Risk Factors and Outcome Analysis of Surgical Site Infections in Chinese Elderly Patients with Intestinal Obstruction After Emergency Surgery

Surgical site infection (SSI) and antimicrobial resistance may adversely affect the clinical outcome of older patients. The objective of the study was to investigate the etiology and outcome of SSIs in elderly patients who underwent emergency bowel surgery. We conducted a retrospective study of all patients aged ≥ 60 years who underwent emergency surgery due to an intestinal obstruction from January 2014 to June 2019 in a tertiary teaching hospital in Western China. Demographic data, comorbidities, perioperative physiological and microbiological data, and information on the surgical technique and duration of hospitalization were extracted from clinical records to assess risk factors for SSIs. Of 125 patients included, 115/125 (92%) had a duration of > 48 h postoperative prophylactic antibiotic use; 37 (29.6%) patients were diagnosed with SSI. All occurred within the period of postoperative antimicrobial prophylaxis or after an extended duration of > 48 h prophylactic antibiotic use. Enterobacteriaceae and Enterococcus were the most frequently isolated species (67.4% and 20.9%, respectively): 93.8% of Escherichia coli (15/16) and 33.3% of other Enterobacteriaceae (4/12) isolated were ceftriaxone-resistant. Incision site classification was an independent risk factor for SSI in multivariate analysis. SSI patients had a significantly longer length of stay than those without (29.81 ± 12.96 days vs. 22.52 ± 10.67 days, respectively; p = 0.001). Higher rates of extended-spectrum beta-lactamase-producing Enterobacteriaceae carriage were associated with higher SSI, despite prolonged antimicrobial prophylaxis. This calls for the improved surveillance of resistance in order to offer an alternative prophylaxis for the prevention of these types of infections.


Introduction
Surgical site infection (SSI) is one of the most common postoperative complications [1]. It delays wound healing and may result in prolonged hospitalization and increased medical costs [2]. Some studies have reported that time to surgery, operation duration, blood loss, and the surgical wound classification were independently associated with SSI in elective abdominal surgery [3]. However, there are few data concerning the prediction of the efficacy of antimicrobial prophylaxis in the current context of increasing antimicrobial resistance [4].
In a previous study, we observed an increasing number of acute bowel obstructions over time among elderly patients [5]. Acute intestinal obstruction can cause a dysregulation of the body metabolism and multiple organ failure, and thus, surgeons must perform an emergency procedure to resolve the obstruction by intestinal resection, by intestinal anastomosis, or by eliminating the adhesion zone. According to the type of obstruction, closed loop obstruction, bowel ischemia, or volvulus may occur. However, an emergency procedure is also an important risk factor for SSI. In a recent study in Pakistan, it was reported that the rate of SSI was greater in older individuals with 44.4% of patients over 60 years who developed SSI postoperatively [6]. Given the global increase in the ageing population [7], future healthcare systems will have to adapt to the challenges associated with the management of an increasing number of elderly persons presenting with surgical emergencies. The aim of this retrospective study was to identify risk factors in emergency bowel obstruction surgery among patients ≥ 60 years in order to inform strategies to decrease the occurrence of SSI and compare patient outcomes between those with and without SSI.

Study Design and Patients
A retrospective chart review was performed of all patients ≥ 60 years who had undergone emergency surgery for bowel obstruction at the Third Affiliated Hospital of Zunyi Medical University, a 2500-bed tertiary teaching hospital in Western China, between January 1, 2014, and June 30, 2019. All patients had reported symptoms such as abdominal pain, abdominal distension, absence of flatus, or vomiting and had undergone blood tests. An abdominal ultrasonography or computed tomography scan had been performed immediately upon presentation to the hospital. Diagnosis of bowel obstruction for all cases was based on the patient's medical history, physical examination, and imaging findings. The diagnosis of SSI was defined by the criteria of the Chinese national guideline for the prevention and control of SSIs and included persistent wound discharge or dehiscence, visible abscess or gangrene, and the presence of bacteria confirmed by discharge liquid culture [8]. One hundred and twenty-one patients underwent procedures with an open access and four patients with a laparoscopic approach. Surgical wound types were classified as follows: level II, clean-contaminated, and level III, contaminated.

Statistical Analysis
The Kolmogorov-Smirnov test was used to test the normality of distribution. Normally distributed continuous data were assessed with Student's t-test. If the data were not normally distributed, continuous data were analyzed with the Mann-Whitney U test. ROC curve analysis was used to determine cutoff values for age, time to onset of infection, operation duration, blood loss, preoperative white blood cell and neutrophil counts, and first-day postoperative albumin. The optimal cutoff values were selected according to the maximal Youden index. Multivariate logistic regression analysis was used to identify the risk factors associated with SSI. All statistical analyses were performed with SPSS software, version 17.0 (SPSS Inc., Chicago, IL, USA). A p value < 0.05 was considered to be statistically significant.

SSI and Isolates
Thirty-seven of 125 patients had a confirmed SSI. All presented deep SSIs, i.e., an infection occurring beneath the incision area in the muscle and tissues surrounding the muscles. Diagnosis  (15/16) and 46.2% of other Enterobacteriaceae (6/13) isolated from the wounds of SSI patients were resistant to ceftriaxone. Enterococcus (9/43) was the second genus isolated, with 44.4% resistant to ampicillin (4/9); all isolates were susceptible to vancomycin, with the exception of Enterococcus gallinarum. The Staphylococcus aureus isolate was resistant to oxacillin. All isolated Enterobacteriaceae were susceptible to imipenem, and both Pseudomonas aeruginosa isolates were susceptible to ceftazidime and ciprofloxacin ( Table 2).
The SSI rate was 30.1% (37/123) for patients with a long duration of postoperative antimicrobial prophylaxis (> 48 h), and those undergoing an unforeseen extended prophylactic use of antimicrobials compared to patients with a short duration of antimicrobial prophylaxis (< 48 h). In particular, 8 The bold numbers mean sum of their own subsets in Table 2, or represent statistical significance with P value less than 0.05 in Table 3 Pathogen Strains ( of the 10 patients with less than 48 h prophylactic use of antimicrobial were diagnosed with SSI within 48 h after procedures and then continued to receive the same antimicrobial or were changed to others according to the treatment proposed. However, the latter group was very small as surgeons tended to favor a long-term prophylaxis strategy.  (Table 4).

Discussion
The incidence of SSI in abdominal surgery has been reported to range from 4.8 to 7.4% in China [9,10]. In our study, we found a high rate of 29.6% (37/125) following emergency abdominal surgery in the elderly. This high rate may be explained by the fact that a patient with acute bowel obstruction cannot receive mechanical bowel preparation and antibiotic bowel preparation prior to surgery. Klinger et al. [11] reported that mechanical bowel preparation combined with antibiotic bowel preparation significantly reduced the occurrence of SSI, organ space infection, wound dehiscence, and anastomotic leak compared to no preparation or antibiotic bowel preparation alone. A recent randomized controlled trial of 396 patients undergoing elective colonic surgery showed no difference in SSI and overall morbidity between the mechanical and oral antibiotic bowel preparation group and the no bowel preparation group [12]. However, confirmation of the relationship between the abundance of bacteria in the intestinal lumina and SSI needs to be addressed by more high-quality studies. E. coli is reported as the most common pathogen causing SSI in abdominal surgery [13,14]. Similarly, our findings showed that E. coli was the leading pathogen isolated from surgical sites (16/43; 37.2%), with most isolates resistant to third-generation cephalosporins (15/16; 93.8%). Although data on resistance of pathogens isolated from abdominal surgical sites in a multicenter cross-sectional study in China were not reported [9,14], this is higher than the prevalence (61.1%) of extended-spectrum beta-lactamase (ESBL)-producing E. coli from community-onset bloodstream infections reported in Southwest China and associated with older age groups [15]. Golzarri et al. also reported that patients colonized by ESBL-producing Enterobacteriaceae (ESBL-PE) were at high risk for SSI [16]. In our study, ESBL-PE isolated from surgical sites accounted for 67.86% (19/28) and certainly contributed to the higher rate of patients with SSIs as the majority received second/third generation of cephalosporins as standard prophylaxis. For this reason, cefoxitin was chosen as the prophylactic agent for the patients in the latter period of the study according to local data collected on antimicrobial susceptibility testing. Preoperative cefoxitin had a better efficacy to prevent SSI, but was less significant (SSI rate: cefoxitin vs. others: 15.4% (2/15) vs. 31.8% (35/110), respectively; p = 0.236). Similarly, in a large study by Poeran et al. [17], cephamycins did not demonstrate a better efficacy for SSI prevention compared with noncephamycins. In another study, it was reported that cefoxitin concentrations were not sufficient in subcutaneous adipose tissue (< 8 mg/mL) at the time of surgical closure in obese patients who received a sleeve gastrectomy [18]. Of note, the WHO guidelines for the prevention of SSI have no recommendation regarding screening for ESBL-PE colonization and the impact on surgical antibiotic prophylaxis [19]. Our findings suggest that it would be beneficial to develop a new prophylactic antimicrobial agent to reduce the risk of SSI particularly in areas with a high prevalence of ESBL-PE, including among patients known to be colonized with ESBL.
The positive effect of preoperative surgical antibiotic prophylaxis on SSI has been reported [19,20], but the optimal timing remains a subject of debate; meanwhile, several guidelines recommend that no prophylaxis be given postoperatively [1,19,20]. In our study, it was difficult to determine the beginning of the actual time of antibiotic administration as this information was not well documented for emergency patients and thus we were unable to assess the optimal timing of preoperative surgical antibiotic prophylaxis.
In our study, 92.0% (115/125) of patients had a longer (> 48 h) duration of postoperative prophylactic antibiotic use than recommended (< 24 h) by the American Society of Health-System Practitioners [21] and national technical and administrative documents [8,20], although 32 (86.5%) cases of SSIs were diagnosed before the end of the prophylaxis course. In addition, 8 of 10 patients who received postoperative antimicrobial prophylaxis for less than 48 h were identified with SSI within 48 h and then continued to receive the same antimicrobial or were changed to others according to the treatment proposed. However, prolonged postoperative antimicrobial prophylaxis focusing on emergency abdominal surgery was very common in many healthcare settings during the same period [22], although it was limited by the guidelines issued by the Chinese Surgical Society of the Chinese Medical Association [20] and rarely supported or reported in the national literature. In the multicenter, prospective, cross-sectional study of adult patients who underwent emergency abdominal surgery in 47 tertiary hospitals in China during 1-2 months from 2018 to 2019 [22], only 3.46% patients were admitted with (less than) 1-day perioperative prophylactic antibiotics. A Cochrane review also showed no benefits for long-term prophylaxis in elective surgery [23]. Furthermore, according to the CDC and WHO guidelines for the prevention of SSI and relevant recommendations according to available guidelines, the panel recommends against the prolongation of surgical antibiotic prophylaxis administration after completion of the procedure for the purpose of preventing SSI [1,19]. By contrast to the previous "guidelines" in China on surgical antibiotic prophylaxis prolongation [24], the latest evidence-based document published after the study provides a similar recommendation in line with international practices. We identified a level III surgical incision as the only independent risk factor for SSI as it is more exposed to a contaminant and thus a high risk of infection. We observed that the duration (≥ 117 min) of the surgical procedure was related to the occurrence of SSI. Complex procedures are often long and thus the incision and the wound are more exposed to the risk of SSI. Sattar et al. [6] reported that SSIs were more likely in procedures with a duration greater than 90 min and particular attention should be paid to the wounds of patients undergoing a lengthy procedure. The serum albumin level has an important role in the development of SSI, but is not well evaluated in elderly patients. Liang et al. [25] also reported that the rate of low albumin levels was 35.1% in SSI patients. In particular, a low preoperative albumin level (< 35 g/L) was associated with a 2.3-fold increased risk of SSI. In our study, we did not test the preoperative albumin level, but we analyzed the effect of the first day of postoperative albumin on SSI and demonstrated that a level of < 25 g/L was significantly associated with SSI in emergency bowel surgery, potentially leading also to tissue edema, a delay of incision recovery, and an increase in the risk of SSI. In a prospective cohort study, a decreased concentration of serum albumin ≥ 10 g/L on postoperative day 1 was associated with a threefold increased risk of overall postoperative complications [26].
WHO recommends that bathing or showering prior to surgery can reduce SSI by decreasing the bacterial load on the skin [19]. However, this is not always possible for emergency patients. We were unable to determine if a history of chronic disease, such as diabetes mellitus, cardiovascular disease, or others, had any adverse influence on SSI. However, several observational studies have shown that hyperglycemia was related to SSI [27] and it would be relatively simple for our surgical teams to comply with the WHO recommendation to control blood glucose to reduce the risk [19]. No deaths occurred among our patients in the 30 days following emergency surgery. Gomila et al. [28] reported that the overall 30-day mortality caused by organ-space SSI was 8.9% (30/336) in elective colon and rectal surgeries, and these findings suggest that more attention should be paid to the prevention of deep SSI to help decrease mortality.
Our study has some limitations. First, this is a singlecenter study and data collection might be biased and some confounding factors may have been generated due to its retrospective nature, e.g., lack of documentation of antimicrobial exposure and previous hospitalization history. Second, it was difficult to determine the beginning of the actual time of antibiotic administration as this information was not well documented for emergency patients. For this reason, we were unable to assess the optimal timing of preoperative surgical antibiotic prophylaxis, although the optimal timing is still a subject of debate [12,29]. Third, underlying risk factors, such as blood transfusion, hypothermia, and preoperative albumin level, were not assessed. Fourth, perioperative management against SSI in this study was not based on evidence-based guidance [8] and the implementation of a bundle strategy for the prevention and management of SSI was not introduced during the study period [19,30].

Conclusion
The incidence and burden of SSIs were high in bowel obstruction patients over 60 years of age who underwent emergency surgery, including the prevalence of ESBLs isolated from SSIs, and highlight the need to rapidly implement preventive strategies to improve quality of care. A level III surgical incision was identified as an independent risk factor for SSI.