Background
Due to the clean-contaminated nature of the wound, rates of surgical site infections (SSI) after colorectal surgery are the highest among elective procedures, exceeding 20% in some institutions [
1‐
3]. It has been suggested that the rates and risk factors for developing an SSI after colon and rectal surgery may be different [
4,
5], due to the differences found in the surgical approach and the degree of bacterial contamination between both surgeries. Nevertheless, most studies carried out to date have analysed colon and rectal surgeries together [
6,
7]. Separate assessments of patients undergoing colon and rectal surgery are scarce [
4,
8].
It has been proposed that incisional SSI (I-SSI) and organ-space SSI (OS-SSI) may have distinct pathogenesis and risk factors. Incisional SSI has been associated with increased body mass index or the presence of an ostomy [
6,
9]. On the other hand, OS-SSI has been more frequently related to blood transfusion, previous abdominal surgery or poor nutritional status [
6,
7,
10]. Interestingly, the development of an OS-SSI has more severe consequences than the development of an I-SSI; in many cases OS-SSI requires reoperation and increases morbidity and length of stay (LOS) [
11,
12]. Moreover, while many of the most significant advances in colon and rectal surgery such as laparoscopy and other minimally invasive techniques have decreased I-SSI rates, they have had a lesser impact on OS-SSI [
13,
14].
Remarkably, the administration of mechanical bowel preparation (MBP) was discontinued in the last decades in most Spanish hospitals due to the lack of effectiveness [
15]. In this scenario, and for reasons not well established, the administration of oral antibiotic prophylaxis (OAP) was discontinued too. Currently, only some hospitals use it in the elective surgery of the colon and rectum in Spain. This situation contrasts with that of other European and American countries, where the OAP is part of the daily practice.
The aim of this study was to compare the incidence, risk factors and outcomes of OS-SSI in patients undergoing elective surgery of the colon or rectum in a large, representative cohort of Spanish hospitals.
Discussion
This large multicentre cohort study found significant differences in the incidence, predictive factors and outcomes of OS-SSI after elective colon and rectal surgery. This suggests that the two procedures should be considered as different surgical interventions.
The separation of procedures according to patients’ characteristics may allow more accurate assessment of their specific risk factors. Comparing colon and rectal populations, we found that they had different characteristics in terms of risk factors for SSI. Patients undergoing colon surgery were older, had more IBD and less laparoscopy, factors related to SSI. On the other hand, patients undergoing rectal surgery were younger but had more rate of malignancy; more frequently received chemoradiotherapy and had longer surgery duration. The surgical techniques were also different, something inherent to the anatomical location of the disease, in special with more ostomies performed in rectal resections. These factors, associated with the fact that the rectum has higher bacterial contamination load, conferred it greater risk of SSI. Accordingly, overall SSI and OS-SSI rates were higher in rectal surgery than in colon surgery. Although these rates were high, they were similar to these reported in previous studies [
8,
21]. Data from surveillance systems in Europe an US vary widely [
22,
23], being in most cases lower than ours, though post-discharge surveillance is not always performed.
We found significant differences in the predictive factors for developing an OS-SSI in colon and rectal surgeries. In colon surgery, independent risk factors predisposing to OS-SSI were male sex and ostomy creation, while laparoscopic surgery and OAP were protective factors. In rectal surgery, independent risk factors for OS-SSI were male sex and longer duration of surgery, whereas OAP was the only protective factor. Male sex was a common risk factor for developing OS-SSI in both colon and rectal surgeries; this association is well established [
5,
7,
24], although the reasons are not known.
Ostomy creation was a strong risk factor for the development of OS-SSI in colon surgery but not in rectal surgery, as previously reported elsewhere [
8]. Ostomies are normally used to divert the faecal stream from a newly created immature anastomosis, or to definitively disconnect the gastrointestinal tract in some extensive colorectal surgeries. Nevertheless, ostomies have been associated with increased rates of SSI in previous studies [
4‐
6,
9] because they allow organisms from the air, contaminated hands, or skin flora to reach the subcutaneous fat and the wound, and eventually the intraabdominal cavity [
25]. In our study, patients with colon surgery who received an ostomy more frequently underwent laparotomy due to complex pathology like IBD or diverticulitis. These diseases have been associated with OS-SSI [
26], and ostomy creation may act, in part, as a marker of this complex pathology.
The laparoscopic approach significantly reduced SSI rates in several large-database studies and also offered other benefits such as faster recovery of pulmonary function, less pain and shorter postoperative stay [
13,
14]. In our study it served as an independent protective factor for the development of OS-SSI in colon surgery, but not in rectal surgery. Probably, the beneficial effect of laparoscopy was exceeded by the higher frequency of risk factors for SSI inherent in rectal surgery.
Importantly, we found that OAP was a protective factor for the development of OS-SSI in both colon and rectal surgeries, although the impact was higher in rectal surgery, probably because the rectum has a higher level of bacterial contamination. During the study period there was not a national or regional recommendation for the application of OAP, and for this reason the use of the measure was decided by each participating hospital (it was only applied in 4 of the 10 hospitals). The findings of the present study lead to a change in the clinical practice of hospitals participating in the VINCat program and in 2016 the use of OAP was institutionally recommended. The OAP combined with intravenous prophylaxis and MBP significantly reduces SSI rates after colon and rectal surgery by decreasing the intraluminal bacterial load [
27‐
30]; in a previous meta-analysis of randomized controlled trials comparing the effectiveness of OAP plus intravenous antibiotic prophylaxis
vs intravenous antibiotic prophylaxis alone, the association of OAP was estimated to reduce the incidence of SSI by 43% [
31]. Nevertheless, the use of MBP has been widely questioned, due to its unpleasant gastrointestinal effects, and in many studies it has failed to reduce SSI rates [
15]. Currently, since almost all studies that demonstrate the effectiveness of OAP have been performed in combination with MBP, the use of MBP will have to be raised again. Last World Health Organization (WHO) recommendations on preoperative measures for surgical site infection prevention suggest using OAP with MBP in all adults undergoing elective colorectal surgery [
32,
33].
Longer duration of surgery was an independent risk factor for the development of an OS-SSI in rectal surgery. This association has often been described in the colorectal surgery population [
21,
34,
35], and it also favours other risk factors for SSI like the hyperglycaemia or hypothermia [
33]. Given the capacity of this parameter to predict SSI, it was included as one of the components of the NNIS risk index. Rectal tumours close to the anal verge usually require extensive surgery with additional organ resection, requiring longer operative time and causing greater bleeding, factors that have been associated with an increased risk of SSI [
24,
36]. Moreover, in these prolonged surgeries, antibiotic redosing is not always administered correctly.
Significantly, mortality of patients with OS-SSI after colon surgery was higher than after rectal surgery. The fact that patients in the colon group were older and more frequently had complicated diseases other than neoplasia could explain this result.
Among the strengths of the study is its multicentre nature, the large number of patients included and the fact that all data were collected by trained infection control staff. However, the study has a number of limitations that should be acknowledged. Firstly, the retrospective analysis of prospectively collected data may lead to bias and is unable to control for confounding factors. Secondly, certain risk factors that have been linked to SSI such as perioperative hyperglycaemia, hypothermia and blood transfusion were not recorded here.
Acknowledgments
We want to thanks the other collaborators at Hospital Universitari de Bellvitge, Doménico Fraccavieri, Rosa Escofet, at Corporació Sanitària Parc Taulí, Xavier Serra-Aracil, Marta Piriz, at Parc Sanitari Sant Joan de Déu de Sant Boi, Vicens Diaz-Brito, Encarna Moreno, at Hospital Universitari Sant Joan de Reus, Antoni Castro, Simona Iftimie, at Consorci Sanitari de Terrassa, Elena Espejo, Lurdes Pagespetit, at Consorci Sanitari de l’Anoia, Montse Brugués, at Fundació Althaia, Fina Obradors, at Hospital de Viladecans, Ana Lérida, Camilo Sanz, at Fundació Privada Hospital Asil de Granollers, Jordi Cuquet, Nares Arroyo and at Hospital Universitari Mútua de Terrassa, Albert Navarro, Núria Freixas.
Writing assistance was provided by the Department of Languages in the Faculty of Medicine of Barcelona University.