ORIGINAL ARTICLE

Increasing Compliance With an Antibiotic Prophylaxis Guideline

to Prevent Pediatric Surgical Site Infection
Before and After Study
Jeannette P. So, MSc,∗ Ilyas S. Aleem, MD,† Derek S. Tsang, MD,‡ Anne G. Matlow, MD, MSc, FRCPC,§
and James G. Wright, MD, MPH, FRCSC∗ ¶; for The SickKids Surgical Site Infection Task Force

Conclusions: A multifaceted intervention improved compliance with a pedi-

Objectives: To evaluate an intervention for improving antibiotic prophylaxis
atric AP guideline.
(AP) guideline compliance to prevent surgical site infections in children.
Background: Although appropriate AP reduces surgical site infection, and Keywords: antibiotic prophylaxis, behavioral change, guideline compliance,
guidelines improve quality of care, changing practice is difficult. To facilitate intervention, pediatric surgery
behavioral change, various barriers need to be addressed.
(Ann Surg 2015;262:403–408)
Methods: A multidisciplinary task force at a pediatric hospital developed an
evidence-based AP guideline. Subsequently, the guideline was posted in op-
erating rooms and the online formulary, only recommended antibiotics were
available in operating rooms, incoming trainees received orientation, antibi-
otic verification was included in time-out, computerized alerts were set for
S urgical site infection (SSI) is a serious adverse event leading
to significant patient morbidity, mortality, and higher costs.1–3
Approximately 50% of SSIs are estimated to be preventable,3 and ap-
inappropriate postoperative prophylaxis, and surgeons received e-mails when
propriate use of antibiotic prophylaxis (AP), including correct drug,
guideline was not followed. AP indication and administration were docu-
dosage, timing, and redosing, is a key intervention to reduce the
mented for surgical procedures in July 2008 (preintervention), September
incidence of SSI.4–7 Practice guidelines that assist physicians in de-
2011 (postintervention), and April–May 2013 (follow-up). Compliance was
cision making are a potential strategy to facilitate appropriate use
defined as complete—appropriate antibiotic, dose, timing, redosing, and du-
of prophylactic antibiotics.8,9 However, despite availability of var-
ration when prophylaxis was indicated; partial—appropriate drug and timing
ious guidelines,4–5,10 prior reports suggest that clinician’s behavior
when prophylaxis was indicated; and appropriate use—complete compliance
has changed little and compliance remains suboptimal,11,12 partic-
when prophylaxis was indicated, no antibiotics when not indicated. Compli-
ularly in pediatric surgery in which the benefits of antibiotics in
ance at preintervention and follow-up was compared using χ 2 tests.
children and adolescents are uncertain.13,14 In a 5-year retrospec-
Results: AP was indicated in 43.9% (187/426) and 62.0% (124/200) of sur-
tive study of general surgery and urology in 22 children’s hospi-
gical procedures at preintervention and follow-up, respectively. There were
tals, only 82% of patients received antibiotics when prophylaxis
significant improvements in appropriate antibiotic use (51.6%–67.0%; P <
was indicated, and 40% of patients were given antibiotics without
0.001), complete (26.2%–53.2%; P < 0.001) and partial compliance (73.3%–
indication.13 A pediatric surgical study in Singapore reported that
88.7%, P = 0.001), correct dosage (77.5%–90.7%; P = 0.003), timing
57% of patients had appropriate antibiotics, 76% timely administra-
(83.3%–95.8%; P = 0.001), redosing (62.5%–95.8%, P = 0.003), and du-
tion, and 41% appropriate duration, with 13% overall compliance.14
ration (47.1%–65.3%; P < 0.002).
Although comparison among studies is difficult because definitions
of compliance differed, few of these studies provided explicit strate-
gies for implementation, which is essential for effective uptake of
guideline recommendations.15 Pathman et al16,17 proposed a 4-stage
model to understanding utilization of clinical practice guidelines:
From the ∗ Department of Surgery, The Hospital for Sick Children, Toronto, awareness, agreement, adoption, and adherence. Various attitude-,
Ontario, Canada; †Division of Orthopaedic Surgery, McMaster University, knowledge-, and organizational-related factors facilitate or impede
Hamilton, Ontario, Canada; ‡Department of Radiation Oncology, University use of guidelines, including awareness, familiarity, agreement, out-
of Toronto, Toronto, Ontario, Canada; §Center for Patient Safety, University
of Toronto, Toronto, Ontario, Canada; and ¶Division of Orthopaedic Surgery,
come expectancy, self-efficacy, motivation or inertia of previous prac-
University of Toronto, Toronto, Ontario, Canada. tice, and patient, guideline, and environmental elements.18 In general,
Members of The SickKids Surgical Site Infection Task Force include the following: behavioral change requires multiple different implementation and
Zoran Bojic, MHSc, PMP, CLA, ASQ-CMQOE; Rita Damignani, MSc BScPT; dissemination strategies to address the different factors. Therefore, in
Annie Fecteau, MDCM, FRCP(C); Igor Luginbuehl, MD; Anne G. Matlow,
MD, MSc, FRCPC; Jean Paul Paraiso, MSc; Lisa Pendergast, RN; Debi Senger,
this study, we implemented and evaluated a multifaceted intervention
CHIM; Jeannette P. So, MSc; Laurie Streitenberger, BSc, RN, CIC; Kathryn for improving compliance with an AP practice guideline to prevent
Timberlake, PharmD; and James G. Wright, MD, MPH, FRCSC. SSI in pediatric patients.
Disclosure: Supported by funds from the Robert B. Salter Chair of Surgical Re-
search held by Dr. Wright. Ms. Timberlake has received consulting fees from
Pfizer for the antifungal voriconazole. For the remaining authors, no conflicts
of interest were declared.
Supplemental digital content is available for this article. Direct URL citation appears METHODS
in the printed text and is provided in the HTML and PDF versions of this article
on the journal’s Web site (www.annalsofsurgery.com). Setting
Reprints: James G. Wright, MD, MPH, FRCSC, The Hospital for Sick Children, The study was performed at The Hospital for Sick Children,
555 University Ave, Black Wing, Room 1254, Toronto, Ontario, M5G 1X8, a 278-bed academic pediatric hospital in Toronto, Canada, that per-
Canada. E-mail: james.wright@sickkids.ca.
Copyright C 2014 Wolters Kluwer Health, Inc. All rights reserved.
forms approximately 12,000 surgical procedures per year. Before
ISSN: 0003-4932/14/26202-0403 2006, the only reference for AP use was the hospital formulary, which
DOI: 10.1097/SLA.0000000000000934 contained incomplete guidelines for all specialties and procedures.

Annals of Surgery r Volume 262, Number 2, August 2015 www.annalsofsurgery.com | 403

Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
So et al Annals of Surgery r Volume 262, Number 2, August 2015

Multifaceted Intervention in these 2 specialties were almost never indicated for prophylaxis. We
The first step was to develop comprehensive evidence-based collected data on eligible surgical procedures performed in July 2008
guidelines. To raise awareness and secure input and support from (preintervention), a random sample of procedures performed in the
all relevant stakeholders, we formed a multidisciplinary task force first 2 weeks of September 2011 (postintervention) and a stratified
in May 2006 with representation from surgery, anesthesia, nurs- random sample of procedures performed in April and May 2013
ing, pharmacy, infection prevention and control, quality improvement (follow-up). For the postintervention and follow-up periods, cases
and patient safety, information technology, and health records. The were listed consecutively by date, and a random number genera-
task force adapted existing evidence-based guidelines,19–21 including tor was used to compile the samples. Cases with confirmed or sus-
those from the American Academy of Pediatrics,22 Centers for Dis- pected preoperative infection were excluded, including patients with
ease Control and Prevention,4 American Society of Health-System otitis media, cholesteatoma, ingrown toenail, perforated appendix,
Pharmacists,23 Infectious Diseases Society of America,10 Surgical In- abscess, sepsis, frank infection, and draining pus. We also excluded
fection Society,24 American Society for Gastrointestinal Endoscopy,25 procedures that involved the oral cavity, isolated endoscopy (includ-
and Surgical Infection Prevention Guideline Writers Workgroup,5 to ing foreign body removal), organ retrieval, and central venous line
create a pediatric surgical AP guideline (see Supplemental Digital removal, which had no risk of postoperative infection. From OR
Content 1 Appendix, available at http://links.lww.com/SLA/A639, logs, anesthesia records, and health records, we extracted patient and
which shows full guideline). The guideline encompassed all special- surgery characteristics and AP administration for each procedure. If
ties, contained both child- and neonate-specific recommendations, antibiotics were given, the drug, timing, dosage, redosing, and dura-
listed alternative regimens for patients with allergies, and set inter- tion were also collected. For procedures in which AP was not given in
vals for dosing and discontinuation. The guideline was subsequently the OR, we manually conducted a chart review to determine whether
reviewed and revised by all surgical disciplines and by pharmacy for patients received them before surgery, on the ward, or in the emer-
completeness and appropriateness. gency department. Physicians were also consulted for clarification
Once the guideline was approved by all disciplines, we initi- on any discrepancies between the OR log and the anesthesia record.
ated multiple implementation strategies. First, we collaborated with Where differences could not be reconciled, we utilized the data from
pharmacy to stock anesthesia carts with guideline-approved antibi- the anesthesia records.
otics and ensured that anesthesiologists were willing to administer For each surgical procedure, we determined whether antibiotics
antibiotics. Second, the guideline was posted on the online hospital were indicated. Moreover, for procedures with prophylaxis indicated
formulary. Third, copies of specialty-specific guidelines were posted and administered, we evaluated compliance with the 5 AP criteria
in the associated operating rooms (ORs). Fourth, AP verification (drug, timing, dosage, redosing, and duration) according to the guide-
was added to the surgical time-out checklist to be completed before line. The guideline contained a list of approved drugs for each type of
surgery. Fifth, we developed an educational intervention for incoming procedure. Patients already on antibiotics for a preexisting condition
residents and fellows. Beginning in July 2008, all new surgical and before surgery did not require additional prophylaxis if the antibiotic
anesthesia residents and fellows received the AP guideline in their spectrum covered the potential pathogens and they were not due for
orientation packages and attended a mandatory educational session redosing. Recommended dosage was calculated on the basis of pa-
given by the surgeon-in-chief. The presentation provided information tients’ age and weight, and doses within 20% of the reference were
on guideline access, reviewed the criteria and evidence for appropriate considered appropriate. The guideline specified that antibiotics be
AP use, and emphasized the importance of compliance for preventing given before applying the tourniquet, if one was required, and within
SSI. Sixth, the final phase of the quality improvement initiative be- 60 minutes before incision. Patients with β-lactam allergy or previous
ginning May 2009 involved informatics. We programmed a reminder methicillin-resistant Staphylococcus aureus colonization or infection
in the computerized patient order entry system that postoperative should have received vancomycin administered within 120 minutes
prophylaxis was not recommended for any but cardiac, transplant, before incision and infused over at least 60 minutes.5,22 According to
and cochlear implant surgical procedures. The electronic alert pro- the guideline, antibiotics were to be redosed every 2 half-lives during
vided a link to the online guideline and was triggered when clinicians anesthesia (“top-up”) to ensure adequate tissue antimicrobial levels
attempted to prescribe postoperative prophylaxis for nonindicated at wound closure.4 For surgical procedures involving excessive blood
patients. In addition, the task force mapped surgical prophylactic loss or extensive burns, intraoperative redosing to maintain antimicro-
indications to every procedural code in the hospital’s Surgical Infor- bial levels was to be considered. Patients already on antibiotics before
mation Systems (SIS, Alpharetta, GA), with input from the surgeons. surgery were redosed if “top-up” was due during the procedure. Pro-
SIS is a software program for managing and interfacing perioperative phylaxis should be discontinued at the end of all procedures, including
information and workflow, including scheduling and intraoperative for patients with open wounds, indwelling catheters, or in situ drains
patient tracking. Details of surgical patients and procedures were all or chest tubes. The exceptions for postoperative prophylaxis discon-
documented in SIS. OR nurses documented AP administration in tinuation were cardiac surgery, transplant, and cochlear implantation,
the electronic OR log. Information technology programmed the SIS which AP was extended for up to 48 hours postoperation.4–5,23
to automatically review all surgical procedures performed and sent We defined complete compliance with the AP guideline as
e-mail notifications to surgeons the next day if a patient was given having correct antibiotic, dosage, timing, redosing, and duration when
antibiotics without indication, not given antibiotic when indicated, or prophylaxis was indicated. Partial compliance was defined as correct
not given antibiotics within 60 minutes before skin incision when AP drug and timing when prophylaxis was indicated. Appropriate use
was indicated. Finally, aggregate data from the SIS was provided to of AP was defined as complete compliance when prophylaxis was
all divisions and departments on a monthly basis. indicated, and antibiotics not given when not indicated.

Statistical Analysis
Data Collection We assessed preintervention and follow-up compliance to eval-
To examine the effects of the intervention on guideline com- uate practice changes. Postintervention compliance was reported to
pliance, we assessed AP use in cardiovascular, general surgery, neu- determine trends but was not included in the statistical analyses.
rosurgery, orthopedic, otolaryngology, plastics, and urology proce- Using an estimated baseline rate of 45% complete compliance, we
dures. We excluded ophthalmology and dentistry because antibiotics required a sample size of 173 cases in each period to detect at least

404 | www.annalsofsurgery.com 
C 2014 Wolters Kluwer Health, Inc. All rights reserved.

Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 262, Number 2, August 2015 Surgical Antibiotic Guideline Compliance

15% change at 80% power and 5% significance level. We used the χ 2

TABLE 2. Surgical Cases by Antibiotic Prophylaxis
test with P < 0.05 (2-tailed) level of statistical significance to com-
Indication and Administration
pare compliance rates and dichotomous variables and 1-way analysis
of variance at P < 0.05 (2-tailed) significance to compare continu- Antibiotics Given
ous variables. Missing values were not imputed. All analyses were
Antibiotics Indicated Yes No Total
conducted in SPSS 20 for Mac (IBM Corporation, Armonk, NY).
Institutional ethics approval was obtained before study initiation. Preintervention (%)
Yes 174 (93.0) 13 (7.0) 187
RESULTS No 68 (28.5) 171 (71.5) 239
We reviewed 773 procedures preintervention, 192 procedures Total 242 184 426
Postintervention (%)
postintervention, and 220 procedures at follow-up. After excluding Yes 98 (94.2) 6 (5.8) 104
patients, 426, 157, and 200 cases that met the inclusion criteria were No 10 (18.9) 43 (81.1) 53
identified for each time period, respectively. As shown in Table 1, Total 108 49 157
surgical procedures at follow-up had a greater proportion of inpa- Follow-up (%)
tients (81% vs 67%), longer mean duration of surgery (115 min vs Yes 118 (95.2) 6 (4.8) 124
97 min), and smaller proportion of patients with American Society of No 6 (7.9) 70 (92.1) 76
Anesthesiologists (ASA) physical status I or II (67% vs 78%). Total 124 76 200
Administering antibiotics when indicated and not giving an-
tibiotics when not indicated according to guideline improved from
81% (345/426) to 94% (188/200; P < 0.001). Reduction in adminis-
tration of unnecessary antibiotics accounted for much of the increase TABLE 3. Appropriate Use of Surgical Antibiotic Prophylaxis
(Table 2). by Service
Appropriate Use of Antibiotic Prophylaxis
Appropriate Use of Antibiotic Prophylaxis
Appropriate use of AP increased from 52% to 67% (P < 0.001). Service (%) Preintervention Postintervention Follow-up P
The highest rate was in cardiovascular surgery, whereas improvements All 51.6 57.3 67.0 <0.001
in orthopedic surgery were statistically significant. There were also Cardiovascular 69.0 62.5 83.3 0.226
trends of improvement in cardiovascular surgery, urology, and plastic General 54.9 44.0 60.4 0.516
surgery. Practice was relatively unchanged in otolaryngology and Neurosurgery 27.3 23.1 11.1 0.416
general surgery, and appropriate AP use declined in neurosurgery Orthopedics 39.6 69.6 69.8 <0.001
(Table 3). Otolaryngology 73.7 50.0 72.7 0.928
Plastics 64.9 61.1 78.6 0.503
Complete and Partial Compliance Urology 48.8 63.2 63.2 0.259
For surgical procedures with prophylaxis indicated, complete Bold values are statistically significant.
compliance with the guideline improved from 26.2% (49/187) to
53.2% (66/200; P < 0.001). Cardiovascular and plastic surgery had
the highest compliance. Complete compliance increased by a total of
ogy. Partial compliance improved in all surgical services from 73.3%
50% in each of plastic surgery, orthopedic surgery, and otolaryngol-
(137/187) to 88.7% (110/200; P = 0.001) overall. Neurosurgery had
the highest partial compliance, followed by orthopedic surgery and
cardiovascular surgery. Otolaryngology had the greatest relative im-
TABLE 1. Characteristics of Surgical Cases provement (Table 4).
Preintervention Postintervention Follow-up
Variable (N = 426) (N = 157) (N = 200) P Drug Selection, Dosage, Timing, Redosing,
Age, mean, yr 6.7 7.2 7.1 0.428 and Duration Compliance
Weight, mean, kg 28.0 29.3 29.5 0.451 Compliance with the recommended dosage, timing, redosing,
Surgical duration, 97.2 126.0 115.2 0.041 and duration increased significantly after the intervention in surgical
mean, min procedures with antibiotics indicated and given (Table 5). There was
Sex (%) 0.101 no significant change in correct antibiotic selection, where compli-
Male 64.3 61.8 57.5 ance was already high.
Female 35.7 38.2 42.5 Guideline compliance by surgical service is provided in
Admission
type (%) <0.001
Table 6. For cardiovascular procedures with prophylaxis indicated
Inpatient 66.9 89.6 80.5 and given at follow-up period, all patients received correct antibi-
Outpatient 33.1 10.4 19.5 otics and redosing for the appropriate duration. Dosing and timing in
ASA physical 0.002 cardiovascular cases also had greater than 90% compliance. Timely
status (%)∗ AP administration increased in general surgery, but further work is
I–II 78.3 67.5 66.5 required to improve correct dosage, redosing, and duration. Neu-
III–V 21.7 26.8 33.5 rosurgery had 100% compliance in all criteria except duration, in
Case type (%) 0.715 which antibiotics were continued postoperatively for most patients.
Elective 74.6 60.0 76.0 Prolonged AP was also the main source of noncompliance in ortho-
Emergency 25.4 40.0 24.0
pedic surgery and otolaryngology. All the other criteria in these 2
Bold values are statistically significant.
∗
specialties had close to 90% compliance or better. Plastic surgery had
Missing data: 3 cases in preintervention and 9 cases in postintervention. one of the highest rates of complete compliance. However, duration
ASA indicates American Society of Anesthesiologists.
compliance remained low, despite considerable improvement. There


C 2014 Wolters Kluwer Health, Inc. All rights reserved. www.annalsofsurgery.com | 405

Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
So et al Annals of Surgery r Volume 262, Number 2, August 2015

TABLE 4. Complete and Partial Compliance by Surgical Service for Cases With Antibiotic Prophylaxis
Indicated
Complete Compliance Partial Compliance
Service (%) Preintervention Postintervention Follow-up Preintervention Postintervention Follow-up
All 26.2 46.2 53.2 73.3 81.7 88.7
Cardiovascular 69.0 62.5 83.3 82.8 93.8 91.7
General 23.3 40.9 47.1 62.8 72.7 85.3
Neurosurgery 25.0 23.1 11.1 85.7 100.0 100.0
Orthopedics 4.5 53.6 55.2 86.4 85.7 96.6
Otolaryngology 0.0 40.0 50.0 16.7 60.0 75.0
Plastics 9.1 45.4 60.0 72.7 72.7 80.0
Urology 34.6 44.4 36.4 57.7 66.7 81.8
Included only surgical procedures with antibiotics indicated.

TABLE 5. Drug, Dosage, Timing, Redosing, and Duration TABLE 6. Drug, Dosage, Timing, Redosing, and Duration
Compliance in Cases With Antibiotic Prophylaxis Indicated Compliance by Surgical Service in Cases With Antibiotic
and Administered Prophylaxis Indicated and Administered
Criterion (%) Preintervention Postintervention Follow-up P Guideline Compliance
Correct drug 94.3 96.9 96.6 0.355 Surgical Service (%) Drug Dosage Timing Redosing Duration
Correct dosage 77.5 88.7 90.7 0.003
Correct timing 83.3 89.8 95.8 0.001 Cardiovascular
Correct redosing 62.5 81.8 95.8 0.003 Preintervention 100.0 85.7 85.7 88.9 89.3
Correct duration 47.1 69.4 65.3 0.002 Postintervention 100.0 93.8 93.8 83.3 100.0
Follow-up 100.0 91.7 91.7 100.0 100.0
Bold values are statistically significant. General
Included cases with antibiotics indicated and given. Preintervention 85.4 72.5 78.1 62.5 53.7
Redosing included only cases that required top-up. Postintervention 85.7 80.0 90.5 66.7 71.4
Follow-up 90.9 78.8 97.0 50.0 63.6
Neurosurgery
Preintervention 100.0 74.1 88.9 16.7 44.4
were compliance improvements for 4 of 5 AP elements in urology, Postintervention 100.0 100.0 100.0 100.0 30.8
but complete compliance was unchanged. Follow-up 100.0 100.0 100.0 100.0 11.1
Our findings prompted changes to perioperative process and Orthopedics
workflow. Extended preoperative positioning and preparation for Preintervention 100.0 76.2 90.5 66.7 9.5
complex procedures had resulted in some antibiotics given out-of- Postintervention 100.0 85.7 85.7 71.4 75.0
window (>60 minutes before incision). Other patients also received Follow-up 100.0 96.6 96.6 100.0 62.1
prophylaxis more than 1 hour before surgery in the emergency de- Otolaryngology
partment or “on call” to the OR. We addressed this with the anesthe- Preintervention 100.0 100.0 100.0 —∗ 0.0
siologists, designating the role of AP administration to anesthesia and Postintervention 100.0 75.0 75.0 —∗ 50.0
Follow-up 100.0 88.9 100.0 100.0 66.7
requesting that prophylaxis be given closer to the start of surgery.26 Plastics
We also worked with surgeons and house staff to restrict prophylaxis Preintervention 100.0 90.0 80.0 66.7 20.0
administration to the OR.27 Antibiotic selection was already high Postintervention 100.0 87.5 100.0 100.0 62.5
preintervention; however, the general surgeons often used a drug that Follow-up 80.0 100.0 80.0 —∗ 60.0
was not on the recommended list for gastrointestinal procedures. Urology
Discussions between the general surgeons and the hospital’s antimi- Preintervention 84.0 76.0 72.0 50.0 68.0
crobial advisory group led to the addition of cefoxitin to the list of Postintervention 100.0 100.0 75.0 100.0 62.5
agents for gastrointestinal procedures and revision of the hospital for- Follow-up 100.0 100.0 100.0 100.0 44.4
mulary. Surgeons and anesthesiologists found having posters showing Included cases with antibiotics indicated and given. Redosing included only cases
the guideline in the ORs very helpful, but they noticed that copies that required top-up.
∗
were missing from some rooms by the follow-up period. Pharmacy, No cases required redosing.
perioperative services, and anesthesia established a new process to
ensure that the guidelines were available in each OR at all times,
with a regular schedule for replacement so that the guidelines were use and compliance with AP in pediatric surgery, including correct
up-to-date. dosage, timing, redosing, and duration. However, results did vary by
specialty, with high rates of compliance in cardiovascular surgery and
DISCUSSION neurosurgery and lower compliance in general surgery, otolaryngol-
Reduction in SSIs is an important goal for clinicians and health ogy, and plastic surgery. Almost all cardiovascular and neurosurgical
care organizations. Appropriate use of AP is one strategy to reduce procedures required AP. The general consensus on antibiotic indica-
SSI.7,9 Clinical practice guidelines, if adopted, can improve quality of tion and repetition of clinical practice may have contributed to the
care.8,15,28,29 We demonstrated that a comprehensive guideline imple- high level of compliance in these surgical specialties. In contrast,
mented with a multifaceted intervention led to increased appropriate indications for procedures in other specialties were more varied and

406 | www.annalsofsurgery.com 
C 2014 Wolters Kluwer Health, Inc. All rights reserved.

Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 262, Number 2, August 2015 Surgical Antibiotic Guideline Compliance

agreement on antibiotic use was less uniform among surgeons. As a challenge. Many factors could account for the challenge of reaching
result, compliance with the guideline was lower. perfect compliance. First, the relative infrequency of SSI may im-
Although difficult to isolate any specific component, we con- pede the impetus for change. Second, in an academic health sciences
sidered physician’s involvement in guideline development and im- center, the frequent changeover in house staff requires continued re-
plementation an important success factor, as has previously been inforcement of change management strategies. Third, guidelines can
described.15 There is little research on appropriate surgical AP use never consider all possible patient situations. Fourth, any approach
in pediatric patients, and existing guidelines are based mainly on that relies solely on auditing behavioral change can never achieve
adult research. Awareness and consensus among physicians, the main 100% compliance. Further efforts of quality improvement need to
guideline users, was vital to adoption and compliance, so at all stages, focus on dynamic (rather than static) strategies for change, such as
we focused on engaging and securing the support of surgeons and the SIS guiding antibiotic indication and choice rather than simply
anesthesiologists. The guiding coalition included all relevant medical recording compliance. Also, new approaches such as standardized
disciplines to ensure that important concerns were addressed and to clinical assessment and management plans may allow greater clini-
facilitate communication with clinicians. The surgeon-in-chief and cian input into guidelines and consider important patient factors that
the infectious disease specialist chaired the task force and champi- prompt physicians to deviate from guidelines.38
oned the project with both anesthesia and surgical representation. We There were several limitations to our study. First, the design
communicated with surgeons on multiple occasions to explain the was not randomized, resulting in differences between the baseline and
project and inform them of the changes in practice. We also required follow-up groups. We found that the mean duration of surgery and
surgeons to review the AP guideline and agree on the prophylaxis the proportion of inpatients and patients with American Society of
indications. Anesthesiologists status higher than II were greater at follow-up. The
Another important factor in guideline implementation was ac- probable explanation for the difference was fewer scheduled complex
curate surveillance. First, we made great efforts to ensure accuracy surgical procedures and subsequently greater proportion of emer-
of AP documentation, including modifying the data entry process so gency cases during the summer months that made up the baseline
that antibiotic documentation must be completed before the OR log group than those that made up the follow-up group. Yet, given that
could be signed off by nurses. The OR nurses were also sent e-mail compliance improved despite these potential biases, which should
reminders about correct documentation. Second, we added a new field have decreased compliance, the effect of the intervention may in fact
to the OR log to distinguish patients who were already on antibiotics be stronger than observed. Second, although we made great efforts to
before surgery from those who did not receive antibiotics. Third, the ensure accuracy, compliance and antibiotic indication may be misclas-
e-mail notifications that surgeons received provided audit and feed- sified. However, because misclassification results in underestimation
back and allowed them to indicate issues with the guideline in general of compliance rates and intervention effects, improvements in com-
and for individual cases. Surgeons were requested to provide reasons pliance may again be potentially greater than our estimates. Third, our
for deviance from guidelines. In some instances, this feedback led study did not assess the effect of guideline compliance on incidence
to minor changes or reclassification of indications in the guideline. of SSI. However, multiple studies have shown that appropriate use
This process of guideline feedback and refinement is consistent with of AP reduces SSI.7 Fourth, we could not determine the individual
a prior study examining the clinical deviation from a standardized impact of each component in the multifaceted intervention. However,
clinical assessment and management plans for aortic valve stenosis, our goal was to maximize improvement in compliance, and to that
arterial switch operation, and hypertrophic cardiomyopathy.30 end, the synergistic effect of a multifaceted intervention was a signif-
Prior research has shown that prophylaxis continued after icant advantage over a single-intervention approach. The strategies
surgery did not further reduce the incidence of SSI and was in fact we implemented were evidence-based and targeted known barriers to
associated with antibiotic-related morbidity, emergence of antimi- guideline adoption and compliance. Finally, our study was based on a
crobial resistance, and higher cost.31,32 However, as with previous single children’s hospital and may not be generalizable to other insti-
investigations, the most common prophylaxis administration error in tutions. We included surgical procedures from all services to obtain a
this study was prolonged use.14,27,33–36 We set electronic alerts to representative sample, but multicentered studies would be necessary
warn clinicians if they ordered unwarranted postoperative prophy- to corroborate our findings.
laxis, but the reminder was often bypassed. Neurosurgery had the
lowest complete compliance rate solely as a result of noncompliance CONCLUSIONS
with recommended prophylaxis duration. Even with partial compli- Using evidence-based guidelines to change physician’s prac-
ance, there were cases in which prophylaxis was prolonged or top-up tice required multiple dissemination and implementation strategies.
doses were missed. However, although redosing in partially compli- A multifaceted intervention that combined access to recommended
ant surgical procedures improved from 91% to 100% compliance by antibiotics, reference material, targeted education, verification of pro-
follow-up, correct duration remained low (66%). The general percep- phylaxis administration at surgical time-out, electronic alerts for inap-
tion of the risk of SSI being greater than that of complications from propriate postoperative prophylaxis, and e-mails notifying surgeons
unnecessary antibiotics was difficult to modify. Eskicioglu et al37 when the guideline was not followed improved compliance with a
speculated that stronger evidence and greater effort were required to pediatric surgical AP guideline. Appropriate AP use, complete and
change behavior associated with recommendations for discontinuing partial compliance, and correct dosage, timing, redosing, and dura-
practice compared with recommendations for adopting practice. Au- tion improved significantly. Further efforts to improve compliance
tomated alerts and physician-specific feedback raised awareness of with AP guidelines may require specialty-specific attention. The im-
prolonged antibiotics, but further attention may be required to induce pact of AP guidelines on SSI rate, cost savings, and other patient- and
compliance. health care–related outcomes are areas of future study.
This study confirms that extended use of multiple strategies
can change physician’s behavior. Although we demonstrated signif- ACKNOWLEDGMENTS
icant improvements in guideline compliance, our results were still The authors thank The Surgical Site Infection Task Force for
short of 100% appropriate use of AP. Despite improvement in all critically reviewing and revising the article for important intellec-
aspects of the guideline (drug, dose, timing, redosing, and duration), tual content; Stéphanie Collin, MSc (Université de Moncton), for
achieving compliance for all 5 criteria in the same patient was a assistance in development of the antibiotic prophylaxis guideline and


C 2014 Wolters Kluwer Health, Inc. All rights reserved. www.annalsofsurgery.com | 407

Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
So et al Annals of Surgery r Volume 262, Number 2, August 2015

indications; and Dan Napier, BSc (Sunnybrook Hospital), for pro- 18. Cabana MD, Rand CS, Powe NR, et al. Why don’t physicians follow clinical
gramming support. practice guidelines? A framework for improvement. JAMA. 1999;282:1458–
1465.
The SickKids Surgical Site Infection Task Force includes the
19. Antimicrobial prophylaxis in surgery. Med Lett Drugs Ther. 1999;41:
following members: Zoran Bojic, MHSc, PMP, CLA, ASQ-CMQOE; 75–79.
Rita Damignani, MSc, BScPT; Annie Fecteau, MDCM, FRCP(C); 20. Post-exposure anthrax prophylaxis. Med Lett Drugs Ther. 2001;43:91–92.
Igor Luginbuehl, MD; Anne G. Matlow, MD, MSc, FRCPC; Jean Paul 21. Gilbert DN, Moellering RC, Sande MA, eds. The Sanford Guide to Antimi-
Paraiso, MSc; Lisa Pendergast, RN; Debi Senger, CHIM; Jeannette P. crobial Therapy 2003. 33rd ed. Hyde Park, VT: Antimicrobial Therapy Inc.;
So, MSc; Laurie Streitenberger, RN, BSc, CIC; Kathryn Timberlake, 2003.
PharmD; and James G. Wright, MD, MPH, FRCSC. 22. The American Academy of Pediatrics. Antimicrobial prophylaxis. In: Ameri-
can Academy of Pediatrics, Pickering, LK, ed. Red Book: 2003 Report on the
Committee on Infectious Diseases. 26th ed. Elk Grove Village, IL: American
REFERENCES Academy of Pediatrics; 2003:773–787.
1. de Lissovoy G, Fraeman K, Hutchins V, et al. Surgical site infection: incidence 23. ASHP Therapeutic Guidelines on Antimicrobial Prophylaxis in Surgery.
and impact on hospital utilization and treatment costs. Am J Infect Control. American Society of Health-System Pharmacists. Am J Health Syst Pharm.
2009;37:387–397. 1999;56:1839–1888.
2. Kirkland KB, Briggs JP, Trivette SL, et al. The impact of surgical-site infections 24. Page CP, Bohnen JM, Fletcher JR, et al. Antimicrobial prophylaxis for sur-
in the 1990s: attributable mortality, excess length of hospitalization, and extra gical wounds. Guidelines for clinical care. [Erratum appears in Arch Surg.
costs. Infect Control Hosp Epidemiol. 1999;20:725–730. 1993128:410]. Arch Surg. 1993;128:79–88.
3. Umscheid CA, Mitchell MD, Doshi JA, et al. Estimating the proportion of 25. Hirota WK, Petersen K, Baron TH, et al. Guidelines for antibiotic prophylaxis
healthcare-associated infections that are reasonably preventable and the related for GI endoscopy. Gastrointest Endosc. 2003;58:475–482.
mortality and costs. Infect Control Hosp Epidemiol. 2011;32:101–114.
26. White A, Schneider T. Improving compliance with prophylactic antibiotic
4. Mangram AJ, Horan TC, Pearson ML, et al. Guideline for prevention of sur- administration guidelines. [Erratum appears in AORN J. 2007;85:711. Note:
gical site infection, 1999. Centers for Disease Control and Prevention (CDC) Dosage error in article text]. AORN J. 2007;85:173–180.
Hospital Infection Control Practices Advisory Committee. Am J Infect Control.
27. Whitman G, Cowell V, Parris K, et al. Prophylactic antibiotic use: hardwiring
1999;27:97–132; quiz 3–4; discussion 96.
of physician behavior, not education, leads to compliance. J Am Coll Surg.
5. Bratzler DW, Houck PM. Antimicrobial prophylaxis for surgery: an advisory 2008;207:88–94.
statement from the National Surgical Infection Prevention Project. Am J Surg.
28. Grimshaw JM, Russell IT. Effect of clinical guidelines on medical prac-
2005;189:395–404.
tice: a systematic review of rigorous evaluations. Lancet. 1993;342:1317–
6. Woods RK, Dellinger EP. Current guidelines for antibiotic prophylaxis of 1322.
surgical wounds. Am Fam Physician. 1998;57:2731–2740.
29. Lugtenberg M, Burgers JS, Westert GP. Effects of evidence-based clinical
7. Bowater RJ, Stirling SA, Lilford RJ. Is antibiotic prophylaxis in surgery a practice guidelines on quality of care: a systematic review. Qual Saf Health
generally effective intervention? Testing a generic hypothesis over a set of Care. 2009;18:385–392.
meta-analyses. Ann Surg. 2009;249:551–556.
30. Farias M, Friedman KG, Powell AJ, et al. Dynamic evolution of practice
8. Talon D, Mourey F, Touratier S, et al. Evaluation of current practices in surgical guidelines: analysis of deviations from assessment and management plans.
antimicrobial prophylaxis before and after implementation of local guidelines. Pediatrics. 2012;130:93–98.
J Hosp Infect. 2001;49:193–198.
31. Alvarez P, Fuentes C, Garcia N, et al. Evaluation of the duration of the antibiotic
9. Anderson DJ, Kaye KS, Classen D, et al. Strategies to prevent surgical site prophylaxis in paediatric postoperative heart surgery patients. Pediatr Cardiol.
infections in acute care hospitals. Infect Control Hosp Epidemiol. 2008;29 2012;33:735–738.
(suppl 1):S51–S61.
32. De Chiara S, Chiumello D, Nicolini R, et al. Prolongation of antibiotic prophy-
10. Dellinger EP, Gross PA, Barrett TL, et al. Quality standard for antimicrobial laxis after clean and clean-contaminated surgery and surgical site infection.
prophylaxis in surgical procedures. Infectious Diseases Society of America. Minerva Anestesiol. 2010;76:413–419.
Clin Infect Dis. 1994;18:422–427.
33. Ozgun H, Ertugrul BM, Soyder A, et al. Peri-operative antibiotic prophylaxis:
11. Lallemand S, Thouverez M, Bailly P, et al. Non-observance of guidelines for adherence to guidelines and effects of educational intervention. Int J Surg.
surgical antimicrobial prophylaxis and surgical-site infections. Pharm World 2010;8:159–163.
Sci. 2002;24:95–99.
34. Gorecki W, Grochowska E, Krysta M, et al. A prospective comparison of
12. Bratzler DW, Houck PM, Richards C, et al. Use of antimicrobial prophylaxis for antibiotic usage in pediatric surgical patients: the safety, advantage, and effec-
major surgery: baseline results from the National Surgical Infection Prevention tiveness of the Surgical Infection Society guidelines versus a common practice.
Project. Arch Surg. 2005;140:174–182. J Pediatr Surg. 2002;37:1430–1434.
13. Rangel SJ, Fung M, Graham DA, et al. Recent trends in the use of antibiotic 35. Kritchevsky SB, Braun BI, Bush AJ, et al. The effect of a quality improve-
prophylaxis in pediatric surgery. J Pediatr Surg. 2011;46:366–371. ment collaborative to improve antimicrobial prophylaxis in surgical patients a
14. Hing WC, Yeoh TT, Yeoh SF, et al. An evaluation of antimicrobial prophy- randomized trial. Ann Intern Med. 2008;149:472–480.
laxis in paediatric surgery and its financial implication. J Clin Pharm Ther. 36. Groselj Grenc M, Derganc M, Trsinar B, et al. Antibiotic prophylaxis for
2005;30:371–381. surgical procedures on children. J Chemother. 2006;18:38–42.
15. So JP, Wright JG. The use of three strategies to improve quality of care at a 37. Eskicioglu C, Gagliardi A, Fenech DS, et al. Can a tailored knowledge trans-
national level. Clin Orthop Relat Res. 2012;470:1006–1016. lation strategy improve short term outcomes? A pilot study to increase com-
16. Mickan S, Burls A, Glasziou P. Patterns of ‘leakage’ in the utilisation of clinical pliance with bowel preparation recommendations in general surgery. Surgery.
guidelines: a systematic review. Postgrad Med J. 2011;87:670–679. 2011;150:68–74.
17. Pathman DE, Konrad TR, Freed GL, et al. The awareness-to-adherence model 38. Rathod RH, Farias M, Friedman KG, et al. A novel approach to gathering
of the steps to clinical guideline compliance. The case of pediatric vaccine and acting on relevant clinical information: SCAMPs. Congenit Heart Dis.
recommendations. Med Care. 1996;34:873–889. 2010;5:343–353.

408 | www.annalsofsurgery.com 
C 2014 Wolters Kluwer Health, Inc. All rights reserved.

Copyright © 2014 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.