Int J Colorectal Dis DOI 10.1007/s00384-016-2662-y
REVIEW
The role of oral antibiotics prophylaxis in prevention of surgical site infection in colorectal surgery Michalis Koullouros 1 & Nadir Khan 1 & Emad H. Aly 1,2
Accepted: 19 September 2016 # Springer-Verlag Berlin Heidelberg 2016
Abstract Background Surgical site infection (SSI) continues to be a challenge in colorectal surgery. Over the years, various modalities have been used in an attempt to reduce SSI risk in elective colorectal surgery, which include mechanical bowel preparation before surgery, oral antibiotics and intravenous antibiotic prophylaxis at induction of surgery. Even though IV antibiotics have become standard practice, there has been a debate on the exact role of oral antibiotics. Aim The primary aim was to identify the role of oral antibiotics in reduction of SSI in elective colorectal surgery. The secondary aim was to explore any potential benefit in the use of mechanical bowel preparation (MBP) in relation to SSI in elective colorectal surgery. Methods Medline, Embase and the Cochrane Library were searched. Any randomised controlled trials (RCTs) or cohort studies after 1980, which investigated the effectiveness of oral antibiotic prophylaxis and/or MBP in preventing SSIs in elective colorectal surgery were included. Results Twenty-three RCTs and eight cohorts were included. The results indicate a statistically significant advantage in preventing SSIs with the combined usage of oral and systemic antibiotic prophylaxis. Furthermore, our analysis of the cohort studies shows no benefits in the use of MBP in prevention of SSIs.
* Emad H. Aly
[email protected] 1
School of Medicine, University of Aberdeen, Foresterhill, Aberdeen AB25 2ZN, UK
2
Laparoscopic Colorectal Surgery and Training Unit, Aberdeen Royal Infirmary, Aberdeen, Scotland, UK
Conclusions The addition of oral antibiotics to systemic antibiotics could potentially reduce the risk of SSIs in elective colorectal surgery. Additionally, MBP does not seem to provide a clear benefit with regard to SSI prevention. Keywords Antibiotics . Prophylaxis . Wound site infection . Colorectal surgery
Introduction The development of surgical site infections (SSIs) continues to be a challenge following colorectal surgery. SSIs have been shown to prolong the patients’ length of stay almost threefold, hence increasing costs of treatment of up to fourfold [1]. This is an important factor to consider as most, if not all, healthcare systems are under increasing financial pressure. Various strategies have been adopted in an attempt to reduce SSIs from occurring following colorectal surgery, such as mechanical bowel preparation (MBP), oral antibiotic prophylaxis, intravenous (IV) antibiotic prophylaxis and different combinations of these have been examined in clinical studies [2, 3]. There has been consistent evidence for the usefulness of antibiotic prophylaxis. A meta-analysis [4] comparing outcomes from 26 RCTs, showed significant improvement in the SSI rates and other secondary outcomes due to the use of different regimens of antibiotic prophylaxis. Their conclusion was that it is unnecessary and unethical to run any more antibiotic vs. control trials [4]. However, there is still uncertainty on the exact role of oral antibiotics when used alone, in addition to intravenous antibiotics with or without bowel preparation. It has been shown that oral antibiotics have an additional benefit in SSI prophylaxis, by decreasing the number of
Int J Colorectal Dis
aerobic and anaerobic micro-organisms in the colon [5], as well as an increased likelihood for a SSI developing in patients not receiving them [6]. Nonetheless, there is some controversy as to whether oral and systemic antibiotics offer additional benefit. To be precise, most evidence indicates that the combination of oral and systemic antibiotics induces a significant decrease in the rate of SSIs when compared with systemic alone [7–10]; however, some studies conclude that this combination carries no benefit with regard to SSI prevention [11]. Current practice in many centres involves an intravenous antibiotic only given at the induction of surgery, to reduce the risk of SSIs. MBP has been shown to provide little or no benefit in SSI prevention [12, 13], and therefore, its use has decreased over the recent years. Guidelines do not recommend MBP to be used as a means to prevent SSIs prior to elective colorectal procedures [14]. However, recent publications looking at the use of pre-operative prophylactic regimens in elective colorectal surgery [15–19], have shown that MBP is still being used, leading to cohort studies investigating MBPs potential benefits with regard to SSI. In the absence of clear evidence from all the available studies to define the exact role of oral antibiotics in reducing SSIs in elective colorectal surgery, a systematic review was conducted and presented in this paper. The goal of such an analysis is to distil the literature for the reader so they have a clear impression of what should be best practice. The primary aim was to identify the role oral antibiotics in reduction of SSI in elective colorectal surgery. The secondary aim was to explore any potential benefit in the use of MBP in relation to SSI in elective colorectal surgery.
Methods Search method The following electronic databases were searched: Medline (1980 to 1st week of October 2015), Embase (1980 to 5th of October 2015), Cochrane’s CENTRAL and Cochrane Database of Systematic Reviews (CDSR). The appropriate Medical Subject Heading (MeSH) terms were used for each of the databases and limited our search to publications related to humans and the English language (Appendix 1). Additionally, more relevant publications were added as identified from results’ references, and concurrent searches in PubMed and Ovid.
prophylaxis in the prevention of postoperative surgical wound infections in patients undergoing colorectal surgery. These studies included comparison between oral antibiotics vs. placebo, oral and IV antibiotics vs. IV alone, oral antibiotics and MBP vs. MBP alone or oral antibiotics alone and combinations of these comparisons. Quality assessment Two reviewers (MK and NK) appraised each of the studies gathered from our database search individually using the Scottish Intercollegiate Guidelines Network (SIGN) checklists, using the appropriate checklist depending on the study type. For any studies for which the two reviewers disagreed on the quality, a decision was made after discussion and use of the SIGN checklist notes. Additionally, the RCTs were appraised for bias using the Cochrane risk of bias criteria, and the cohorts were appraised using the Newcastle-Ottawa scale. Inclusion and exclusion criteria Inclusion criteria were a minimum of an ‘acceptable’ quality as per the SIGN checklists, comparison of oral antibiotics either on their own or in combination with another prophylactic strategy to another pre-operative protocol and SSI rates as a measurable outcome. Exclusion criteria were any study that compared various antibiotic prophylactic regimens without a control group that did not receive prophylactic antibiotics, studies in which patients had received antibiotics for any other reasons or were on an ongoing course of antibiotics that was taken more than 5 days prior to the procedure and studies performed before 1980 to keep the results more relevant to current practice.
Definitions The following definitions were used when analysing the results in the study. Mechanical bowel preparation MBP is an oral preparation given before surgery to clear faecal material from the bowel tract. A variety of preparations were included in this review including polyethylene glycol, mannitol and sodium picosulphate [20].
Types of studies Randomised controlled trials (RCTs) and cohort studies were included, which assess the effectiveness of oral antimicrobial
Surgical site infection SSI is expressed at three levels—superficial incisional, deep incisional or organ infection—and is defined by expression of pus or by the presence of local signs
Int J Colorectal Dis
of inflammation such as temperature, pain, redness, swelling and separation of the edges of the incision. [14]. SSI rates In this study, SSI rates refer to a number of patients who get a surgical site infection from the total number of patients in a particular category, and this is expressed as a percentage.
given before this, then the study was classified as ‘before MBP’, if oral antibiotics were given at the same time the study was classified as ‘during MBP’, and if oral antibiotics were given after MBP initiation the study was classified as ‘after MBP’ [20].
Data extraction Elective colorectal surgery Elective colorectal surgery is any surgery undertaken on a planned basis for any condition of the colon or rectum requiring bowel resection and primary anastomosis. This includes colorectal carcinoma, diverticular disease and inflammatory bowel disease. Colectomy When looking at the surgical techniques, we classified left, right, transverse, sigmoid, subtotal and total colectomies or hemicolectomies in the ‘colectomy’ column.
Two reviewers (MK and NK) gathered the following categories of data from all the studies that passed quality assessment and inclusion and exclusion criteria: title, first author’s last name, sample size, blinding, randomization, number of participants withdrawn, types of oral antibiotics used, any additional prophylactic measure to the oral antibiotics, the comparison group to the oral antibiotics, type of surgery and rates of surgical site infection. Types of participants
Timing of administration When deciding when the oral antibiotics were administered, or started to be administered, in relation to MBP, we regarded the administration of any laxative or enema as the initiation of MBP. If oral antibiotics were
Fig. 1 PRISMA flowchart
Participants include patients of all ages undergoing elective colorectal surgery, in whom infection was not suspected preoperatively.
Int J Colorectal Dis Fig. 2 Oral vs. IV (RCTs)
Types of interventions All antibiotic prophylaxis regimens delivered orally to prevent postoperative infection were considered. Studies where the antibiotics were administered preoperatively due to a suspected or confirmed infection were excluded. Types of outcome measures For the purpose of this review, the rate of surgical site infection was used to be the outcome measured to evaluate the effectiveness of oral antibiotic preparations in colorectal surgery. Most studies required pus expression to define a SSI, in concordance with the pre-mentioned definition.
Cohort studies based on incidence of SSI events per study group. Data were analysed as forest plots using Review Manager 5.2 software.
Results The search strategy used yielded 23 RCTs [8, 9, 11, 21–40] and 8 cohort studies [15–19, 41–43] after screening and applying exclusion criteria (Fig. 1). Overall, there were 5325 participants in the RCT analysis and 58,107 participants in the cohort studies. The data from RCTs is presented as a meta-analysis outlined in Figs. 2, 3, 4, and Table 1, and the Cohort studies are presented in terms of risk ratios shown in Figs. 5, 6, 7, and as percentages in Table 1.
Statistical methods Quality assessment We investigated the relationship between different antibiotic preparation strategies using odds ratios (OR) and confidence intervals (CI, 95 %). The odds ratios were calculated for each RCT based on incidence of SSI events per study group expressed as rates. The risk ratio (RR) was calculated for
Fig. 3 Oral + IV vs. IV (RCTs)
The RCT quality assessments according to SIGN checklist are summarised in Table 2. Five of the RCTs were single blinded [21, 22, 24, 27, 28], four were double blinded [9, 23, 26, 39] and thirteen were not blinded [8, 11, 25, 29–37, 40]. In one
Int J Colorectal Dis Fig. 4 Oral vs. oral + IV (RCTs)
study [38], blinding was not appropriate, since the two study groups were not investigated at the same time. Twenty-one studies had groups which were randomised adequately [8, 9, 11, 21–32, 34–37, 39, 40]; however, two studies [33, 38] were non-randomised trials. The RCT risk of bias have been assessed as per the Cochrane risk of bias checklist (green: low risk, red: high risk, white: unclear risk) (Figs. 8 and 9). The risk of bias in the cohort group of studies has been assessed as per the Newcastle-Ottawa checklist (Figs. 10). There was significant variation with regard to type, dosage and duration of antimicrobial prophylaxis. Overall, there was a low participant withdrawal from the RCTs included in the analysis. However, in the case of Condon et al. [24], there was a large withdrawal (75 %) from the initial recruited sample. This is because the study initially included patients who were later deemed to be inappropriate for inclusion, regardless of their compatibility with their study protocol for a variety of reasons. Seven RCTs disclosed a funding source for their trials that are indicated in Table 2. The definition of SSIs was similar between studies and agreed with definition stated by NICE. RCT patient characteristics Overall, the patients’ age and sex were similar across the different RCTs with the exception of one study [29] where the participants of the study participants were significantly younger due to study focusing on patients with ulcerative colitis. It was not possible to determine an overall age average since studies reported the data in various formats (Table 3). Three of the RCTs included [8, 28, 39] were threearm trials whereas twenty RCTs [9, 11, 21–27, 29–38, 40] were two-arm trials within the scope of this review. Eleven studies compared oral vs. IV [8, 21, 22, 26, 28, 31, 34, 36, 37, 39, 40], twelve studies compared oral + IV vs. IV [8, 11, 27–30, 32, 33, 35, 37–39] and three studies compared oral vs. oral + IV [23–25]. All participants in the RCTs included in the meta-analysis received elective colorectal operations. Most RCTs used
antibiotic regimens that covered both aerobic and anaerobic bacteria. Cohort studies patient characteristics Overall, the patients’ age and sex were similar across the different Cohort studies (Table 4). With the exception of one study [15] where the participants of the study were mainly male as the study was conducted by veteran’s association medical research service. It was not possible to determine an overall average since studies reported the data in various formats. All studies used IV antibiotics and two studies [41, 42] used MBP in addition. Six studies compared oral and IV vs. IV [15, 16, 19, 41–43], five studies compared oral and IV vs. IV with MBP as background [15–19] and three studies compared oral and IV vs. oral and IV [15, 16, 19] in addition to MBP. All participants in the cohorts included in the analysis received elective colorectal operations. Four studies used data of patients that were operated within a period of 2 years [16, 18, 42, 43], one study used looked at a duration of 4 years [15], one study for 1 year [19], one study for 1 year and 4 months [41] and one cohort study did not reveal the timeframe [17]. RCT SSI rates Oral vs. IV Eleven studies compared oral antibiotics with IV antibiotics [8, 21, 22, 26, 28, 31, 34, 36, 37, 39, 40]. The results show a statistically significant benefit of IV antibiotic use in SSI prevention compared with oral antibiotics (OR, 1.82 (95 % CI, 1.28, 2.58); p = 0.008) (Fig. 2). Oral + IV vs. IV Twelve studies compared oral + IV antibiotics with IV antibiotics alone [8, 9, 11, 27–30, 32, 33, 35, 38, 39]. The results
1980
RCT Stellato et al. [39] Subtotal Subtotal SSI rates (%) Cannon et al. [15] Cohort Cohort Englesbe et al. [41] Cohort
Takesue et al. [33] Weaver et al. [34] Kling and Dahlgren [36] Lewis et al. [37] Rohwedder et al. [38]
RCT RCT RCT RCT RCT After MBP
N/A N/A N/A
2012 2010 2015
After MBP After MBP After MBP After MBP After MBP
After MBP After MBP After MBP After MBP
1990
2000 1986 1989 1981 1993
2002 2013 1988 2014
After MBP
Lewis [9] Oshima et al. [29] Raahave et al. [31] Sadahiro et al. [32]
After MBP
1986
After MBP After MBP After MBP After MBP After MBP
After MBP
During MBP During MBP
During MBP
Before MBP
Before MBP
Antibiotic timing
1987
RCT RCT RCT RCT
1982 1988
1983 1982 2005 2007 1988
Lazorthes et al. [8] Playforth et al. [30]
RCT RCT
1982
Dion et al. [26]
Subtotal RCT
1997
2001
Condon et al. [24] Coppa et al. [25] Espin-Basany et al. [27] Kobayashi et al. [11] Lau et al. [28]
Yabata et al. [40]
RCT
RCT RCT RCT RCT RCT
Ishida et al. [35]
RCT
Date
Subtotal Subtotal SSI rates (%) RCT University of Melbourne Colorectal Group [21] RCT University of Melbourne Colorectal Group [22] RCT Beggs et al. [23]
Author
Overall SSI rates and impact of MBP
Type of study
Source
Table 1
3 139 11.75 % N/A N/A N/A
N/A 12 1 5 N/A
N/A N/A 6 N/A
30 22 N/A N/A 18
6
24
6 8.70 % 12
5
N/A 1
5
N/A
N/A N/A N/A
44 1183
N/A 29 27 60 N/A
N/A N/A 50 N/A
540 120 N/A N/A 62
46
121
84
69
30
39
46
N/A
N/A N/A N/A 358 113 104
N/A N/A N/A N/A N/A
N/A N/A N/A N/A
N/A N/A N/A N/A N/A
N/A
N/A
N/A N/A N/A
N/A N/A
N/A N/A
N/A
N/A
SSI events
SSI events
Total
IV
Oral + MBP
SSI rates
1978 915 1270
N/A N/A
N/A N/A N/A N/A N/A
N/A N/A N/A N/A
N/A N/A N/A N/A N/A
N/A
N/A
N/A
N/A
N/A N/A
N/A
N/A
N/A
Total
N/A N/A N/A 60 29 3
N/A N/A N/A N/A N/A
N/A N/A N/A N/A
N/A N/A N/A N/A N/A
N/A
N/A
N/A N/A N/A
N/A N/A
N/A N/A
N/A
N/A
SSI events
Oral + IV
723 638 117
N/A N/A
N/A N/A N/A N/A N/A
N/A N/A N/A N/A
N/A N/A N/A N/A N/A
N/A
N/A
N/A
N/A
N/A N/A
N/A
N/A
N/A
Total
3 96 5.61 % 311 N/A 31
2 N/A N/A N/A 3
5 6 N/A 6
20 8 15 17 6
5
N/A
10 10.99 % N/A
1 9
N/A N/A
N/A
8
SSI events
3400 N/A 1386
51 1710
38 N/A N/A N/A 100
104 97 N/A 99
542 121 200 242 65
51
N/A
N/A
91
30 61
N/A
N/A
72
Total
Oral + IV + MBP
2 234 12.28 % 768 N/A 150
4 3 2 18 96
17 22 2 17
N/A N/A 6 26 7
N/A
10
23 18.11 % 2
4 16
N/A 3
5
17
SSI events
IV + MBP
3839 N/A 2248
51 1905
45 31 27 70 718
103 98 50 95
N/A N/A 100 242 67
N/A
125
83
127
30 58
39
51
71
Total
