Discussion

We compared the incidence of SSIs before and after adding anaerobic coverage to antimicrobial prophylaxis for abdominal hysterectomy. The addition of metronidazole was associated with a decrease in the abdominal hysterectomy complex SSI rate.

Postoperative infection remains one of the most common complications of surgical procedures in gynecology. The development of SSIs results in great patient morbidity.^{Reference Lefebvre, Allaire and Jeffrey4} Current antimicrobial prophylaxis guidelines recommend the use of cefazolin, cefotetan, cefoxitin, or ampicillin/sulbactam as a single agent before vaginal or abdominal hysterectomy.^{Reference Bratzler, Dellinger and Olsen5} The American College of Obstetricians and Gynecologists also recommends cefazolin.⁶ The widespread implementation of antibiotic prophylaxis prior to surgery, and the recognition of modifiable risk factors, has led to a reduction in SSIs in recent years. According to NHSN data for 2006 to 2008, the incidence of all-depth SSIs after abdominal hysterectomy was 2.7%.^{Reference Edwards, Peterson and Mu7} A more recent study from the Michigan Surgical Quality Collaboration cohort demonstrated an all-depth SSI rate of ~2.0%.^{Reference Morgan, Swenson and Streifel8,Reference Uppal, Harris and Al-Niaimi9} Uppal et al^{Reference Uppal, Harris and Al-Niaimi9} revealed that the SSI rate was 1.8% with prophylaxis with a β-lactam, compared to 3.1% with β-lactam alternatives and 3.7% with nonstandard prophylaxis.^{Reference Uppal, Harris and Al-Niaimi9} The rate of SSI at our institution decreased from 3.2% to 1.6% after adding metronidazole to cefazolin for SSI prevention. A recent large retrospective study with 18,255 hysterectomies by Till et al^{Reference Till, Morgan and Bazzi3} demonstrated that the risk of SSI was lower, at 1.4%, for patients who received cefazolin and metronidazole, compared to 1.8% with cefazolin alone, and this rate was 2.1% with a second-generation cephalosporin. Currently, no published prospective studies comparing cefazolin and metronidazole to cefazolin alone are available.

Cefazolin, a first-generation cephalosporin, is the most widely used preoperative antibiotic. However, organisms isolated from SSIs after abdominal hysterectomy include anaerobic vaginal flora, most commonly Bacteroides spp, Prevotella spp, Peptostreptococcus spp, and Gardnerella spp.^{Reference Soper, Bump and Hurt10} In our study, the most common organisms were Bacteroides spp. Although anaerobic coverage is recommended as prophylaxis for colorectal surgery,^{Reference Nelson, Glenny and Song11} its role has not been proven for abdominal hysterectomy. Some institutions have shifted to using second-generation cephalosporins to improve anaerobic coverage. However, resistance of Bacteroides spp to second-generation cephalosporins is 15%–25%.^{Reference Aldridge and Sanders12} Our study showed similar results to Till et al^{Reference Till, Morgan and Bazzi3} in complex SSIs. Although all-depth SSI rates trended downward compared to before the implementation of metronidazole, the difference was not statistically significant at our institution. The lack of difference in all-depth SSI after the addition of metronidazole might be because microorganisms associated with superficial infections do not typically include anaerobes. However, given that deep infections are more serious than superficial infections, the impact of adding metronidazole to cefazolin is promising and should be further investigated.

This study has several imitations. It was conducted at a single center and the results may not be generalizable to other institutions. We did not compare differences in baseline characteristics. Comorbidities such as malignancy, obesity, and diabetes mellitus might have affected the development of SSI. However, we calculated the standardized infection ratios according to the NHSN guidelines, which account for such potential confounders, and we were still able to show a decrease in SIRs (Table 1). We only investigated the association between prophylaxis and SSI rates; we did not evaluate postoperative course or mortality. The reduction could have been affected due to the change of procedure approaches over the study period. However, the SSI rates decreased for both surgical approaches. Other SSI prevention activities already ongoing might have affected the reduction. Finally, change of surgeons over the study period could have affected the SSI rate.

In conclusion, hospitals should assess the microbiology of abdominal hysterectomy SSIs and could consider adding metronidazole to their antimicrobial prophylaxis if the rate of anaerobic infections is high.