
Obermeyer

2005
2005
2007
2012
2014

80%
50%
50%
25%
75%

2. Methods
An IRB approved protocol (IRB #14-03-WC-0034) was obtained to
perform a retrospective chart review on patients who underwent minimally invasive repair of PE and subsequently developed an infection
between 10/1/2005 and 6/30/2013. The chart review included evaluation of demographic, clinical, surgical and postoperative variables of patients. We dened an infection as being any case treated with antibiotics
for cellulitis, supercial infection with active drainage, or deep infection
involving hardware, regardless of the time of presentation. We also report our postoperative infection rate as dened by the CDC (Centers for
Disease Control and Prevention) National Healthcare Safety Network
(NHSN), which only includes infections with active drainage or hardware involvement that occur within 90 days of surgical orthopedic
hardware placement. Descriptive statistics, frequencies, Fisher's Exact
test, chi-square analysis, logistic regression, and general linear models
were used to evaluate risk factors, infection characteristics, management, and outcomes among patients who developed postoperative infections. All analyses were conducted using SPSS version 19 statistical
software. Finally, the type of infections and bar preservation rate after
1717 primary and redo Nuss repairs for pectus excavatum in 854 patients over our recent experience (10/20056/2013) was compared to
our initial report of 863 patients over eighteen years (1/1987-/2005).
3. Results
3.1. Incidence
In over 25 years, we have experienced an overall infection rate of
2.5% (43 of 1717 patients), but in the last 8 years (20052013), the infection rate increased to 3.5% (30 of 854 patients). This rate is signicantly higher than the 1.5% (13 of 863 patients) rate observed in our
previous report from 1987 to 2005 (2(5.96) = 2.15, p = .007). The infection characteristics were the similar between the two different time
periods. Both the initial and recent time period demonstrated a higher
percentage of deep bar infections (Table 2). When using the CDC
NHSH denition of a postoperative infection, our infection rate was 2%
(17 out of 854) in the last 8 years, which is below reported rates of infection for clean cases in over 600,000 cases reviewed by Ortega et al.
[4] using the ACS-NSQIP database. Using this same denition the infection rate in our initial report was 0.7% (6 out of 863) and our overall incidence over 25 years is 1.3% (23 out of 1717).
3.2. Potential risk factors
In the recent time period, patients who were given clindamycin preoperatively (5 of 26 patients) had higher infection rates than those who
received cefazolin (25 of 828) (19% vs 3%, (2(1) = 15.12, p b .001). A

155

total of 156 patients were treated with a peri-incisional ON-Q (I-Flow,

Kimberly-Clark, Irvine, CA) at our facility. Thirteen of those patients
(8.3%) developed an infection which is signicantly higher than the infection rate observed in patients who did not receive an ON-Q (2.4%);
( 2(1) = 10.37, p b .001). Multivariate analysis was not performed
but only two patients got both clindamycin preoperatively and an ONQ postoperatively. After January 1, 2010, our protocol regarding duration of postoperative antibiotics changed from 72 h or more to 48 h or
less. Patients who received 72 h of postoperative antibiotics had a
lower infection rate (11 of 404) than patients with 48 h (19 of 450)
but this was not statistically different (2.7% vs 4.2%, 2 (1) = 0.925,
p = .34). There was one case of Clostridium difcile diarrhea in a patient
who received 24 h of antibiotics. Recently (May 2012), in an effort to
lower our infection rate we began using Sage applicator cloths (2%
chlorhexidine gluconate, Sage Products, Cary, IL) as part of our patient's
preoperative preparation. We currently instruct the patients to use the
cloths the day before surgery and immediately before going back to
the operating suite. Since initiating this, our rate of infection decreased
to 2.1% (3 of 144) compared to 3.8% (27 of 710). Although this addition
indicates a promising trend, it is not statistically signicant ( 2(1) =
2.15, p = .143). To date, our infection rate with the use of chlorhexidine/alcohol skin antiseptic (ChloraPrep, chlorhexidine gluconate
(CHG) 2% w/v and isopropyl alcohol (IPA) 70% v/v, CareFusion, San
Diego, CA) is 1.6% (1 of 63) while the infection rate with betadine is
3.7% (29 of 791). This variable also did not reach statistical signicance
(Fisher's Exact = 0.718) and the low number of patients in the CHG
group makes the analysis suboptimal (Table 3). Of note, the only patient
that developed a wound infection after both preoperative skin preparation with CHG cloths and intraoperative CHG/IPA skin preparation also
received IV clindamycin as their antibiotic prophylaxis.
3.3. Patient characteristics
All patients with an infection presented a median of 33.5 days
(range: 8595 days) after the initial operation. Presenting characteristics included erythema (73%), drainage (70%), swelling (43%), pain
(43%), and fever (33%). All patients were male with an average age of
16.7 years. Two bars were placed in 16 (53%) of these patients and
two (6.7%) required titanium hardware.
3.4. Cellulitis/supercial wound infections
Eighteen patients developed infections we classied as either cellulitis or supercial open wounds. These patients presented at a median of
34 days but ranged from 8 to 595 days after surgery. All bars were preserved with antibiotic treatment and conservative wound management. Intravenous (IV) antibiotics were used in 10 of the 18 patients
(55.6%) for an average of 15.5 days (260 days) with the most common
antibiotic being clindamycin followed by cefazolin. A peripherally
inserted central catheter (PICC) line was used in 28% (n = 5) of these
cases. All patients were treated with oral antibiotics for a variable
amount of time ranging from 10 days to over a year. The most common
oral antibiotics used included cephalexin, clindamycin, and sulfamethoxazole/trimethoprim (SMZTMP). Some patients were treated at outside facilities so we were unable to capture all antibiotic regimens. None
of these patients required drainage in the operating room. Wound cultures were not done in all of these patients because some had no drainage to culture. Positive culture results were documented in only three

Table 2
Infection characteristics.
Time period

Total repairs

Total infections

Cellulitis

Supercial infection

Deep infection

18 years (1/19879/2005)
8 years (10/20056/2013)
Totals

863
854
1717

13 (1.5%)
30 (3.5%)
43 (2.5%)

4 (31%)
8 (27%)
12 (28%)

3 (23%)
10 (33%)
13 (30%)

6 (46%)
12 (40%)
18 (42%)

156

R.J. Obermeyer et al. / Journal of Pediatric Surgery 51 (2016) 154158

Table 3
Incidence of infection vs. potential risk factors.

Table 5
Infection prevention bundle.

Variable

Used

Not used

CHG cloths
Preoperative clindamycin
CHG/IPA prep
Postop antibiotics N72 h
Peri-Incisional On-Q

2.1%
19.0%
1.6%
2.7%
8.3%

3.8%
3.0%
3.7%
4.2%
2.4%

.143
b.001
.718a
.23
b.001

Preoperative
precautions
Intraoperative
precautions

Fisher's Exact.

cases (17%) and 2 of these were coagulase-negative Staphylococcus

(CoNS) while one was methicillin sensitive Staphylococcus aureus
(MSSA).

Operative
technique
Postoperative
precautions

Metal allergy skin testing

Chlorhexidine wipes the day before and day of surgery
Chlorhexidine/alcohol skin preparation
Cefazolin IV (30 mg/kg up to 2000 mg)
Give antibiotic dose within 30 min of incision and repeat every
4h
Penicillin allergic patients (non-anaphylactic): test dose cefazolin
while under anesthesia
First generation cephalosporin allergic patients: clindamycin or
vancomycin (3060 min infusion time)
Vancomycin in patients with MRSA colonization or in facilities
with recent outbreaks
Sterile technique
Three layered wound closure
Avoid use of ON-Q catheters near incision (Peri-incisional)
Continue IV antibiotics for up to 48 h postoperatively

3.5. Deep bar infections

5. Infection prevention
Twelve patients developed a deep infection. These patients presented a median of 33 days after surgery but ranged from 16 to 118 days.
Positive culture results were obtained in nine patients: 6 were MSSA
and one was methicillin resistant S. aureus (MRSA). Among patients
with deep wound infections, 50% (n = 6) were managed with a PICC
line. Compared to the patients with cellulitis or supercial wound infections, the average duration of IV antibiotics was longer at 24 days
(188 days) while the average duration of oral antibiotic therapy was
shorter at 120 days (1 to 188 days). Patients with deep bar infections required an average of 2.2 operations (range 16) with 25% (n = 3) requiring removal of one or two stabilizers. Only 3 patients in the recent
series required early bar removal and all were associated with a deep
bar infection. This resulted in a bar preservation rate of 90% for all infections and 75% for deep bar infections. All three patients with a deep bar
infection had two bars placed for correction and, as of this writing, none
have developed a recurrence at 24 years post-bar removal. The earliest
bar removal occurred at 4 months and only required removal of a lower
bar since the infection did not communicate with the upper bar. Another patient developed a metal allergy after a negative preoperative metal
allergy skin test. This case resulted in a complicated medical course and
ultimately the patient requested removal of his bars 6 months after insertion. The nal patient requested removal of both bars at 19 months
because of ongoing discomfort and chronic wound drainage (Table 4).

4. Conclusions
Accepted incidences of infections depend on the surgical wound
class and have been established as follows: clean = 1%5%, clean
contaminated = 3%11%, contaminated = 10%17%, and dirty N 27%.
More recent literature evaluating the ACS-NSQIP database found average rates of infection in these four categories to be 2.6%, 6.7%, 8.6%,
and 11.8% respectively [4]. The Nuss repair wound classication is
clean; therefore, our overall infection incidence of 1.3% using the CDC
NHSN classication is acceptably below the ACS-NSQIP reported average. After identifying a spike in our infection rate we decided to seek potential risk factors and develop an infection prevention bundle
described below (Table 5).

5.1. Type of antibiotic prophylaxis

Patients that are reported to have a penicillin allergy are often given
a preoperative dose of IV clindamycin as an accepted alternative [10]. In
our study, patients given preoperative clindamycin had a statistically
signicant higher infection rate. This may be caused by the fact that
clindamycin is predominately bacteriostatic with slower bactericidal effect while cefazolin is known to be rapidly bactericidal [11]. Even when
a patient has a known allergy to penicillin, some of our surgeons
proceeded with a test dose of cefazolin while under anesthesia. No patients had a cross-allergy reaction, which is historically cited as being
as high as 10%. Part of the reason for this high estimate is that up until
1982, pencillins and cephalosporins were both commercially
manufactured using Acremonium fungus which was previously called
Cephalosporium. More recent literature reviews suggest that this crossreactivity occurs in only 1% of patients treated with all rst-generation
cephalosporins and only 2.6% of those with a conrmed penicillin allergy. Since cefazolin has a side chain that is dissimilar to penicillin, amoxicillin, and ampicillin this makes cross-reactivity less likely. Patients
with anaphylactic type reactions to penicillin are prone to multiple
drug allergies [12,13]. Given these ndings, we recommend considering
an intraoperative test dose of cefazolin when a patient has a known
nonanaphylactic allergy to penicillin. Although we have not had a patient have an anaphylactic reaction to cefazolin in any of these circumstances, it would be safer and easier to manage the reaction under
general anesthesia. When patients have a known anaphylactic type reaction to penicillin or an allergy to rst-generation cephalosporins,
clindamycin should be used. Vancomycin may be considered when patients are colonized with MRSA or in facilities with recent MRSA outbreaks [14].

5.2. Duration of postoperative antibiotics

The duration of postoperative IV antibiotic therapy in clean cases involving implanted hardware is recommended to not exceed 24 h [15].
The CDC supports this recommendation with the caveat that patients

Table 4
Bar preservation with deep infections.
Time period

Deep infections

Bar preservation

Duration bars in place (months)

Patients with recurrence

18 years (1/19879/2005)
8 years (10/20056/2013)

6
12

3 (50%)
9 (75%)

3, 18, 18
4, 6, 19

1
0

R.J. Obermeyer et al. / Journal of Pediatric Surgery 51 (2016) 154158

undergoing cardiac surgery can continue postoperative IV antibiotics for

up to 48 h [16]. With the knowledge that the Nuss bar is near the heart
and involves dissection of the pericardium with reported cases of pericarditis, our group decreased our initial postoperative use of IV antibiotics from more than 72 h to 48 h or less. Our analysis showed that
there was no difference in infection rates when antibiotics were given
longer. The current recommendation is that prophylactic antibiotics
for the Nuss procedure not extend beyond 48 h of therapy.
5.3. Preoperative skin preparation
The use of preoperative CHG cloths has been demonstrated to decrease infection rates in the orthopedic literature after total hip
arthroplasty from 1.6 to 0% by Johnson et al. [17] in over 1134 patients.
In his second report, the infection rate was 0.5% with the use of CHG
clothes compared to 1.7% with standard in-hospital bathing in over
2458 patients [18]. Our experience with preoperative CHG cloths demonstrated a trend toward lowering infections, but did not reach statistical signicance at the time of this report. Some of our patients had skin
irritation after application and could not tolerate the CHG cloths but
otherwise there were no adverse effects. The perioperative use of CHG
cloths is recommended given the support of similar literature.
5.4. Intraoperative skin preparation
Our data show a trend toward lower infection rates with intraoperative CHG/IPA skin preps and other groups have indicated an infection
rate as low as 0.7% when using CHG scrub after Nuss bar insertion [19].
Chlorhexidine has been shown to decrease positive skin cultures more effectively than iodine in a systematic review [20], and in a recent Cochrane
review CHG/IPA has been shown to be superior to povidoneiodine for
preventing postoperative infections in clean cases [21]. These alcohol
based skin preps require special care to prevent intraoperative res and
burns. Placing dry towels around the chest during the prep with subsequent removal and ensuring the prep is dry prior to the use of cautery
are essential in this effort. At our institution CHG/IPA surgical skin preparation is recommended, at the discretion of the surgeon, as a method to
lower infection rates for this and other appropriate clean cases.
5.5. ON-Q catheters
ON-Q catheters typically secrete over 30 mL per day of local anesthesia around and potentially into the surgical wound bed depending on
how they are positioned. The catheters communicate through the skin
for up to 4 days which may allow for colonization. As a result of the statistically signicant increase in postoperative infections when using the

Table 6
Infection treatment bundle.
Controlling gross
infection

Gross infection
controlled

Antibiotic therapy

Drain purulence and culture

Irrigate consider pulse antibiotic irrigation
Debridement
Remove all nonabsorbable suture in the eld
If gross pus pack wet-to-dry and irrigate in operating room
daily
If wound clean but cannot be closed use Wound Vac every
23 days or daily wet-to-dry
If wound can be closed attempt layered closure with nylon
skin suture
Close follow-up
Initial IV antibiotics for 12 weeks (consider PICC line)
Modify antibiotics per culture results
Consider repeat allergy testing if not improving
Home antibiotics PO clindamycin and rifampin (CBC and LFT
q 3 weeks)
After 12 months if CBC, ESR, and CRP return to normal then
stop clindamycin and rifampin
Consider long-term trimethoprimsulfamethoxazole

157

ON-Q catheters around the incisions, this method of postoperative pain

control was largely discontinued in our practice. It is recommended to
place the ON-Q catheters away from the thoracic incisions when used.
5.6. Allergy testing
One patient in this cohort developed an allergy after a negative preoperative skin metal allergy test, which has been previously reported
[22]. Although this scenario is rare, preoperative skin metal allergy
test is recommended to prevent this avoidable complication since
chronic skin irritation has the potential to compromise the skin integrity
and result in a deep bar infection [23].
6. Managing infections
Our treatment of cellulitis, supercial infections, and deep bar infections has varied, and while no denite therapy can be endorsed, certain
strategies can be recommended based on our experience. All patients
with signs of an infection should have close follow-up to monitor
their progress. The development of a rash should prompt consideration
to repeat metal allergy skin testing even if preoperative testing was negative. Patients with cellulitis or supercial wound infections can almost
always be managed with a combination of intravenous and oral antibiotics as well as conservative local wound care as needed. This may include long-term suppressive oral therapy, typically with SMZTMP or
clindamycin. Patients with deep bar infections often required a longer
course of IV antibiotics; PICC lines should be considered since they provide excellent access with a low complication prole. We typically used
IV clindamycin followed by several months of oral SMZTMP. More recently, we have started using oral clindamycin and rifampin for 1
2 months, then switch to oral SMZTMP. Although clindamycin was
found to be a poor preoperative antibiotic in our study, it may still be appropriate treatment for a postoperative infection because of excellent
oral bioavailability and tissue absorption, including bone. Rifampin is
often used in orthopedic hardware infections because of its unique ability to penetrate the biolm formed by Staphylococcus [24]. It is helpful to
explained to the patient and family that treatment of a deep bar infection may require multiple courses of antibiotics and potentially multiple
trips to the operating room. They should also understand that removal
of the stabilizer and even the bar may be required if the infection cannot
be controlled. Initial surgical treatment should focus on drainage with
cultures, irrigation, wound capsule debridement, removal of nonabsorbable braided suture (Fiberwire, Arthrex, Inc.), and wet-to-dry packing or wound vac. Serial dressing changes are frequently required.
Following control of the gross infection, delayed closure may be performed with absorbable suture in the deep tissue and nonabsorbable
loose mattress sutures in the skin (Table 6). Although our recent experience of bar preservation with cellulitis or supercial infections is 100%
and 75% after deep bar infections, these percentages should only be
used as guidelines when counseling families because every case is
unique and our limited number of infections makes predictions of outcome impossible.
7. Final statements
Strategies to lower Nuss bar infections may include perioperative
and intraoperative CHG skin preparations, preferential use of cefazolin
over clindamycin for antibiotic prophylaxis when appropriate, and
avoidance of peri-incisional ON-Q catheters for postoperative pain control. Cellulitis and supercial infections are unlikely to necessitate bar
removal and even deep bar infections will infrequently require bar removal with aggressive antibiotic and surgical management. The optimal
amount of time to attempt bar preservation in the face of a Nuss bar infection is not known, but a longer period of bar preservation should help
reduce pectus excavatum recurrence.

158

R.J. Obermeyer et al. / Journal of Pediatric Surgery 51 (2016) 154158

Appendix A. Discussions
Presented by Robert Obermeyer, Norfolk VA
JENNIFER ALDRINK (Columbus, OH) What was the antibiotic of choice
in the rst cohort of patients?
ROBERT OBERMEYER The early experience? I think it was probably very
similar. I think we used Ancef and clindamycin when appropriate, when patients had an allergy. I will mention that the
clindamycin use amongst our group is variable, so when somebody has a noted allergy to penicillin many people go right to
clindamycin. I think most of us know the number of 10% as
the crossreactivity so I am going to stay away from Ancef because they had a listed allergy to penicillin, but when you review the literature a little bit more deeply you discover that
actually Ancef is quite dissimilar from penicillin. In fact, of the
cephalosporins, its side chain is very dissimilar and the
crossreactivity rate is probably only about 3%. I have been
very liberal with Ancef in those cases. I give a test dose in the
operating room when they're asleep. What better place to
have an allergic reaction then under that kind of control and
so far I have had none. I think Ancef is a bactericidal and clinda
is known to be bacteriostatic, so there may be some benets.
DAVID LANNING (Richmond, VA) Robert, nice presentation. On these
patients that have had an infection, have you removed any
of those bars? If so, has it been similar to those that have
not been infected or is it more difcult?
ROBERT OBERMEYER Have we removed bars? I'm sorry?
DAVID LANNING That have had deep infections.
ROBERT OBERMEYER Yes. You mean later removal?
DAVID LANNING When they've been removed, how has that been?
ROBERT OBERMEYER No different that I can detect. I think it's interesting that some bars are very easy to remove and some are very
hard to remove, and I'm not sure that plays a role in the difculty of removal. I think some people form more calcium for
some reason, their biochemical makeup, and other people
I've had people come to me ve years later to have their
bars removed. They've been bad patients and they slide out
easily, so I don't think that has played a role. Thank you for
the question.
DAVID NOTRICA (Phoenix, AZ) Bob, great presentation, thank you. I
know that there has been a strong trend towards using
more bars over time and so is any of this component of increased bar infections related to increased number of bars
being implanted per patient?
ROBERT OBERMEYER Thank you for the question, Dave. I don't know.
In this cohort, we looked at the patient characteristics and
didn't have enough time but 100% of them were males and
about 50% had two bars, so it may be a coin ip. I'm not sure.
SHERIF EMIL (Montreal, QC) Thanks again for sharing this great series
with us. I've yet to deal with one of those and I was just telling
someone I haven't done enough yet obviously. I am just curious if you looked at any patient-associated factors. One of my
colleagues actually just had a patient with a bar infection and
the young man had acne and with that was actually a very
predictable bacteria that often colonizes patients with acne
for example and that is not an unusual problem in a lot of
teenagers. I am just curious if there are any patientassociated factors that you could sort of predict ahead of
time that might increase the risk.
ROBERT OBERMEYER Thank you for the question. I don't know. Congratulations on your excellent series. Part of the reason that I
wanted to present this work is there was recent literature
that came out that suggested that all deep bar infections require bar removal and why I present this to you and hopefully
we can get it in the literature is that is not the case, that you

have hope and you can give your patients hope in these difcult situations and we can't give the lawyer hope.
Propionibacterium only grew in one of our patients. The majority of the bacteria we grew was MSSA, so I can't detect
any other risk factors. We did have one MRSA patient but if
you're a MRSA carrier there is the whole thing of do you
give that patient vancomycin or clindamycin instead of
Ancef. I'm not sure.
ROBERT GATES (Greenville, SC) We have had two infections in the last
ve years. I drilled down and looked at the OR records on
those two patients and noted that we had an increasing number of personnel in and out: scrub techs changing out, circulators, anesthesiologists, anesthetists coming in and out. I noted
in the last three years when I had really drilled down and got
rid of the turnover of the OR personnel that we've not had any
infections. I think it really was related to over 20 people in and
out of the OR on those two cases.
ROBERT OBERMEYER Thank you for that comment.
References
[1] Nuss D, Kelly Jr RE, Croitoru DP, et al. A 10-year review of a minimally invasive technique for the correction of pectus excavatum. J Pediatr Surg 1998;33:54552.
[2] Fonkalsrud EW. Current management of pectus excavatum. World J Surg 2003;
27(5):5028.
[3] Kelly RE, Shamberger RC, Mellins RB, et al. Prospective multicenter study of surgical
correction of pectus excavatum: design, perioperative complications, pain, and baseline pulmonary function facilitated by internet-based data collection. J Am Coll Surg
2007;205(2):20516.
[4] Ortega G, Rhee DS, Papandria DJ, et al. An evaluation of surgical site infections by
wound classication system using the ACS-NSQIP. J Surg Res 2012;174(1):338.
http://dx.doi.org/10.1016/j.jss.2011.05.056 [Epub 2011 Jun 24].
[5] Tanaka K, Kuwashima N, Ashizuka S, et al. Risk factors of infections of implanted device after the Nuss procedure. Pediatr Surg Int 2012;28:8736.
[6] Rightmire E, Zurakowski D, Vrahas M. Acute infections after fracture repair: management with hardware in place. Clin Orthop Relat Res 2008;466(2):46672.
[7] Van Renterghem KM, Von Bismarck S, Bax NMA, et al. Should an infected Nuss bar
be removed? J Pediatr Surg 2005;40:6703.
[8] Shin S, Goretsky MJ, Kelly Jr RE, et al. Infectious complications after the Nuss repair in
a series of 863 patients. J Pediatr Surg 2007;42:8792.
[9] Calkins CM, Shew SB, Sharp RJ, et al. Management of postoperative infections after
the minimally invasive pectus excavatum repair. J Pediatr Surg 2005;40(6):10048.
[10] Wheeless III Clifford R. Wheeless' textbook of orthopaedics. online reference http://
www.wheelessonline.com; 2012.
[11] Lincosamides, oxazolidinones, and streptogramins. Merck manual of diagnosis and
therapy. Merck & Co.; 2005
[12] Campangna JD, Bond MC, Schabelman E, et al. The use of cephalosporins in
penicillin-allergic patients: a literature review. J Emerg Med 2012;42(5):61220.
[13] Pichichero ME. Cephalosporins can be prescribed safely for penicillin-allergic patients. J Fam Pract 2006;55(2):10612.
[14] American Society of Health-System Pharmacists ASHP. Therapeutic guidelines on
antimicrobial prophylaxis in surgery. American Society of Health-System Pharmacists. Am J Health Syst Pharm 1999;56(18):183988.
[15] American Academy of Orthopedic Surgeons. Information statement, recommendations for the use of intravenous antibiotic prophylaxis in primary total joint
arthroplasty; 2014.
[16] Fry DE. Surgical site infections and the Surgical Care Improvement Project (SCIP):
evolution of national quality measures. Surg Infect 2008;9(6):57984.
[17] Johnson AJ, Daley JA, Zywiel MG, et al. Preoperative chlorhexidine preparation and
the incidence of surgical site infections after hip arthroplasty. J Arthroplasty 2010;
25(6 Suppl.):98102.
[18] Kapadia BH, Johnson AJ, Daley JA, et al. Pre-admission cutaneous chlorhexidine
preparation reduces surgical site infections in total hip arthroplasty. J Arthroplasty
2013;28(3):4903.
[19] Sacco Casamassima MG, Goldstein SD, Salazar JH, et al. Perioperative strategies and
technical modications to the Nuss repair for pectus excavatum in pediatric patients:
a large volume, single institution experience. J Pediatr Surg 2014;49(4):57582.
[20] Lee I, Agarwal RK, Lee BY, et al. Systematic review and cost analysis comparing use of chlorhexidine with use of iodine for preoperative skin antisepsis to prevent surgical site infection. Infect Control Hosp Epidemiol 2010;21(12). http://dx.doi.org/10.1086/657134.
[21] Dumville JC, McFarlane E, Edwards P, et al. Preoperative skin antiseptics for preventing
surgical wound infections after clean surgery. Cochrane Database Syst Rev 2013;3:159.
[22] Sesia SB, Haecker FM, Shah B, et al. Development of metal allergy after Nuss procedure for repair of pectus excavatum despite preoperative negative skin test. J Pediatr
Case Rep I 2013:1525.
[23] Shah B, Cohee A, Deyerle A, et al. High rates of metal allergy amongst Nuss procedure patients dictate broader pre-operative testing. J Pediatr Surg 2014;49(3):
4514.
[24] Zheng Z, Stewart PS. Penetration of rifampin through Staphlococcus epidermidis
biolms. Antimicrob Agents Chemother 2002;46(3):9003.