In our study, more than 33% of patients > 65 years old. Advanced age is a risk factor for postoperative infection and malnutrition [15,16,17,18,19]. Most surgeries in the study were classified as clean-contaminated (89.3%) procedures. According to Altemeier’s surgical wound classification based on the risk of surgical infection, the risk of postoperative infection was as follows: 1–5% (clean incisions), 5–10% (clean-contaminated incisions) . A cohort study showed that the rate of postoperative infection was 5.9% in cases of clean open surgery and 46.4% and 3.2% in open and laparoscopic clean-contaminated surgery, respectively .
The appropriateness of prophylactic antibiotic indication in our study was 97.2%, which is higher than that reported by Malavaud et al. , Durando et al. , and Pittalis et al. , who observed rates of 88.1%, 70.3% and 95%, respectively. This finding may be attributable to the increasing prevalence of nosocomial infections and drug resistance in Vietnam and the greater attention to postoperative infectious complications and the use of preoperative antibiotic therapy. Most patients in our study received prophylactic cefoxitin (64.2%) and metronidazole (21.7%). Cefoxitin is a 2nd generation broad-spectrum cephalosporin with a wide range of anaerobic coverage and is therefore commonly indicated for prophylaxis in gastrointestinal procedures, such as cholecystectomy, colorectal surgery, and operations for small bowel obstruction [21, 22]. Metronidazole and cefazolin combination therapy is recommended for operations with a high risk of anaerobic bacterial contamination, such as biliary tract surgery, laparoscopy, laparoscopic appendectomy, operations for small bowel obstruction, and colon surgery . Statistical evidence suggests that cefoxitin and metronidazole constitute a high percentage of prophylactic antibiotics administered for gastrointestinal surgery. Most patients were prescribed a single prophylactic antibiotic (76.3%). Commonly used prophylactic antibiotics include cefoxitin (49.5%), ceftriaxone (6.2%), and cefazolin (9.3%).
We observed that the choice of antibiotics used was appropriate in 68.6% of patients in this study. This rate is lower than that reported by Malavaud et al.  (91.9%) and Pittalis et al.  (84.5%), which is attributable to the differences in hospital practices and policies adopted in our study and those followed by the hospitals included in the aforementioned studies. In fact, prophylactic combination therapy using cefoxitin and metronidazole is widely used during gastrointestinal surgery for effective coverage against anaerobic bacteria, although this combination is not reasonable [24,25,26]. A systematic review of 18 studies that investigated adherence to guidelines for antibiotic prophylaxis observed that the rates of appropriateness of antibiotic prophylaxis ranged from 22 to 95% .
We observed that prophylactic antibiotic usage was 100% within 30–60 min before creation of the skin incision. Current guidelines recommend that prophylactic antibiotics should be administered within 60 min preoperatively and close to the time of creation of the skin incision (120 min for vancomycin and fluoroquinolones) [28, 29]. The timing of administration is an important determinant of successful antibiotic prophylaxis. Microbes that cause postoperative infection primarily invade the body during surgery (between the time of creation of the skin incision and wound closure) . In a study performed by Koch et al. , the postoperative infection rate was lower in those who received prophylactic antibiotics within 60 min before the skin incision than in those who received prophylactic antibiotics later than 60 min before the skin incision. Several studies have reported that the effectiveness of antibiotic prophylaxis is maximum when administered during the 60-min period before creation of the skin incision [23, 24]. A study by Weber et al.  showed that the postoperative infection rate was higher in those who received prophylactic antibiotics between 0 and 30 min before the skin incision than in those who received prophylactic antibiotics over approximately 30–60 min prior to creation of the skin incision. In contrast, Steinberg et al.  observed no difference in the postoperative infection rate in patients who received prophylactic antibiotics within 0–30 min or 30–60 min.
No patient in our study received additional doses intraoperatively. Based on the clinical practice guidelines for antimicrobial prophylaxis in surgery, patients in whom the surgical time is > 2 times the T1/2 of the drug and those with intraoperative blood loss > 1500 mL in adults tend to require additional antibiotic prophylaxis to ensure adequate blood and tissue antibiotic concentrations [4, 20]. In our study, we observed that the surgical time did not exceed 120 min in patients who received cefoxitin; therefore, the interval between repeat doses of cefoxitin was not exceeded. Moreover, repeat dose intervals of ceftriaxone (6 h), cefazolin (4 h), and metronidazole (6 h) were longer than the surgical time. Additionally, intraoperative blood loss > 1500 mL did not occur in any patient in our study (based on patients’ surgical record). Therefore, repeat doses of prophylactic antibiotics were not required. The rate of appropriate dose supplementation in this study was 100%, which was higher than the rates reported by Goede et al.  (54.9%).
Prophylactic antibiotic usage beyond 24 h after treatment completion was observed in 52.6% of patients, of which 12.4% patients continued treatment without a justifiable indication. A study performed across 9 provincial and central hospitals in 2009 reported that 94.6% of patients received prolonged postoperative prophylactic antibiotics , which may have been due to the notion prevalent among surgeons regarding the need for long-term administration of antibiotics to prevent postoperative infections. However, per recommendations, most routine surgeries usually require administration of only a prophylactic dose. Antibiotic prophylaxis is not required beyond 24 h postoperatively for most surgeries (or beyond 48 h for cardiac surgery) . The appropriateness of the rate of prolonged antibiotic prophylaxis in our study was 87.6%, which was higher than that reported by Pittalis et al.  (48%). Gouvea et al.  showed that the reasonable rate of duration of antibiotic prophylaxis ranged from 5.8 to 91.4%.
The overall rate of appropriateness of antibiotic prophylaxis in our study was 54.7%, which is similar to that reported by Malavaud et al.  (58.3%). The similarity in results may be attributed to the fact that the types of surgeries performed and the criteria for criteria for antibiotic selection were nearly identical between our study and the study performed by Malavaud et al. .
Postoperative antibiotic use
In this study, the rate of appropriate postoperative antibiotic type was 51.3%, and combination antibiotic therapy (dual anaerobic coverage therapy) resulted in inappropriate antibiotic treatment. The overall rate of appropriate antibiotic treatment was 38.5%. Culture or antibiogram were not performed in any patient in this study, and treatment was based exclusively on experience. This is one of the limitations associated with the use of antibiotics in the Department of Surgery, which should be addressed in future research. Adherence to recommended guidelines is important to reduce inappropriate antibiotic use and to improve treatment effectiveness, to benefit both patients and hospitals.
Surgical site infection is one of the most common infections associated with health care. Therefore, there are many studies to investigate postoperative infection, including surgical site infection after gastrointestinal surgery. In this study, 34.8% of patients had at least one sign of infection, which is higher than that reported by Bhangu et al. , Wang et al. , Alkaaki et al. , who reported rates of 2–5%, 5.2% and 16.3%, respectively. This difference can be explained as follow, we recorded cases of patients with at least one sign of infection, while other studies recorded the proportion of patients with accurate diagnosis of surgical site infection.
The median length of hospitalization was 14 (10–20) days, which is similar to that reported by Jakobson et al.  (14.5 ± 10 days) and higher than that reported by Alkaaki et al. , the median postoperative hospital stay was 2 days for patients without SSI, compared to 13 days for those with SSI.
We observed that increasing age increased the risk of postoperative, which is similar to findings reported by Neumayer et al. , who observed a higher risk of postoperative infection in patients aged over 40 years old than in those aged under 40 years. A study performed in the United States showed that the risk of postoperative infection increased 1.2-fold with every 1-year increase in age among patients aged above 65 years .
Furthermore, we observed that advanced age was associated with longer length of hospitalization; this result is similar to that reported by Chong et al. , who observed that advanced age was associated with the longer length of hospitalization in patients who underwent laparoscopic cholecystectomy.
Following are the limitations of this study: (a) This retrospective observational study was based on data obtained from patients’ medical records; therefore, we did not consider all potential confounders that may have affected our results. (b) Owing to the small sample size and single-center study design, our findings may not accurately reflect the effect of prophylactic and therapeutic antibiotics in patients who undergo gastrointestinal surgery. Further large-scale studies are warranted to gain a deeper understanding of this subject.