Assessment of perioperative antibiotic prophylaxis for gynecological surgeries at an academic hospital in Brazil

The aim of this study was to assess the adequacy of physicians' practice patterns regarding the use of perioperative antibiotic prophylaxis for gynecological surgeries in an academic hospital specialized in gynecology located at Rio de Janeiro city, Brazil. This is a retrospective study assessing all gynecological surgeries performed over one year. Appropriateness of antibiotic prophylaxis was determined according to criteria adapted from evidence-based guidelines. Clinical practice regarding the use of perioperative antibiotic prophylaxis was considered appropriate for 58.4% of 416 surgeries. The non-indicated use of antimicrobial prophylaxis was the main factor determining the low percentage of overall adequacy. Three variables were independently associated with inappropriate administration of perioperative antibiotics: patients age, breast surgeries and longer procedures. Antibiotic prophylaxis compliance to published recommendations is low. Women undergoing gynecological surgery are exposed to unnecessary risks associated to non-indicated use of antibiotic prophylaxis. Strategies aimed to improve compliance to evidence-based guidelines are necessary.


Introduction
Surgical site infections (SSI) are the most frequent surgical complication in developing countries and affect as many as 11% of patients undergoing surgical procedures (Allegranzi et al., 2011). In Brazil, they account for 14% to 16% of all nosocomial infections (ANVISA, 2009), with the highest incidence being among obstetric or gynecological patients (Clifford & Daley, 2012;Jaiyeoba, 2012). Most of SSI are preventable if perioperative antimicrobial prophylaxis (PAP) is available (Allegranzi et al., 2011). According to the World Health Organization (WHO), appropriate PAP is defined as "administering an effective antimicrobial agent prior to exposure to contamination during surgery" (WHO, 2018).
However, improper use of antibiotic prophylaxis compromises its efficacy. Many studies have shown that irrational antibiotic use for PAP is still a common problem worldwide and it is frequently related to inappropriate selection of a broadspectrum antibiotic, prolonged duration of PAP, incorrect timing of prophylaxis and un-indicated use (Abubakar et al., 2018;Alemkere, 2018;Viamonte & Cherres, 2016;Wright et al., 2013). There is great variability in PAP used for gynecological surgeries and it frequently does not conform to the published recommendations (Abubakar et al., 2018, Joyce et al, 2015Kremer et al., 2018;Schimpf et al., 2015;Wright et al., 2013). The consequences of improper utilization include adverse drug reactions, development of bacterial resistance and increased healthcare costs (Burke, 2001;van Kasteren et al., 2003).
Therefore, the quality and use of PAP have been the subject of various studies (Abubakar et al., 2018;Burke, 2001;Joyce et al, 2015;Kremer et al., 2018;van Kasteren et al., 2003;Schimpf et al., 2015;Wright et al., 2013). Drug utilization studies could give useful insight about PAP patterns and provide valuable information for the improvement of PAP practices to ensure proper and effective administration of PAP. Nevertheless, there is scarce literature on PAP utilization for gynecological procedures in Brazil. Thus, the objective of this study was to assess physicians' practice patterns regarding the use of PAP in the light of evidence-based guidelines at a university hospital in Brazil.

Methodology
This is a retrospective cross-sectional study on the adequacy of PAP in gynecological and breast surgery. All women over 18 years old who underwent surgery in the Gynecology Institute of the Federal University of Rio de Janeiro, Brazil, in 2017 were included in the study population. Women whose medical records were incomplete or those undergoing antibiotic treatment were excluded. Data were obtained manually from patients' medical charts by a clinical pharmacist. Data collection was carried out using a standardized form and included: age, weight and height of the patient; admission and discharge dates; type of surgery; date timing and duration of surgery; type of antibiotics used (if used), dose, timing of administration and duration (including first and subsequent doses); blood loss greater than 1,5 L. Surgeries were classified into two groups: gynecological surgery (which included all types of procedures except of those on breasts) and breast surgery.
PAP appropriateness was assessed according to criteria adapted from the recommendations of the American Society of Health-system Pharmacists (ASHP) (Bratzler et al., 2013)  Practice pattern regarding the use of PAP was considered adequate if: (i) it was not indicated and was not used or (ii) it was indicated and used properly, which meant meeting all following five criteria: 1. choice of antibiotic: first-generation cephalosporin cefazolin as first choice.
2. dose and administration: 2.0 g intravenous or 3.0 g if patient weight > 120 Kg.
3. timing of administration relative to time of surgery: administration within 60 min before incision.
4. duration of administration: do not exceed 24 h after conclusion of surgery.
5. perioperative re-administration: only necessary if surgery lasts longer than 4h after first dose or in case of blood loss > 1.5 L.
Antibiotic consumption was calculated as the number of Defined Daily Dose (DDD) per 100-procedures using the Anatomical Therapeutic Chemical/Defined Daily Dose (ATC/DDD) index (2020)

Results
In total, 555 women underwent gynecological or breast surgical procedures during the study period. Among them, 416 were eligible for inclusion in the analysis. The mean age, weight and body mass index (BMI) were 47.7 ± 14.1 years, 71.8 ± 15.5 Kg and 28.2 ± 5.6 Kg/m2, respectively. A total of 306 (73.6%) women underwent gynecological surgery and 110 (26.4%) breast surgery. The most frequent gynecological procedures were total abdominal hysterectomy, cervical tissue excision and gynecological exeresis/biopsy. Regarding breast surgeries, mastectomy (bilateral or unilateral) was the most frequent followed by nodulectomy. Mean duration of the surgeries and hospitalization were 103.1 ± 63.9 min and 2.5 ± 1.5 days, respectively. Patient and surgery characteristics are shown in Table 1.
Overall proportion of PAP use was 92.8%. Intravenous cefazolin (ATC J01DB04) was used as single antibiotic for PAP in all procedures. Antimicrobial prophylaxis was administered within 60 min prior to incision in 97.4% of the surgeries.
Duration of PAP was within intraoperative phase for 89.4% of procedures and a second dose of cefazolin was used in 5.7% of the cases. PAP characteristics are shown in Table 2.
Clinical practice regarding the use of PAP was considered appropriate for 58.4% of 416 surgeries. Overall adequacy of PAP was significantly different (p < 0.05) between surgery types. Gynecological procedures showed a lower proportion of PAP adequacy (52.9%) compared to breast surgeries (73.6%). The use of antimicrobial prophylaxis when it was not recommended was the main factor determining the low percentage of overall adequacy. Out of 113 surgeries with no indication for PAP, it was performed in 101 (89.4%). The opposite also occurred as PAP was not performed in 21 (6.9%) out of 303 surgeries for which antibiotic prophylaxis should be used. Adequacy proportion per criteria and type of surgery is shown in Table 3. It is worth noting that among the cohort of patients for whom PAP was recommended and performed (N=282), Research, Society and Development, v. 10, n. 8, e30810817299, 2021 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v10i8.17299 4 antibiotic use met all five quality criteria in 231 cases, resulting in a overall adequacy of 81.9% in this group of patients.   Univariable analysis revealed that patient age, type of surgery and procedure duration were significantly associated (p = 0.000) with practice pattern appropriateness. Obese patients and those who spent more than 2 days in hospital after surgery were no more likely to receive adequate PAP (Table 4). According to the multivariate analysis, clinical practice regarding PAP for patients older than 45 years old was more likely to be adequate (OR = 2.280; 95% CI, 1.503 -3.457), the same occurred in the case of breast surgeries (OR = 2.823; 95% CI, 1.687 -4.724) and surgeries that lasted more than 100 minutes (OR = 2.537; 95% CI, 1.456 -4.421). The variables obesity and hospitalization length did not significantly affect the odds of PAP adequacy (Table 5). Antibiotic consumption was 72,9 DDD of cefazolin per 100 surgeries. Inappropriate use of PAP accounted for 27,1% of all antibiotic doses that were administered.

Discussion
The aim of this study was to analyze the compliance of physicians' practice patterns with evidence-based guidelines regarding the use of antibiotic prophylaxis. Our data suggest that while the density of perioperative antimicrobial use is high, the incidence of inappropriate utilization is frequent and affected 58,4% of the analyzed procedures. Non-indicated use of PAP was the determinant factor for the poor alignment to guidelines recommendations.
The inadequacy of antibiotic prophylaxis for gynecological procedures has already been described in the literature. Wright et al. (2013) studied 545,332 gynecological procedures performed at more than 500 hospitals in the United States between 2003 and 2010 and found that 40.2% of women received non-indicated preoperative antibiotics. Joyce et al. (2015) who studied 326 surgeries performed from 2012 to 2013 at tertiary care hospital found that PAP was used in 53.7% of procedures with no indication. More recently, Kremer et al. (2018) analyzed 2,961 gynecological cases at a tertiary hospital and found that the overall proportion of preoperative antibiotic use for procedures in which antibiotics were not indicated was 19%. The study from Abubakar et al. (2018) that evaluated compliance with surgical antibiotic prophylaxis in obstetrics and gynecological surgeries performed at 3 tertiary hospitals in Nigeria found the proportion of unnecessary antianaerobic combination to be as much as 89.3%.
In our study, the proportion of non-indicated preoperative antibiotics was significantly high, reaching almost 90%. In fact, there was no record of occurrences that could justify such use, for instance, prolonged surgery, unplanned entrance into the abdomen or excessive blood loss (Kremer et al., 2018). The overuse of antibiotics is well known to increase patient morbidity and healthcare costs. Besides subjecting patients to unnecessary toxicity, it may contribute to antimicrobial resistance. These consequences emerge as even more important considering the scarcity of resources and budget constraints of the Brazilian healthcare system. Some factors were significantly associated with appropriateness of PAP when other variables were controlled. Older patients were more likely to receive adequate PAP as well as patients undergoing breast surgery and those requiring longer surgeries. Interestingly, Wright et al. (2013) also found that older age is protective from using non-indicated PAP in gynecological procedures whereas the study conducted by Kremer et al. (2018) revealed that inappropriate use of PAP is less likely to occur the longer the surgery lasts. One possible explanation could be that older patients would require more complex procedures which in turn could result in longer surgeries. Perhaps such characteristics could influence physicians´ practice, making them more prone to comply with recommendations.
Even though previous studies about antibiotic prophylaxis adequacy in gynecological procedures were carried out in different settings and countries, their findings are aligned regarding the unsatisfactory PAP compliance to evidence-based guidelines. In our study, although the surgeries were performed at a teaching hospital specialized in gynecological procedures, there is no stablished local guideline for antibiotic prophylaxis nor explicit recommendations for the adoption of published ones. The fact that literature on antibiotic prophylaxis in various gynecological procedures is scarce, and hence recommendations on specific procedures are unavailable could, at least in part, explain practice variations and PAP misuse.
Patients and students could both benefit from the implementation of strategies designed to promote evidence-based practice dissemination. Besides the reduction of unnecessary treatments and their consequent economic burden (Brook, 2011), the adoption of a formalized protocol could improve patient security and health outcomes (Stulberg et al., 2010).

Conclusion
Inappropriate use of antibiotic prophylaxis in women undergoing gynecological surgeries is prevalent and the results of the present study corroborate it. The implementation of antibiotic stewardship strategies in our institution is necessary to reduce the use of PAP in women that are unlikely to benefit from it. Additionally, further broadening of existing guidelines and the inclusion of other types of gynecology procedures would contribute to reduce antibiotic misuse. Future studies addressing the factors that influence PAP could contribute to improvement of clinical practices.