In the United States, surgical procedures account for up to 450,000 pediatric admissions each year,Reference Somme, Bronsert, Morrato and Ziegler1,Reference Tzong, Han, Roh and Ing2 with antibiotic prophylaxis administered for >50% of surgeries.Reference Sandora, Fung, Melvin, Graham and Rangel3–Reference Bratzler, Houck and Richards5 National guidelines provide procedure-specific recommendations for antimicrobial prophylaxis, including drug and dosing, in an effort to reduce surgical site infections.Reference Bratzler, Houck and Richards5 The 2017 Centers for Disease Control and Prevention (CDC) guideline recommends only a single dose of perioperative prophylaxis for clean and clean-contaminated cases.Reference Berrios-Torres, Umscheid and Bratzler6
Despite these guidelines, inappropriate surgical prophylaxis use continues to be common. In a pediatric study of surgical antibiotic prophylaxis in 348,119 procedures, 35.4% of antibiotic prophylaxes were considered inappropriate, with interhospital variability ranging from 15.6% and 52.7%.Reference Sandora, Fung, Melvin, Graham and Rangel3 In another study, Voit et alReference Voit, Todd, Nelson and Nyquist7 found that 28% of surgical procedures had excess duration of antibiotic prophylaxis prior to the new recommendation eliminating postoperative doses in low-risk surgeries. In this multisite study, we aimed to determine the prevalence of inappropriate surgical prophylaxis among hospitalized children.
Methods
Study sample
A point-prevalence survey (PPS) was conducted in 32 children’s hospitals to document antimicrobial prescribing among hospitalized patients. Data were collected during 6 quarterly cycles from September 2016 to December 2017. Patients <18 years of age with an active antimicrobial order at 8:00 am on the day of the PPS were eligible for inclusion. This methodology has been used in similar studies.Reference Versporten, Bielicki, Drapier, Sharland and Goossens8,Reference Versporten, Sharland and Bielicki9 A chart review of the electronic medical record was performed, and patient data (eg, age, sex, medical service type, and underlying chronic conditions) and antimicrobial characteristics (eg, name, route, indication, appropriateness) were recorded in a Research Electronic Data Capture (REDCap) online database. Additionally, for antibiotics, data were collected regarding when or if the antimicrobial stewardship program (ASP) would have routinely reviewed this antibiotic. For this post-hoc analysis, we selected those patients who were receiving an antibiotic with an EHR-documented indication for surgical prophylaxis.
Study outcome
The primary outcome was whether the antibiotic was inappropriate or not, which was recorded on the day of the PPS. The determination of inappropriate antibiotic administration was completed by the physician(s) and/or clinical pharmacist(s) involved with the institution’s ASP. Several factors were considered when assigning appropriateness, including drug, route, and indication. If the antibiotic was recorded as inappropriate, the reviewer indicated the primary reason for inappropriateness from a prespecified list (ie, pathogen–drug mismatch, surgical prophylaxis duration >24 hours, unnecessary duplicate therapies, intravenous medication that could be administered orally, or other with free text response). A standard operations manual that provided clear definitions of inappropriateness was given to each participating institution.
Data analysis
The frequency of inappropriate surgical prophylaxis was calculated, stratified by surgical specialty (otolaryngology, orthopedic, cardiovascular, neurosurgery, urology, cosmetic or reconstructive, general surgery, other) and whether the ASP would routinely review the antibiotic. Additionally, we evaluated the variability in inappropriate surgical prophylaxis across the participating institutions. Analyses were completed using SAS version 9.4 software (SAS Institute, Cary, NC). Institutional review board approval was obtained for all sites.
Results
Overall, 32 hospitals participated in at least 1 of 6 PPSs. Clinical characteristics from 13,051 patients who were actively receiving antimicrobial treatment were recorded during the study period. Of these, 1,324 patients (1,477 orders) were receiving antibiotics for surgical prophylaxis. The most commonly prescribed surgical prophylaxis antibiotic was cefazolin (n = 788, 53.4%) followed by clindamycin (n = 85, 5.8%), vancomycin (n = 85, 5.8%), cefoxitin (n = 69, 4.7%), and piperacillin/tazobactam (n = 54, 3.7%).
Overall, 485 surgical prophylaxis antibiotics (33.0%) were categorized as inappropriate. The most common reason was due to prophylaxis of >24 hours (n = 387, 79.8%). Other inappropriate reasons for surgical prophylaxis included prophylaxis not indicated (n = 32, 6.6%) and antibiotic too broad (n = 29, 6.0%). The frequency of inappropriate surgical prophylaxis was higher among otolaryngologic surgery patients (62.7%; 95% confidence interval [CI], 52.6–72.1) cosmetic or reconstructive surgery patients (40.7%; 95% CI, 30.0–52.2), and neurosurgery patients (40.3%; 95% CI, 34.2–46.6) compared with orthopedic surgery patients (15.5%; 95% CI, 11.1–20.7) and cardiovascular surgery patients (24.5%; 95% CI, 20.1–29.3). Of the 485 surgical prophylaxis prescriptions reviewed that were determined to be inappropriate, most (n = 258, 53.2%) would not have been routinely reviewed by the ASP. Inappropriate surgical prophylaxis varied significantly across the 32 hospitals, from 0.0% to 62.8% (Fig. 1).
Fig. 1. Inappropriate surgical prophylaxis by study hospital.
Discussion
This study demonstrates the continued and consistent inappropriate use of surgical antibiotic prophylaxis in children. Similar to previous studies,Reference Sandora, Fung, Melvin, Graham and Rangel3,Reference Rangel, Fung, Graham, Ma, Nelson and Sandora4,Reference van Kasteren, Mannien and Kullberg10 33% of surgical antibiotic prophylaxis was considered inappropriate. The most common reason for inappropriate use was prolonged duration of prophylaxis for >24 hours. In light of the new CDC guideline recommending no doses for low-risk procedures, our inappropriate rate is likely an underestimate. In addition, institutional-level inappropriateness in this study varied from 0 to 63%, which agrees with prior research.Reference Sandora, Fung, Melvin, Graham and Rangel3
This study has several limitations. First, the PPS methodology provides a 1-day glimpse of antimicrobial use within a hospital which may not be entirely comprehensive of all surgeries, including procedures where prophylaxis was not given. However, the prevalence of inappropriate use from our study shows similarity with prior research. Lastly, categorizing an antibiotic as inappropriate or not was a perceived determination by ASP team members within each institution, which may have introduced some differential classification. However, an operations manual was used to help standardize the definition of inappropriateness. Moreover, only a trained ASP physician and/or clinical pharmacist was permitted to make the determination.
In conclusion, a significant portion of surgical antibiotic prophylaxis is inappropriate. This study specifically highlights the prolonged durations of prophylaxis being provided. Future studies are needed to better estimate the overall rate of inappropriate surgical prophylaxis, the factors that drive prolonged surgical prophylaxis, and the best interventions to improve the use of surgical antibiotic prophylaxis.
Acknowledgments
This study could not have done without the 100+ researchers who help collect data. We thank all team members who were part of this PPS study. We also thank the SHARPS collaborative.
Financial support
This work was supported by an investigator-initiated grant from Merck.
Conflicts of interest
All authors report no conflicts of interest relevant to this article.
