Perioperative Antibiotic Therapy and Postoperative Infectious Complications in Appendicitis

Abstract

Appendicitis, first described in 1886, remains a condition with only partially understood pathophysiology. While appendectomy is the worldwide gold standard treatment, the paradigm of surgery is increasingly challenged as non-operative antibiotic management is being studied as an alternative to surgery. The first German guideline for the treatment of appendicitis in adults was published in 2021. This retrospective medical chart review study described a cohort of 1346 consecutive patients undergoing emergency appendectomy at a German university hospital within a six-year time frame and investigated preoperative symptoms, surgical treatment, postoperative course, and especially infectious complications following the procedure. Special attention was given to the microbial spectrum from intraoperatively obtained intraabdominal swabs and post-appendectomy culture results. Cases were classified as CA, UA, NA, or NC according to an algorithm that followed suggestions of several international guidelines. CA was defined by the presence of appendix inflammation plus either one or a combination of the following: gangrene of the appendix, perforation of the appendix, or periappendicular abscess. We were able to show that UA affected predominantly young patients, whereas CA patients were significantly older, had more pre-existing conditions, higher ASA scores, and more severe symptoms, as well as a higher variability in abdominal pain localization before admission. Hospital stay duration was more than twice as long for CA (4.8 days) compared to UA (2.0 days). CA patients had higher preoperative leukocyte counts and CRP levels, and surgeries involved extended resection of the coecum or ileocoecal resection more often than in UA. Histopathological evaluation was not always equal to surgical assessment of appendix perforation and often suggested milder stages of inflammation than indicated by operation reports. The most common combination of agents for perioperative antimicrobial prophylaxis was Cefotaxime + Metronidazole, and antibiotic treatment was continued in more than 80% of CA patients for an average duration of 3.7 days compared to 21% and 1.7 days in the UA group. Microbiological sampling performed during appendectomy was dependent on the surgeon’s choice and performed almost twice as frequently in CA (79%) compared to the UA (41%). Post-appendectomy microbiological sampling was performed in 4.1% of all patients (UA: 0.5%, CA: 10.9%). Culture results showed no growth of bacteria in 43% (post-appendectomy: 39%), single species infections in 9% (post-appendectomy: 18%), and mixed infections in 48% (post-appendectomy: 44%) of cases. The average number of different species identified for UA was 2.8 (post: 1.5) and for CA 3.7 (post: 2.6). Anaerobes were present in 80% of intraoperatively obtained positive cultures and in 44% of post-appendectomy positive cultures. The microbial spectrum from post-appendectomy cultures was similar to the culture results from appendectomies, with the four most commonly detected species being Bacteroides (intra: 43%, post: 20%), Escherichia (intra: 33%, post: 35%), Streptococcus (intra: 16%, post: 20%), and Enterococcus (intra: 14%, post: 25%). The proportion of Bacteroides spp. was significantly lower in post-appendectomy cultures, whereas the proportion of Enterococcus spp. was significantly higher. While the overall distribution of bacterial genera was similar between the intra and post-appendectomy culture results, the overlap for each individual patient was rather low. A comparison of the frequencies of bacterial species showed significant differences from findings in other studies, supporting the importance of the consideration of locoregional bacterial spectra for guiding empirical postoperative therapy in CA. Given that microbiological culture results were considered for adjustments in antibiotic therapy in only a small fraction of cases and were poor predictors of pathogens found in post-appendectomy cultures, we recommend standardizing the indications for collecting cultures. Postoperative complications were more common in the CA group. SSIs were documented in 5% of patients, most commonly intraabdominal, followed by superficial incisional infections. SSIs occurred approximately ten times as often in the CA group compared to UA (13.7% vs 1.4%). Out of all emergency appendectomies performed, 3.6% required reoperation or abscess drainage postoperatively. Almost half of the revision procedures were performed during the initial hospital stay, whereas the remaining procedures were performed after an emergency revisit and readmission of patients. Most revision procedures were performed within ten days following appendectomy, and less than 10% were performed later than one month after appendectomy. Intraabdominal abscesses were the most common diagnosis on revision procedures in the CA group. The second most common diagnosis for the CA group was GI-tract perforation or staple line leak. For the UA group, the most common diagnosis was intraabdominal adhesion, including bowel obstruction. Revisits were 2.5 times more frequent in CA cases, readmissions 5.8 times more frequent, and revision surgeries or abscess drainages 4.2 times more frequent compared to UA. Overall, the mortality rate was low at 0.12% for UA and 0.25% for CA. In the NA group, no emergency revision procedures were performed, suggesting the safety of the procedure and allowing for the attribution of revision procedures in UA and CA to factors inherent to the disease rather than treatment. The results underlie the limitations and potential biases of retrospective medical chart reviews, as data was available or missing dependent on documentation in a real-world hospital setting, patients were discharged and not followed up, and data from only one hospital was considered.