Retrospective Clinical and Microbiologic Analysis of Patients with Anorectal Abscess

Patients and Methods

Results

Between January 2011 and December 2017, 505 patients underwent incision and drainage of anorectal abscess in our operating room. The mean age was 39.5 years (SD 12.4). There were 411 men (81.4%). One hundred fifteen patients (22.8%) had diabetes mellitus, and fifteen patients (3.0%) had inflammatory bowel disease. Table 1 summarizes the locations of the abscesses. Three hundred eighty-five patients (76.2%) received peri-operative broad-spectrum antibiotics.

Three hundred twenty-one patients (63.5%) were discharged within 24 hours post-operatively. There were 184 patients (36.4%) who required admission for more than 24 hours with a median length of stay of two days (IQR 2, 4). Table 2 compares the length of stay with patient characteristics.

Samples were obtained for culture from the abscess at the time of incision and drainage from 211 patients (41.8%). In 22 patients (10.4%), there was no growth, 97 patients (46.0%) had a single organism, and 92 patients (43.6%) had multiple organisms. A total of 266 microorganisms were recovered. Table 3 summarizes the frequency of these microorganisms. Escherichia coli was the most common, accounting for 37.6%, followed by Bacteroides (13.2%) and Streptococcus spp. (13.2%). Methicillin-resistant Staphylococcus aureus (MRSA) was retrieved from 4.9% of the cohort, with 56.5% of all cultures of S. aureus being MRSA. All MRSA were found to be susceptible to fusidic acid, linezolid, teicoplanin, and vancomycin, whereas 76.9% were susceptible to clindamycin. It was noted that eighteen E. coli and one Klebsiella pneumoniae from the samples were extended-spectrum beta-lactamase (ESBL) producers, representing 7.1% of the micro-organisms.

Table 4 compares the most frequent micro-organisms isolated from patients with diabetes mellitus relative to non-diabetics. Escherichia coli accounted for 34.9% of micro-organisms cultured from patients with diabetes mellitus, followed by Streptococcus spp. (27%) and Klebsiella pneumoniae (20.6%). Despite the significant difference in the proportion of organisms grown from patients with and without diabetes mellitus, E. coli was the most common causative agent in both groups.

Table 5 illustrates the frequency of micro-organisms in patients who had mono-microbial versus poly-microbial infection. We note that gut-derived organisms were associated with poly-microbial infection as opposed to S. aureus, which presented mostly as a mono-microbial infection.

Discussion

This study presents the clinical and microbiologic profile of anorectal abscess. To the best of our knowledge, this is the first such report from the Middle East. Our study supports the current evidence that E. coli is the most frequently cultured micro-organism in anorectal abscess as previously noted in the literature [6–9]. Interestingly, our data showed that there is a significantly lower proportion of E. coli among patients with diabetes mellitus relative to non-diabetics. We also found that bacterial infections are the cause of abscess almost exclusively, with only one patient yielding Candida albicans

Liu et al. showed a higher proportion of Klebsiella pneumoniae among patients with diabetes mellitus than in non-diabetics [9]. Patients with diabetes mellitus have an impaired chemotactic response, lower phagocytosis, and defective neutrophil function, particularly in those with poor glycemic control, which may indicate why this is a common organism in this patient population [10, 11]. We found a similar trend, although it did not reach statistical significance. This may be explained by the retrospective nature of both studies and selection bias, in that the operating surgeon did not obtain cultures on all patients. On the contrary, our findings revealed a significantly higher proportion of Streptococcus and a lower proportion of Bacteroides and E. coli among patients with diabetes mellitus.

Our results revealed that more than half of the S. aureus cultured were MRSA. Albright et al. studied the bacteriology of anorectal abscess and found MRSA to be the causative organism in 24 of 69 patients (34.8%) [12]. Another study by Brown et al. found that MRSA was isolated from 19% of patients and that when S. aureus was present, it was MRSA 75% of the time [13]. These MRSA-associated abscesses typically are marked by extensive induration and erythema with a small amount of or no purulence, where drainage alone frequently is insufficient for cure, with patients requiring antibiotic therapy [12]. Present guidelines covering the empiric use of antibiotics after incision and drainage restrict them to specific indications such as the presence of cellulitis, diabetes mellitus, or immunosuppression [3]. Even though the incidence of community-acquired MRSA skin and soft tissue infections is on the rise, our results showed a low incidence of MRSA among patients with anorectal abscess [14].

We noted that longer hospital stays were more common for older patients, patients with diabetes mellitus, and those having an ischiorectal abscess. This finding is expected, as these patients required discharge planning and control of their blood glucose concentration prior to discharge.

The true incidence of anal fistulas in patients with acute anorectal abscess is unknown, with studies suggesting an incidence range between 26% and 37% [15–18]. The application of adjuvant antibiotic therapy and its role in preventing fistula has re-surfaced recently with two randomized controlled trials investigating the role of adjuvant antibiotics after incision and drainage of anorectal abscess in preventing fistula. The two studies reached contradictory conclusions. Ghahramani et al. examined this question by performing a single-blinded randomized controlled trial over a one-year period where patients were randomized to receive a seven-day course of antibiotics (ciprofloxacin and metronidazole) or not. In this study, the use of antibiotics was protective against the development of fistula [4]. In contrast, the trial by Sozener et al., also carried out over a one-year period, found no association between the use of amoxicillin–clavulanic acid and the subsequent development of a fistula [5]. It is important to note that in both articles, culture of the abscess was not performed, so the antibiotic choice was empiric and may have provided inadequate coverage. Given the increasing prevalence of community-acquired multi-drug-resistant strains such as MRSA and ESBL-producing pathogens, the antibiotics chosen may have provided inadequate coverage. Uncertainties about the benefit of antibiotic therapy in preventing fistula formation will remain until future research takes into consideration culture results and targeted antibiotic therapy.

This study is limited by its retrospective design and the selection bias for obtaining culture samples by the operating surgeons. We also did not account for the patients who had their abscesses drained at locations other than the operating room. Despite that, we here present one of the largest series with an analysis of the microbiology of anorectal abscess.