We used criteria of the US Centers for Disease Control and Prevention (CDC) NNIS System to diagnose SSI. Infections were classified as superficial incisional, deep incisional, or organ/space SSI 20. The ASA score was used to characterize the patients' physical status as 1 (healthy), 2 (mild systemic disease), 3 (severe systemic disease), 4 (severe life-threatening systemic disease), or 5 (moribund) 21. Patients' final diagnoses and operations were coded according to the International Classification of Diseases 10th Revision (ICD-10) and the International Classification of Diseases 9th Revision, Clinical Modification (ICD-9 CM), respectively. The appendectomy procedures were also classified according to the NNIS 22.

We defined uncomplicated appendicitis as acutely inflamed appendicitis without perforation (clean-contaminated wound). Patients with gangrenous appendicitis, peritonitis, or abscess formation were not included.

The rate of SSI was computed by dividing the number of infections by the number of operations performed and multiplying by one hundred.

Contingency tables were constructed to analyze the relations between SSI and the other study variables: use and duration of antibiotic prophylaxis, sex, age, length of preoperative stay, type of operation, ASA score, and duration of operation. We conducted logistic regression analysis to estimate the relative risk (RR) of SSI for the main study variables.

All analyses were performed using STATA statistical software, version 7 (Stata Corp, College Station, TX).

During the study period, 2139 patients, 53.1% of them women, underwent open appendectomy. The median age was 26 years (interquartile range 16 to 39). Twenty-one percent of patients were hospitalized preoperatively with a median length of preoperative stay of 1 day (interquartile range 1 to 1). For postoperative or total hospital stay, median length was 3 days (interquartile range 2 to 4). Among 2139 operations, 72.4% were classified as emergency. The median operation duration was 58 minutes (interquartile range 42 to 83).

Twenty-six SSIs were identified in 2139 operations, yielding an overall SSI rate of 1.2 infections/100 operations. Superficial and deep incisional SSIs occurred most frequently (46.2% each). Of the 26 SSIs, 15 (57.7%) were detected after hospital discharge, and a half, within seven days after surgery. The median onset of SSI was 8 days (interquartile range 5–11).

The following variables were associated with the risk of SSI in the crude analyses: duration of antibiotic prophylaxis, age, elevated ASA score, prolonged preoperative hospital stay, duration of operation, emergency surgery, and sex. However, after adjustment, only prolonged duration of operation remained significantly associated with an increased risk of SSI (RR = 3.29; 95% CI 1.44–7.52) (Table 1).

Prophylactic antibiotics were administered in the course of 1972/2139 (92.2%) operations. In 89.8% of these cases, antibiotics were given preoperatively, but they were given within one hour before the incision in only 38.9% of cases. The most common prophylactic antibiotics were metronidazole and gentamicin (64.2%). The combination of a single dose of metronidazole and gentamicin was used in 38.8% of cases. Antibiotic prophylaxis was administered for one day in 54.2% of cases. The median duration of antibiotic prophylaxis was 1 day (interquartile range 1 to 2). Antibiotic prophylaxis was extended for over one day in 38.0% of patients (Table 2).

The doses of antibiotic prophylaxis administrated in this study were metronidazole 500 mg, gentamicin 80 mg, penicillin 2 million units, amoxicillin 500 mg, amoxicillin/clavulanate 1.2 g, ampicillin 1 g, cloxacillin 1 g, cefazolin 1 g, cephalexin 1 g, cefotaxime 1 g, ceftriaxone 2 g, ceftazidime 1 g, cefdinir 100 mg, cefoxitin 1 g, ofloxacin 200 mg, norfloxacin 400 mg, cotrimoxazole 480 mg, chloramphenical 1 g, fosfomycin 1 g, clindamycin 600 mg, and amikacin 500 mg.

Antibiotic prophylaxis was associated with decreased risk of SSI, while timing of administration – preoperatively vs. intraoperatively – had no effect on risk. Compared with no antibiotic prophylaxis, receiving one-day, or multiple-day antibiotic prophylaxis was each associated with about one third of the SSI risk, adjusted RR were 0.30 (0.11–0.88) and 0.32 (0.10–0.98), respectively (Table 3). This reduced risk was also found when receiving only a single dose of metronidazole in combination with gentamycin (adjusted RR = 0.28 (0.09–0.90)), Data on duration, administration time, and antibiotic agent associated with SSI, adjusted for age, sex, ASA score, and duration of operation, are shown in Table 3.

The strengths of this study are its large sample size and prospective cohort design with complete early follow-up. Furthermore, we were able to include all patients admitted with uncomplicated appendicitis, who underwent open surgery in all of the eight hospitals.

Among the study limitations was the failure to account for differences in surgical technique, preoperative and postoperative practices. However, appendectomies are performed quite uniformly in Thailand and we consider large variation in these practices unlikely. Owing to high cost of post-discharge surveillance, we were unable to follow all patients for 30 days after surgery. Curtailed follow-up may cause underestimation of the SSI rate. Still, the SSI rate in our study (1.2%) was similar to that observed in the NNIS system report (1.3%) 2. Some SSIs were potentially misclassified when the exact layer of tissue or organ/space involved in the infection was unclear. To limit such misclassification, an expert was consulted in these cases. In addition, the surgeon and at least one person from the infection control team had to agree to the diagnosis and classification of SSI in all cases.

A prolonged duration of operation has been reported as a risk factor for SSI in other studies 12324. Earlier investigations have also reported increasing age 25, and emergency surgery 26 as risk factors of SSI. Our study did not show substantial association between these factors and the risk of SSI. Under-accounting for other predictors of SSI not included in our analysis might explain our failure to observe an association between age or emergency surgery and SSI.

Twenty-one different antibiotic agents were administered to appendectomy patients in our study, highlighting the lack of consensus among Thai surgeons in prescribing practices. Metronidazole plus gentamicin were most commonly used agents, and our study confirmed their effectiveness in reducing SSIs reported by others 1819. The prophylaxis was effective despite the fact that timing of prophylaxis followed international guidelines in only 39% of the cases 2728. This finding corroborates the notion that the timing of the administration – pre-, intra- or post-operation – may not be crucial for preventing SSIs 1. At the same time, higher SSI rate was reported among patients who received antibiotic prophylaxis after surgical incision 29, probably because the antibiotic serum concentration at the surgical closure is strongly associated with SSI 30.

The American Society of Health System Pharmacists (ASHP) 27 recommends prophylaxis with cephalosporins for uncomplicated appendicitis 2728, with metronidazole and gentamicin only considered an alternative in cases of penicillin allergy. However, the combination of metronidazole plus gentamicin may have an economic advantage. A Thai study indicated that the estimated cost for these combined agents was 210 baht per 24 hours, compared with 1160 baht for cefoxitin 31. Adverse effects were not documented for our patients.

Extended duration of antibiotic prophylaxis was less frequent in our study than in a Malaysian study 32, but our finding is consistent with other studies conducted in France 8 and Spain 9.

The improper use of antibiotic agents and inappropriately prolonged duration of antibiotic prophylaxis are likely to cause antimicrobial resistance 333435363738. Surgeons and surgical departments need to update their practices of antibiotic prophylaxis to comply with standard guidelines 2728 and updated evidencebase 1.

Antibiotic prophylaxis is associated with a decreased risk of postoperative SSIs. This finding is in agreement with other studies 4141516171819. In addition, our data indicate that in uncomplicated appendicitis cases, one-day antibiotic prophylaxis is just as effective in reducing SSIs as multiple-day antibiotic prophylaxis. Thus, our study does not offer justification for routine administration of oral antibiotics upon hospital discharge as in a previous study 39. Furthermore, our findings correspond to findings in other reports 2740, a single-dose antibiotic 40, such a combined a single dose of metronidazole and gentamicin, was efficient to reduce SSIs in uncomplicated appendicitis 27.