Methods

2‐arm RCT

Setting: NR; appears to be single hospital in Saudi Arabia

Participants reportedly followed up for 1 month, no additional details

Participants

254 adults and children (aged 5‐80 years, mean age 21 (Group I) and 24 years (Group II)) undergoing appendectomy for acute appendicitis were randomised; 249 analysed

Inclusion criteria: people undergoing appendectomy through gridiron incision for clinically suspected acute appendicitis

Exclusion criteria: allergy to ampicillin; systemic diseases requiring systemic antibiotic administration

Interventions

Group I: (saline): wound irrigation with 100 mL normal sterile saline at closure (134 participants randomised; 132 analysed; 2 participants withdrawn post‐randomisation)

Group II: (Ampicillin): wound irrigation with 1 g Ampicillin powder dissolved in 100 mL normal sterile saline (120 participants randomised; 117 analysed; 3 participants withdrawn post‐randomisation)

Co‐interventions: IV metronidazole (500 mg for adults; 15 mg/kg for children) and gentamicin (75 mg for adults and 1.5 mg/kg for children) 1 h before surgery. If appendix was found to be gangrenous or perforated antibiotics were continued for 5 ds postoperatively.

Outcomes

Primary outcome: SSI (defined as presence of purulent discharge in wound, regardless of culture results, or as occurrence of serous discharge with a positive culture) within 1 month

Group I (Saline): 7/132 (134 randomised)

Group II (Ampicillin): 1/117 (120 randomised)

Secondary outcome: adverse events including abscess:

Abscess

Group I (saline): 0/132

Group II (Ampicillin): 0/117

Other specific post‐operative complications were reported but total number of participants with adverse events was not clear.

Secondary outcome: hospital stay: reported to be reduced by 2.5 d by avoidance of wound infection. Median reported for participants with (5.5, range 3‐11 d) and without infection (3.0, range 2‐11 d) but not for each treatment group.

Notes

Funding: NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "254 patients fulfilled the criteria and were randomized into two groups using sealed envelopes"

Comment: no information on how the randomisation sequence was generated.

Allocation concealment (selection bias)

Unclear risk

Quote: "254 patients fulfilled the criteria and were randomized into two groups using sealed envelopes that were opened intraoperatively"

Comment: although sealed envelopes were used it is not clear that they were opaque or that the allocation sequence was fully concealed at all times.

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Comment: no direct quote but it is clear that personnel were made aware of allocation once the envelopes were opened. Unclear if participants were aware

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Comment: no direct quote but it is unclear whether the outcome assessment was performed by individuals aware of group allocation.

Incomplete outcome data (attrition bias)
All outcomes

High risk

Comment: there were a small number of post‐randomisation exclusions for protocol violations described. However there were low numbers of events relative to these exclusions, increasing the risk of attrition bias impacting the results.

Selective reporting (reporting bias)

Unclear risk

Comment: rhe data for one of the secondary outcomes (bed‐stay) were not reported on a per‐group base making this outcome difficult to evaluate.

Other bias

Unclear risk

Comment: there was no evidence of other sources of bias but reporting was not clear enough to be certain.

Methods

2‐arm RCT

Setting: single centre; 1 surgical unit in UK

Follow‐up: close monitoring during hospital stay; full inquiry into possible infection signs at 6‐week outpatient clinic follow‐up

Participants

330 participants undergoing elective colorectal surgery (mean ages 61 (Group I) and 63 years (Group II)); 300 analysed

Inclusion criteria: participants undergoing elective colorectal surgery

Exclusion criteria: NR

Interventions

Group I (taurolidine PVP): peritoneal lavage in 2 stages with 250 mL 2% taurolidine in 5% PVP (150 participants)

Group II (saline): peritoneal lavage in 2 stages with 250 mL normal saline (150 participants)

In each group 250 mL lavage solution diluted in a further 250 mL normal saline was placed in the abdomen as a washout and then removed with suction. This was followed by instillation of a second 250 mL undiluted lavage solution, which was left in the abdomen. If abdominal drains were present these were clamped for at least 20 min.

Cointerventions: all participants (except 11 who had severe constricting colonic lesions and were in imminent danger of bowel obstruction) received up to 8 doses of magnesium sulphate (4 g by mouth) for 48 h starting 72 h before surgery followed by 2 sachets of sodium picosulphate (Picolax; Fering Pharmaceutical, Feltham, UK) given in the 24 h immediately before surgery. Participants with severe constricting colonic lesions were prepared according to the wishes of the surgeon; 8 received Klean‐Prep (Norgine, Oxford, UK) in place of Picolax and 3 no preparation. All participants received cefotaxime 1 g and metronidazole 500 mg IV at induction of anaesthesia, and 8 h and 16 h later. 5 participants with penicillin allergy received gentamicin 160 mg on induction, and 120 mg at 8 h and 16 h after induction; doses were individually adjusted according to body mass, renal function and age.

Outcomes

Primary outcome: SSI (defined as spontaneous or incisional discharge from the wound, either of pus or serous fluid, with an infective organism positively identified on culture)

Group I (taurolidine PVP): 17/150 (10 superficial, 7 deep)

Group II (saline): 17/150 (12 superficial, 5 deep)

Secondary outcome: 30‐day mortality

Group I (taurolidine PVP): 4/150

Group II (saline): 4/150

Secondary outcome: adverse events including abscess:

Pelvic abscess

Group I (taurolidine PVP): 2/150

Group II (saline): 1/150

Other specific post‐operative complications were reported but total number of participants with adverse events was not clear.

Secondary outcome: length of hospital stay

Group I (taurolidine PVP): median 11 d for 133 participants without infection; 18 d for 17 with infection (paper reports N = 134, error suspected)

Group II (saline): median 11 d for 133 participants without infection; 18 d for 17 with infection (mean NR, range NR)

Notes

Funding: NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Quote: "This paper reports a randomized controlled trial.....Bottles of lavage fluid were dispensed in identical containers according to a computer‐generated randomized code held by the hospital pharmacy with no stratification for severity of contamination or procedure. "

Comment: appears there was a computer‐generated randomisation sequence

Allocation concealment (selection bias)

Unclear risk

Quote: "Bottles of lavage fluid were dispensed in identical containers according to a computer‐generated randomized code held by the hospital pharmacy with no stratification for severity of contamination or procedure."

Comment: although the allocation sequence was held by the hospital pharmacy it is not clear whether it was adequately concealed.

Blinding of participants and personnel (performance bias)
All outcomes

Low risk

Quote: "Bottles of lavage fluid were dispensed in identical containers according to a computer‐generated randomized code held by the hospital pharmacy... The trial and control solutions were indistinguishable to users."

Comment: personnel appear to have been blinded; although there is no direct information on participants it is likely that they were also blinded.

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

Quote: "All patients were monitored closely after operation until hospital discharge for clinical signs of abdominal sepsis and wound infection by an independent (non‐operating) trained assessor (J.AJ.)."

Comment:

Incomplete outcome data (attrition bias)
All outcomes

High risk

Quote: "Sixteen patients were withdrawn from the trial after consent because the operative procedure performed did not constitute elective colorectal surgery ....... A further eight patients were withdrawn as lavage was not undertaken for logistical reasons, such as breakage of solution bottles. A further six patients were withdrawn at the time of surgery because of overt sepsis or severe faecal spillage, which rendered the intraoperative lavage a therapeutic rather than a prophylactic measure. Thus 300 patient reports were available for analysis."

Comment: almost 10% of randomised participants were not included in the analysis. Although full reasons are given for this the number of withdrawals is almost equivalent to the number of events.

Selective reporting (reporting bias)

Low risk

Comment: the data for one of the secondary outcomes (bed‐stay) were not fully reported (measure of variance lacking) but no other evidence of selective reporting

Other bias

Unclear risk

Comment: there was no evidence of other sources of bias but reporting was not clear enough to be certain.

Methods

Parallel‐group RCT

Setting: single hospital in USA

Follow‐up: 6 weeks (review of records after discharge)

Participants

223 women undergoing cesarean section

Inclusion criteria: women delivered by caesarean section

Exclusion criteria: allergy to penicillin or cephalosporins, taken an antibiotic within 7 d of surgery or required antibiotics around time of surgery for other reasons. Participants with temperature elevated to 38^oC or with foul amniotic fluid prior to or immediately after surgery were considered to have infection and excluded.

High risk and low risk participants were separated according to duration of labor prior to cesarean section, with 6 h arbitrarily chosen as the division point. Each group contained both high risk (more than 6 h labour) and low risk (less than 6 h labour) participants

Interventions

Group I: irrigation with 2 g cefamandole in 1000 mL normal saline (73 participants)

Group II: irrigation with 1000 mL normal saline (75 participants)

Group III: no irrigation (75 participants)

Outcomes

Secondary outcome: length of stay

Group I: low risk: 5.2 (0.3) d (N = 46); high risk: 5.3 (0.2) (N = 27)

Group II: low risk: 5.9 (0.4) d (N = 40); high risk 6.8 (0.6) (N = 35)

Group III: low risk 5.8 (0.3) d (N = 44); high risk: 6.9 (0.4) (N = 31)

Secondary outcome: adverse events

Only a specific event (metritis) was reported

Group I: low risk: 2 (4.3% of 46); high risk: NR

Group II: low risk: 4 (10% of 40); high risk: NR

Group III: low risk: 9 (20.5% of 44); high risk: NR

Infection data were also reported but were not clearly SSI and were not reported for all participants

Group I: low risk: NR; high risk: 3 (11.1% of 27)

Group II: low risk: NR; high risk: 17 (48.6% of 35)

Group III: low risk: NR; high risk: 17 (54.8% of 31)

Notes

Funding: NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Quote: "members of each group were then assigned to a cohort according to a computer‐generated table of random numbers under the direction of the hospital pharmacy"

Comment: appropriate method used to generate randomisation sequence; randomisation stratified by duration of labour: > 6 h vs < 6 h

Allocation concealment (selection bias)

Unclear risk

Quote: "under the direction of the hospital pharmacy"

Comment: unclear whether adequate methods were used to conceal allocation

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Quote: "Physicians who performed the operation and provided postoperative care were unaware of the type of irrigation provided"

Comment: physicians were unaware of the type of irrigation used but are likely to have been aware of whether irrigation was used or not. It is unclear whether participants were aware of treatment allocation.

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Quote: "Patients were followed postoperatively by the resident and attending physicians on service"

Comment: not clear whether outcomes were determined by personnel blinded to treatment allocation.

Incomplete outcome data (attrition bias)
All outcomes

Unclear risk

Quote: "of all 451 patients who had cesarean sections during the study period, 223 were included"

Comment: it appears that the 223 participants described as included were all included in the analyses but it's not completely clear that this is the total number who were randomised.

Selective reporting (reporting bias)

High risk

Comment: not all data relating to outcomes of infection, adverse events and postoperative hospitalisations were reported.

Other bias

Unclear risk

Comment: no obvious source of additional bias but reporting insufficient to be certain

Methods

Parallel‐group RCT

Setting: 16 referral centres in the USA

Follow‐up: 28‐56 d

Participants

449 women (age 32.6 years in Adept group vs 32.3 in lactated Ringer's solution group) undergoing laparoscopic gynaecological surgery. Primary diagnoses included pelvic pain, infertility endometriosis and known adhesions.

Inclusion criteria: aged > 18 years and in good health. Laparoscopic surgery was planned for a gynaecologic procedure that included adhesiolysis followed by a second follow‐up laparoscopy 4–8 weeks later.

Exclusion criteria: preoperative: the use of concomitant systemic corticosteroids, antineoplastic agents, and/or radiation; pregnancy; diagnosis of an active pelvic or abdominal infection, or cancer; and a known allergy to starch‐based polymers. Intraoperative exclusion criteria included women requiring an additional non obstetric/gynaecologic surgical procedure to be performed during the laparoscopic procedure; unplanned surgery necessitating opening the bowel (excluding appendectomy); any laparotomy procedure; and use of another adhesion reduction agent. Adhesion site exclusion criteria included women having < 3 of the available anatomical study sites with adhesions or, if fewer than three were lysed, removal of any anatomical sites being scored for the purposes of the study; and an inability to visualise clearly all available anatomical score sites.

Interventions

Group I: irrigated with a minimum 100 mL Adept (icodextrin 4% solution) solution every 30 min during surgery; any remaining solution at end of surgery was aspirated and then 1 L instilled from a fresh supply of solution (227 ITT, 205 PP participants)

Group II: irrigated with a minimum 100 mL lactated Ringer's solution every 30 min during surgery; any remaining solution at end of surgery was aspirated and then 1 L instilled from a fresh supply of solution (222 ITT 205 PP participants)

Outcomes

Postoperative infections were reported but unclear whether these referred to SSI

Group I (icodextrin): 1% of 227 calculated as 2
Group II (Ringer's solution): 3% of 222 calculated as 7

Secondary outcome: adverse events

Group I (icodextrin): 221/227 of which 44 serious; 55 considered related, reported as serious 8 participants (25 events)

Group II (Ringer's solution): 218/222 of which 36 serious; 38 considered related, reported as serious 11 participants (19 events)

Secondary outcome: mortality

Group I (icodextrin): 0/227

Group II (Ringer's solution): 0/222

Notes

Funding: Innovata Limited, Vectura Group

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Quote: "Treatment was randomized by computer‐generated randomization on a 1:1 basis"

Comment: an appropriate method of generating the randomisation sequence was reported.

Allocation concealment (selection bias)

Low risk

Quote: "Patient numbers were allocated to treatment group before labelling of the blinded study treatment
bags. The study solutions were presented in identical 1 L infusion bags, and each bag had an outer wrap that contained
the study code and patient number on an identification label"

Comment: adequate method for concealment of treatment allocation reported

Blinding of participants and personnel (performance bias)
All outcomes

Low risk

Quote: "Double‐blinding was possible because Adept and LRS are both clear and odourless solutions
with similar viscosities to water."

Comment: blinding appears to have been undertaken for personnel and participants.

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

Quote: "Safety was assessed by serious adverse events (SAEs), adverse events, and changes in laboratory values. Patients completed
diary cards between initial surgery and follow‐up surgery. At postoperative checkup (visits 3 and 4), cards
were assessed to monitor progress. They allowed the patient to record their well‐being and all concomitant medications.
All adverse events whether they were considered related to study solutions or not, were investigated, and the details of
nature, severity, duration, outcome, and relationship to study device were recorded"

Comment: safety outcomes were assessed by participants who were blinded to treatment allocation.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Quote: "Safety was assessed in the intent‐to‐treat (ITT) population, which included all patients who had the study solution instilled.
Efficacy results are presented for the per protocol (PP) population. These patients were those who had completed both first‐ and second‐look laparoscopies without major protocol violations."

Comment: the outcomes relevant to this review were assessed using the ITT population so almost all randomised participants were included in the analyses.

Selective reporting (reporting bias)

Low risk

Comment: outcomes were prespecified; all planned outcomes appeared fully reported.

Other bias

Low risk

Comment: no evidence of other sources of bias and reporting is sufficient to be reasonably confident that this is the case

Methods

Parallel‐group RCT

Setting: single hospital in UK

Follow‐up: NR

Participants

35 participants with gross peritonitis or frank fecal soiling and a positive culture swab at operation

Inclusion criteria: gross peritonitis or frank fecal soiling and a positive culture swab at operation

Exclusion criteria: NR

Interventions

Group I: 2% taurolin in 5% PVP solution (normal saline) up to 200 mL instilled prior to closure of abdomen or afterwards through a tube drain. Additional 200 mL could be instilled daily if required for 7 d (17 participants)

Group II: 5% PVP solution (normal saline) to 200 mL instilled prior to closure of abdomen or afterwards through a tube drain. Additional 200 mL could be instilled daily if required for 7 d (18 participants)

Additional antibiotic use was documented as including gentamycin, lincomycin, cephalosporin [cephalosporine], ampicillin

Outcomes

Secondary outcome: mortality

Group I (taurolin): 3/17

Group II (PVP): 0/18

Notes

Outcome was classed as "good" or "bad" where a normal recovery with normal wound healing and no sepsis was a good result and all other outcomes were bad
Funding NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "Envelopes containing cards, previously randomly arranged were available in theatre for selection of solution A or solution B"

Comment: no information on how the random sequence arrangement was generated

Allocation concealment (selection bias)

Unclear risk

Quote: "Envelopes containing cards, previously randomly arranged were available in theatre for selection of solution A or solution B"

Comment: no information on whether or how the random sequence arrangement was concealed.

Blinding of participants and personnel (performance bias)
All outcomes

Low risk

Quote: "Two solutions were prepared ....in identical bottles labelled A and B"

Comment: it appears that measures were taken to blind personnel and participants.

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

Quote: "Thirty‐five patients entered the trial. It was intended to include a much larger number but at this stage there was a marked difference in results which was statistically significant..... so the code was broken"

Comment: it appears that measures were taken to ensure blinded outcome assessment

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Comment: it appears that all randomised participants were included in the analysis. However many fewer participants than planned were randomised (see other sources of bias).

Selective reporting (reporting bias)

High risk

Outcomes were not prespecified and "a good result" was defined only in the results section.

Other bias

High risk

Quote: "For ethical reasons it was decided the trial should be stopped as soon as a statistically significant difference between the two groups emerged............Thirty‐five patients entered the trial. It was intended to include a much larger number but at this stage there was a marked difference in results which was statistically significant..... so the code was broken."

Trial was stopped very early (a long way short of planned recruitment). Although this was preplanned this approach to early stopping is highly likely to produce an artefactual difference between groups

Methods

Parallel‐group RCT

Setting: single hospital in Norway

Follow‐up: 6 weeks postoperatively

Participants

85 participants with perforated appendicitis and generalised peritonitis

Inclusion criteria: diagnosis of perforated appendicitis and generalised peritonitis verified at laparotomy

Exclusion criteria: age < 6 years; known allergy to ampicillin or tinidazole, localised infiltration or abscess around the appendix.

Interventions

Group I: 24 hs postoperative lavage with 0.9% saline 1 L x 20 for adults, 0.5 L for children (39 participants)

Group II: no postoperative lavage (44 participants)

Cointerventions: intra‐operative peritoneal lavage with 2 L of saline; 2 g ampicillin every 6 hs and 800 mg tinidazole daily until oral fluids commenced then pivampicillin 500 mg 3/d and 1 g tinidazole daily orally; children received pivampicillin 100 mg/kg/d and tinidazole 400 mg daily (rectal) for 5 d

Outcomes

Primary outcome: SSI (wound infection defined as temperature > 38.5 C for > 24 h plus localised, drainage‐confirmed accumulation of fluid in the abdominal incision

Group I (postoperative lavage): 9/39

Group II (no postoperative lavage): 2/44

Secondary outcome: length of stay

Group I (postoperative lavage): median 5 d (range 3‐11) 39 participants

Group II (no postoperative lavage): median 5 d (range 4 ‐12) 44 participants

Notes

Treatment (postoperative lavage) was discontinued early in 10/39 participants

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "As soon as the diagnosis of perforated appendicitis with generalized peritonitis was verified at laparotomy, the patient was randomized..."

Comment: no information on how the randomisation sequence was produced.

Allocation concealment (selection bias)

Unclear risk

Quote: "As soon as the diagnosis of perforated appendicitis with generalized peritonitis was verified at laparotomy, the patient was randomized..."

Comment: no information as to whether allocation was adequately concealed.

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Comment: allocation to postoperative lavage versus no postoperative lavage would be evident to both personnel and participants.

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Comment: unclear whether outcome assessment was performed by blinded individuals

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Comment: 2/85 randomised participants were withdrawn for a documented reason (ampicillin allergy); although both were in the same group the number is low and is unlikely to have been a source of bias.

Selective reporting (reporting bias)

Unclear risk

Comment: outcomes were not all prespecified in the methods although a priori definitions for intra‐abdominal and wound infection were given.

Other bias

Unclear risk

Comment: no evidence of other source of bias but reporting insufficient to be certain

Methods

Parallel‐group RCT

Setting: appears to be single hospital in USA

Follow‐up: 4‐6 weeks postoperatively

Participants

40 women undergoing caesarean section at high risk of infection

Interventions

Group I: irrigation with cefazolin; 2 g in 1000 cc normal saline; 700 cc intrauterine 100 cc in each gutter and 100 cc subcutaneously (20 participants)

Group II: irrigation with 1000 cc normal saline 700 cc; intrauterine 100 cc in each gutter and 100 cc subcutaneously (20 participants)

Cointerventions: none reported

Outcomes

Primary outcome: SSI (defined only as "wound infection")

Group I (cefazolin): 1/20

Group II (saline): 1/20

Notes

Abstract only. Funding NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "Patients at high risk of infection were randomly placed in two groups"

Comment: no information as to how the randomisation sequence was generated

Allocation concealment (selection bias)

Unclear risk

Quote: "Patients at high risk of infection were randomly placed in two groups"

Comment: no information as to whether there was adequate concealment of allocation

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Comment: no specific quote, no information as to whether these groups were blinded to treatment allocation

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Comment: there is no information as to who performed the outcome evaluation or whether they were blinded to treatment allocation

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Quote: "Only 2 of 40 high risk patients who received prophylactic irrigation developed...."

Comment: it appears that all randomised participants were included in the analysis.

Selective reporting (reporting bias)

Unclear risk

Quote: "The objective of this study is to determine the impact of copious antibiotic irrigation versus normal saline (ns) on the incidence of post‐cesarean wound infections"

Comment: the primary outcome was specified and reported but it is not clear from the abstract which other outcomes the study may have planned to assess.

Other bias

Unclear risk

Comment: there were no additional sources of bias noted but the abstract reporting was insufficient to be certain.

Methods

2‐arm RCT

Setting: single centre; 1 surgical unit in the UKm

Follow‐up: included outpatient assessment 6 weeks post‐surgery

Participants

54 women undergoing planned Patey mastectomy for carcinoma of the breast (mean age 56; range 32‐75 years) randomised (52 women and 53 breasts analysed)

Inclusion criteria: Patey mastectomy

Exclusion criteria: participants with allergy to tetracycline

Interventions

Group I (tetracycline): lavage of 1 g tetracycline in 100 mL saline (23 women)

Group II (saline): lavage of 100 mL saline (30 women)

Lavage was given at wound closure and was contained within the axilla and skin flaps as much as possible during closure.

Cointerventions: drainage was standardised to Vygon suction drains to axilla and skin flaps; drains removed at request of surgical staff when drainage for previous 24 hs appeared minimal

Outcomes

Primary outcome: SSI (not defined)

Group I (tetracycline): 0/23

Group II (saline): 1/30

Primary outcome: wound dehiscence
Group I (tetracycline): 0/23

Group II (saline): 1/30 (described as "minor")

Notes

One woman underwent bilateral mastectomy and was randomised for each breast.

Funding: NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "....were randomized to receive at wound closure"

Comment: no information on how the randomisation sequence was generated

Allocation concealment (selection bias)

Unclear risk

Quote: "....were randomized to receive at wound closure"

Comment: no information on whether allocation was adequately concealed

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Comment: no direct quote but no information on whether participants and personnel were blinded

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Quote: "Records were kept on a standard form by the nursing staff, who were unaware of the patient's randomization."

Comment: no information on whether the blinded nursing staff performed outcome assessment at follow‐up.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Quote: "Two women were excluded after randomization because they subsequently did not undergo Patey mastectomy"

Comment: the number of exclusions was very small and both were accounted for by a substantive operative protocol deviation

Selective reporting (reporting bias)

Unclear risk

Comment: there is no evidence of selective reporting but because most outcomes were not specified in the methods section it is not clear whether it may have occurred

Other bias

Unclear risk

There may be a minor unit of analysis issue due to the randomisation of one woman twice for each breast. Otherwise there is no evidence of other bias but the reporting is not full enough to be sure.

Methods

2‐arm RCT

Setting: single centre, Emergency Department in Mexico

Follow‐up: 2 and 4 weeks after operation

Participants

350 participants entered into study; 283 considered evaluable (mean age of 27.99 years (SD 12.81 years), range from 9‐82 years); (67 rejected from final analysis due to finding of another pathology different from appendix)

Inclusion criteria: adults and children of both sexes admitted with a clinical diagnosis of acute abdomen suggestive of acute appendicitis with aid of laboratory and X‐ray, confirmed during operation and by histopathologic study

Exclusion criteria: age < 5 years, allergy to metronidazole or aminoglycosides, antibiotic therapy within 72 h preceding operation, pregnancy, those with other intraperitoneal bacterial infection not originating from the appendix, and those with any immune deficiency (diabetes mellitus, chronic renal insufficiency, malnourishment, chemotherapy, radiotherapy, corticosteroid therapy, asplenism)

Interventions

Group I: no irrigation (156 participants)

Group II: syringe pressure irrigation with saline: after closure of the fascial planes, subcutaneous fat tissue irrigated with 300 mL of normal saline solution, delivered with a 20‐mL syringe with a 19‐gauge IV catheter, applying to the embolus the force of one hand, at a distance of 2 cm from the wound tissues, aspirating the fluid collected in the wound with a bulb syringe (127 participants)

Cointerventions: each participant was administered metronidazole (30 mg/kg/day) 3/d, plus amikacin (15 mg/kg/d) once daily IV 30–45 min before skin incision. In cases of uncomplicated appendicitis they were stopped within the first 24 h, whereas in cases of complicated appendicitis they were maintained for a minimum of 7 d

Outcomes

Primary outcome: SSI (Definition: "A wound was considered to be infected...when there was a collection of pus or a positive bacteriologic culture from a wound discharge")

Group I: (no irrigation): 39/156

Group II: (syringe pressure irrigation): 11/127

Secondary outcome: adverse events. The authors stated that "Antibiotics used for prophylaxis were well tolerated without any case of allergy or intolerance." The proportion of participants with any adverse event was not reported.

Notes

Funding: NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Quote: "All patients included were randomly assigned by a computerized assignment system into 2 groups of the trial."

Comment: randomisation sequence generated by computer

Allocation concealment (selection bias)

Low risk

Quote: "The randomization chart was kept by our statistician, who was blind to the follow‐up, until June 1995."

Comment: appears that allocation sequence was concealed from trial personnel

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Quote: "Statistical analysis of the results...was conducted by our statistician who was blind to the surgical procedures and follow‐up."

Comment: control arm did not include comparator intervention and so unable to conceal allocation to staff present at operation. Study design indicated as "double blind," so assumption that participants were not told of their treatment allocation

High risk of physicians not being blinded; unclear or low risk for participants

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

Quote: "...sought by daily examination of all patients by one of the members of the research team who was blind to the random allocation and the surgical procedures, until discharge."

Quote: "...reevaluated at the outpatient consultation 2 and 4 weeks after operation by the responsible author who was blind to the random assignment and the surgical procedures."

Comment: appears that outcome assessment for SSI was conducted by blinded assessor

Incomplete outcome data (attrition bias)
All outcomes

High risk

Quote: "A total of 350 patients were entered into the study, and 283 (80.9%) were considered evaluable. The reason for rejection of the 67 (19.1%) patients from the final analysis was the finding of another pathology different from the appendix."

Comment: the number of exclusions was high and the study did not achieve its aim of including 133 participants in each arm. Therefore confirmation of appendicitis during surgery appears to be an inclusion criterion and so exclusions based on pathology do not violate inclusion criteria. There was a high rate of exclusion relative to event rate for the primary outcome.

Selective reporting (reporting bias)

Unclear risk

Comment: main end point (defined surgical wound infection) was reported overall and for complicated and uncomplicated types of appendicitis. It is unclear whether there were any other end points specified in the study protocol.

Other bias

Unclear risk

Comment: there was no evidence of other sources of bias but reporting was insufficient to be certain.

Methods

2‐arm RCT

Setting: single centre at hospital in Taiwan

Follow‐up: at 2 weeks, 1 month and 3 months after operation, and then every 3 months until end of study (approximately 19 months)

Participants

244 participants (age range 20‐89 years; Group I: average 67.1 years, (range 20‐82 years); Group II: average 65.4 years (range 22‐89 years)).

Inclusion criteria: primary instrumented lumbosacral posterolateral fusion for degenerative spinal disorder with lumbar or lumbosacral segmental instability defined by chronic back, buttock and/or leg pain and degenerative spondylolisthesis, degenerative scoliosis or isthmic spondylolisthesis

Exclusion criteria: prior spinal surgery, spinal trauma, malignant tumour, infectious spondylitis, rheumatoid arthritis, ankylosing spondylitis, metabolic bone disease, skeletal immaturity or immunosuppressive treatment

Interventions

Group I (povidone‐iodine): wounds irrigated with 0.35% povidone‐iodine solution to soak for 3 min, followed by irrigation with 2000 cc normal saline to remove povidone‐iodine solution (120 participants)

Group II (saline): wounds irrigated with only 2000 cc normal saline (124 participants)

Cointerventions: wound closure by layer after suction drainage applied; drain removed 48 h or 72 h post‐operatively. Routine analgesic pain control applied for 3 d. Pre‐operative IV bolus injection of cefazolin (1000 mg) and gentamicin (60 mg); additional cefazolin (1000 mg/6 h) and gentamicin (60 mg/12 h) also given for 48 hs after surgery, and then oral cefazolin (500 mg/6 h) for 3 d

Outcomes

Primary outcome: SSI (Definition: "Infections were classified as superficial (above lumbosacral fascia) or deep (below lumbosacral fascia), and as early onset (within 2 weeks postoperatively) or late onset (otherwise). All deep infections were confirmed by laboratory parameters including erythrocyte sedimentation rate (ESR) and level of C‐reactive protein (CRP) and a positive culture of biopsy."

Group I (povidone‐iodine): 0/120

Group II (saline): 6/124 (2 early onset; 4 late onset)

Primary outcome: wound dehiscence within 30 d (time NR but it says all others healed with sutures removed on day 14 so can presume < 30 d. No infection found in wounds)

Group I: 1/120

Group II: 2/124

Secondary outcome: proportion of participants with postoperative SSI using systemic antibiotics within 30 d of surgery

Group II: (saline): "After radical debridement and parenteral antibiotics (according to sensitivities) for 6 weeks and oral antibiotics for 2 months, a satisfactory outcome has been reached except in two cases."

Secondary outcome: occurrence of infections showing antibiotic resistance

Group II: (saline): MRSA cultured from 5/6 cases

Secondary outcome: surgical re‐intervention rates

Group I: (povidone‐iodine): 3 participants underwent exploration of the non‐union site and re‐arthrodesis with autogenous bone graft

Group II: (saline): 4 participants underwent exploration of the non‐union site and re‐arthrodesis with autogenous bone graft

Group II: (saline): "After radical debridement and parenteral antibiotics (according to sensitivities) for 6 weeks and oral antibiotics for 2 months, a satisfactory outcome has been reached except in two cases." These 2 cases had implants removed 4 months post‐operatively as infection could not be eradicated

Notes

Interventions: no information given on duration of irrigation of wounds with normal saline for either group.

Funding: Quote: "No funds were received in support of this work. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this manuscript."

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Quote: "Patients...were randomly assigned to either treatment group. An independent person unaware of the subject characteristics and the study design delivered pre‐coded sealed enveloped randomly (containing serial numbers from 1 to 300) to the assignment of the subjects into the two groups."

Comment: clearly states how sequence was generated

Allocation concealment (selection bias)

Unclear risk

Quote: "An independent person unaware of the subject characteristics an the study design delivered pre‐coded sealed enveloped randomly (containing serial numbers from 1 to 300) to the assignment of the subjects into the two groups. The sealed envelope was not opened until the middle of the surgery before wound irrigation."

Comment: although sealed envelopes were used it is not clear that they were opaque or that the allocation sequence was fully concealed at all times.

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Comment: no direct quote, but no information on how personnel might have been blinded to the treatment performed. Unclear whether participants were blinded, but report states it was "single blind"

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Quote: "All clinical and radiographic assessments were made by independent observers other than the treating surgeons."

Comment: unclear whether observers were aware of treatment group

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Comment: no direct quote, but no evidence of attrition

Selective reporting (reporting bias)

Unclear risk

Comment: no evidence of selective reporting but not enough information to be certain

Other bias

Unclear risk

Comment: there was no evidence of other sources of bias but reporting was insufficient to be certain

Methods

2‐arm RCT

Setting: single centre at hospital in Taiwan

Follow‐up: at 2 weeks, 4 weeks and 2 months after operation, and then every 3 months until end of study (mean follow‐up 15.5 months for both groups)

Participants

417 consecutive eligible participants enrolled. 3 who died during the follow‐up period were excluded (1 case in Group I and 2 cases in
Group II); 414 were included (average age 64 years (Group I) and 61 years (Group II))

Inclusion criteria: pre‐operative diagnosis of degenerative scoliosis or stenosis; degenerative disc disease; disc prolapse; traumatic spinal fracture; spinal metastasis lesion. Undergoing procedure such as decompression for degenerative stenosis; decompression, fusion and fixation for degenerative scoliosis or stenosis; fixation of traumatic spinal fracture; discectomy for disc prolapse; excision with fixation for spinal metastatic lesions.

Exclusion criteria: those with overt or suspected pyogenic vertebral osteomyelitis, discitis, or any form of pre‐operative spinal infection were excluded. Those with fever or other suspected sources of infection also excluded

Interventions

Group I (povidone‐iodine): surgical wound soaked with dilute povidone‐iodine solution for 3 min after operation. Commercially available Betadine solution used had a concentration of 10% povidone‐iodine (100 mg povidone‐iodine per 1 mL solution). Approximately 5 mL povidone‐iodine was diluted with normal saline to achieve a 0.35% povidone‐iodine (3.5% Betadine) solution for use during the operation. The wound was irrigated with copious amounts of normal saline (2000 mL) after Betadine solution irrigation (208 participants)

Group II (saline): irrigation with copious normal saline (2000 mL) performed alone (206 participants)

Cointerventions: each participant received 1 dose of parenteral cefazolin (1000 mg) and gentamicin (60 mg) 1 h before surgery. Cefazolin (1000 mg) every 6 hs and gentamicin (60 mg) every 12 hs were then given for 48 hs after surgery. Additional doses of antibiotics were given to maintain antibiotics levels during prolonged surgery. Following IV antibiotics, cefazolin (500 mg every 6 hs) was continued orally for 3 d. Drains were retained until < 100 mL of output was observed.

Outcomes

Primary outcome: SSI (Definition: "Infection was suspected when unusual pain, tenderness, erythema, induration, fever, or wound drainage was noted. Such findings were investigated with measurement of erythrocyte sedimentation rate, C‐reactive protein, and bacteriological cultures from the operative site or blood. Cultures were obtained from blood and wound discharge by aseptic methods.")

Group I (povidone‐iodine): 0/208

Group II (saline): 7/206 (one superficial and 6 deep)

Secondary outcome: occurrence of infections which show antibiotic resistance

Group II (saline): MRSA cultured from 5/7 cases

Secondary outcome: surgical re‐intervention rates

Group II (saline): 7/206 (all those with highly suspected wound infection underwent surgical debridement)

Notes

Funding: Quote: "No funds were received in support of this work. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this manuscript."

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Quote: "Patients were randomly assigned to two groups, using pre‐coded sealed envelopes containing serial numbers from 1 to 500. Patients with odd serial numbers were group 1 (study group) and those with even serial number were group 2 (controls)."

Comment: clearly states how sequence was generated

Allocation concealment (selection bias)

Unclear risk

Quote: "Patients were randomly assigned to two groups, using pre‐coded sealed envelopes containing serial numbers from 1 to 500. Envelopes were not opened until the end of surgery, before wound irrigation. Patients with odd serial numbers were group 1 (study group) and those with even serial number were group 2 (controls)

Comment: although sealed envelopes were used it is not clear that they were opaque or that the allocation sequence was fully concealed at all times.

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Comment: no direct quote, but no information on how personnel might have been blinded to the treatment performed. Unclear whether participants were blinded, but report states it was "single blind"

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Comment: no direct quote; no information given regarding who collected outcome assessment data

Incomplete outcome data (attrition bias)
All outcomes

High risk

Quote: "Three patients who died during the follow‐up period were excluded (one case in group 1 and two cases in
group 2)."

Comment: number of exclusions is low but similar to number of events

Selective reporting (reporting bias)

Unclear risk

Comment: no evidence of selective reporting but not enough information to be certain

Other bias

Unclear risk

Comment: there was no evidence of other sources of bias but reporting was not clear enough to be certain

Methods

Parallel RCT

Setting: single centre in Republic of Korea

Follow‐up: 2 weeks

Participants

34 patients undergoing gastrectomy

Inclusion criteria: naive stomach cancer patients

Exclusion criteria: history of diabetes, pneumonia, urinary tract infection, chemotherapy

Interventions

Group I: saline exchange after gastrectomy (17 participants)

Group II: no saline exchange during surgery (17 participants)

Co‐interventions: preoperative cefametazol 1 g

Outcomes

Primary outcome: SSI

Superficial, deep SSI defined by Horan 1992

Group I (saline exchange): 1/17

Group II (no saline exchange): 3/17

Notes

Funding: NR

Reported in Korean. Data extraction and risk of bias assessment performed by translator

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Comment: random number table was used

Allocation concealment (selection bias)

Unclear risk

Comment: method of allocation concealment was not described

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Comment: personnel were blinded because they were under anaesthesia but personnel would have been aware of the allocation due to the nature of the comparison

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

Comment: outcome assessors were blinded

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Comment: all participants were included

Selective reporting (reporting bias)

Low risk

Comment: pre‐specified outcomes were reported

Other bias

Low risk

Comment: not detected

Methods

Parallel‐group, 5‐arm RCT

Setting: single hospital in USA

Follow‐up: unclear

Participants

360 women undergoing caesarean section. Both caesareans in labour and without labour were included. Mean ages between 24.59 and 27.52 years. Gestational ages between 37.85 and 39.31 weeks

Inclusion criteria: women undergoing caesarean section

Exclusion criteria: history of penicillin or cephalosporin allergy, taking antibiotics, known infectious process (e.g. chorioamnionitis or urinary tract infection)

Interventions

Group 1: saline lavage (800 mL)

Group 2: 2 g cephapirin sodium lavage

Group 3: 2 g cefamandole nafate lavage

Group 4: 2 g moxalactam disodium lavage

Group 5: 2 g ampicillin sodium lavage

Inferred that each antibiotic lavage used 800 mL

Cointerventions: none reported

Outcomes

Primary outcome: SSI (wound breakdown with positive culture or presence of cellulitis)

Group 1 (saline): 3/77

Group 2 (cephapirin): 3/70

Group 3 (cefamandole): 2/64

Group 4 (moxalactam): 2/79

Group 5 (ampicillin): 0/70

Secondary outcome: adverse events including abscess

There were 0 abscess events; other adverse events reported were infection‐related morbidity as follows

Group 1 (saline): 22/77

Group 2 (cephapirin): 17/70

Group 3 (cefamandole): 8/64

Group 4 (moxalactam): 19/79

Group 5 (ampicillin): 10/70

Notes

Funding NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Quote: "A computer‐generated table of pseudo‐random numbers.... was used by the pharmacy to assign each patient to one of five groups"

Comment: an acceptable method of sequence generation was reported.

Allocation concealment (selection bias)

Unclear risk

Quote: "A computer‐generated table of pseudo‐random numbers.... was used by the pharmacy to assign each patient to one of five groups"

Comment: there was no information on how allocation concealment was undertaken.

Blinding of participants and personnel (performance bias)
All outcomes

Low risk

Quote: "A vitamin .... was added to each solution for disguise"

"The patients and physicians were unaware of the group assignment until after completion of the study and chart review by the authors"

Comment: blinding of both participants and physicians was undertaken.

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Comment: no specific quote but it was unclear who performed the outcome assessments and hence whether they were blinded to group allocation.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Comment: all randomised participants were included in the analyses.

Selective reporting (reporting bias)

Unclear risk

Comment: the outcomes were not defined in the methods section so it is unclear whether all planned outcomes were fully reported.

Other bias

Unclear risk

Comment: no apparent sources of additional bias but reporting insufficient to be certain

Methods

2‐arm RCT with 2 phases

Setting: NR, but appears to be a general surgery department at a hospital in the Netherlands

Follow‐up: at 4, 8 and 14 d, and 4 weeks after surgery

Participants

592 participants, of which 34 excluded (18 in the control group and 16 in the povidone‐iodine group) because they died before the end of the control period or had to be operated upon again through the same wound during this period. 2 wounds were present in 21 participants (9 in the control group and 12 in the povidone‐iodine group). A total of 582 wounds were evaluated in 558 participants.

Inclusion criteria: all elective and acute patients who underwent intra‐abdominal operations or operations for inguinal hernia

Exclusion criteria: children < five years of age and those undergoing vascular reconstruction.

Interventions

Group I (control): quote "No special measures were taken"

Group II (povidone‐iodine): carried out in 2 phases. Subcutaneous tissues irrigated with a povidone‐iodine solution at the end of the operation. Lavage with an ample amount of 1% aqueous povidone‐iodine solution (Phase 1) or 10% aqueous povidone‐iodine solution (Phase 2). Lavage was performed after closure of the fascia with interrupted polyglactin 910 sutures. After lavage for 1 min, excess fluid was aspirated and skin closed with interrupted with nylon sutures. If present, drains were brought out through a second wound.

Outcomes

Primary outcome: SSI (Definition: "Diagnosis of wound infection made if a purulent discharge form the wound was seen within a period of four weeks after the operation or if culturing of fluid from the wound was positive.")

Group I (control): Phase 1: 21/142 wounds Phase 2: 15/137 wounds (270 participants? ‐ participant numbers unclear)

Group II (povidone‐iodine): Phase 1: 17/154 wounds Phase 2: 22/149 wounds (291 participants? ‐ participant numbers unclear)

Notes

Participants: age NR

Outcomes: these appear to refer to number of wounds, not participants

Funding: NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "The patients were divided at random into two groups."

Comment: it is unclear how randomisation was performed.

Allocation concealment (selection bias)

Unclear risk

Quote: "The patients were divided at random into two groups."

Comment: no information on whether the randomisation sequence was concealed

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Quote: "In the first group, no special measures were taken."

Comment: the control arm did not involve an intervention as a comparator, and so unable to conceal allocation to staff present at the operation. Unclear whether all staff were aware of the different phases of the study (and concentrations of solution used as the intervention).

Unclear whether participants were aware of treatment allocation

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Quote: "Postoperatively, all wounds were assessed by the same investigator..."

Comment: unclear as to whether the investigator was blinded to the treatment allocation

Incomplete outcome data (attrition bias)
All outcomes

High risk

Quote: "34, 18 in the control group and 16 in the povidone‐iodine group, were excluded because they either died before the end of the control period or had to be operated upon again through the same wound during this period."

Comment: number of exclusions is high. No reasons given for cause of death or re‐operations

Selective reporting (reporting bias)

Unclear risk

No evidence of reporting bias, but report is not complete enough to be sure

Other bias

Unclear risk

Comment: there was no evidence of other sources of bias but reporting was not clear enough to be certain

Methods

4‐arm RCT

Setting: 2 hospitals in USA

Follow‐up: 6 weeks

Participants

'High risk' patients (for developing post operative febrile morbidity) undergoing cesarean section for a variety of reasons

158 women included in study

Inclusion criteria: women in active labour or with ruptured membranes, at least one digital vaginal examination (i.e. high risk from developing postoperative febrile morbidity)

Exclusion criteria: allergy to cephalosporins or penicillin, presence of fever ≥ 37.8 C during labour with suspicion of chorioamnionitis, maternal use of antibiotics in 2‐week period before delivery

Interventions

Group I: 8 doses of IV cefoxitin 2 g (1st dose after umbilical cord clamp, then every 6 hs) (39 participants)

Group II: irrigation of uterus and peritoneum with 2 g cefoxitin (in 1000 mL of normal saline). After delivery of the placenta, the fundus of the uterus was irrigated with 300 mL, the uterine incision with 150 mL, after closure of the first layer 150 mL, bladder flap 150 mL, remainder used to irrigate peritoneal cavity and excess suctioned away before closure of the abdomen (42 participants)

Group III: combination of IV antibiotic (8 doses of 2 g cefoxitin) and irrigation with cefoxitin (in 1000 mL of normal saline) i.e. treatments of groups I and II combined (38 participants)

Group IV: control group who received no prophylactic antibiotics (39 participants)

Outcomes

Primary outcome: SSI (wound infection)

Group I (IV antibiotics only): 0/39

Group II (irrigation with antibiotics only): 0/42

Group III (IV antibiotics plus irrigation with antibiotics): 0/38

Group IV (no IV and no irrigation):1/39

Secondary outcome: hospital stay (mean (SD) d)

Group I: 4.8 (1.1)

Group II: 4.9 (1.0)

Group III 4.9 (1.2)

Group IV: 5.4 (1.4)

Secondary outcome: adverse events

Infectious (endomyometritis, urinary tract infection, wound infection (1 case, see above), pulmonary infection, septicaemia)

Group I: 2/39

Group II: 3/42

Group III: 2/38

Group IV: 14/39

Non‐infectious (seroma, transfusion reaction, atelectasis)

Group I: 1/39

Group II: 1/42

Group III: 1/38

Group IV: 0/39

Notes

Funding: NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Quote: "Randomization into one of four treatment groups was performed by using a table of random numbers"

Comment: an appropriate method appears to have been used

Allocation concealment (selection bias)

Unclear risk

Comment: there is no information about allocation concealment

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Comment: there is no information about blinding. It is possible that participants were blinded, but personnel would be aware of treatment as the protocols are quite different.

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Comment: there is no information about blinding or who performed outcome assessment

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Comment: all participants are accounted for in the results

Selective reporting (reporting bias)

Unclear risk

Comment: the outcomes do not appear to have been pre‐specified, although febrile morbidity was extensively defined

Other bias

Unclear risk

Comment: there is no evidence of additional sources of bias but the reporting is insufficient to be confident that there were none.

Methods

2‐arm RCT

Setting: single hospital in UK

Follow‐up: 1 month

Participants

129 patients undergoing elective and emergency colorectal surgery

Age: unknown;

Type of operations: unknown

Inclusion criteria: NR

Exclusion criteria: NR

Interventions

Group I: 1000 mL saline lavage at the end of the operation (65 participants)

Group II: 1000 mL saline lavage with 1 g cefotetan at the end of the operation (64 participants)

Co‐interventions: Groups I and II both received 500 mg metronidazole and 120 mg gentamicin IV at anaesthesia induction

Outcomes

Primary outcome:

SSI (defined as discharge of pus from the wound "wound sepsis")

Group I (saline): 18/65

Group II (cefotetan): 15/64

Notes

Funding: NR

Limited information from paper

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "Patients were randomly allocated to receive either 1 liter of saline lavage or 1 liter of saline containing 1g of cefotetan..."

Comment: the method of randomisation is not described

Allocation concealment (selection bias)

Unclear risk

Comment: there is no mention of allocation concealment

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Comment: there is no mention of blinding of participants or personnel

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Quote: "Post‐operatively, patients were assessed regularly by a single observer for the development of wound sepsis..."

Comment: there is no mention of blinding of the observer

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Comment: participants are all accounted for in the outcome data of interest

Selective reporting (reporting bias)

Unclear risk

Comment: the outcome of interest (SSI) is reported but it is not clear that related results are fully reported

Other bias

Unclear risk

Comment: there is not enough methodological information to judge whether there were any additional sources of bias

Methods

2‐arm RCT

Setting: single hospital in Turkey

Follow‐up: participants were examined at 2 and 6 weeks after surgery. Wounds examined twice daily during hospitalisation. After discharge, women were instructed to contact investigators immediately if any of the listed symptoms appeared. Women who contacted the investigators were examined within 12 h

Participants

520 women with indications for elective or emergency caesarean section (incidence of emergency surgery (45.5 vs 51.5%; P  = 0.53))

Inclusion criteria: past 37 weeks' gestation and required a caesarean section (elective or emergency).

Exclusion criteria: anaemia (haemoglobin: < 7 g/dL), chorioamnionitis and fever on admission

Interventions

Group I: underwent wound irrigation before wound closure with 100 mL of sterile saline with a 30–60 mL syringe (260 participants)

Group II: no wound irrigation before wound closure (260 participants)

Outcomes

Primary outcomes: SSI (wound drained purulent material or serosanguineous fluid in association with induration, warmth and tenderness)

Group I (saline): 17/260

Group II (no irrigation): 19/260

Secondary outcomes: mean length of hospital stay

Group I (saline): 2.05 (0.21) d

Group II (no irrigation): 2.04 ( 0.20)

Notes

Funding: NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "Consenting patients were preoperatively randomised using numerically ordered cards in sealed envelopes"

Comment: method of sequence generation is not reported

Allocation concealment (selection bias)

Unclear risk

Quotes: "Consenting patients were preoperatively randomised using numerically ordered cards in sealed envelopes" "The investigator was not blinded to the procedure allocation" "The allocated envelope was opened by the clinician just before surgery"

Comment: the use of sealed envelopes suggests an attempt to conceal some aspect of allocation but the authors state that the investigator was not blinded to allocation; envelopes are not stated to be opaque

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Quote: "The allocated envelope was opened by the clinician just before surgery. The procedure allocation was recorded in the women's charts"

Comment: personnel and participants were both aware of treatment

Blinding of outcome assessment (detection bias)
All outcomes

High risk

Quote: "The procedure allocation was recorded in the women's charts" "The investigator was not blinded to the procedure allocation"

Comment: it is not explicitly stated but the report suggests the outcome assessors were not blinded

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Comment: data are reported for all participants

Selective reporting (reporting bias)

Unclear risk

Comment: apart from SSI, it is unclear which outcomes were prespecified

Other bias

Low risk

Comment: there is no evidence of other bias

Methods

Parallel‐group RCT

Setting: single hospital in UK

Follow‐up: 4 weeks postoperatively

Participants

192 participants undergoing appendectomy via grid iron incision

Inclusion criteria: appendectomy via a right iliac fossa incision

Exclusion criteria: female participants of child bearing age not adequately protected by contraceptive practice

Interventions

Group I: 50 mL 2% taurolin in 5% PVP in "saline sufficient to produce solutions of equal tonicity"; wound irrigated for 2 min; then 10 mL instilled through a quill after closure of the skin

Group II 50 mL of 5% PVP in "saline sufficient to produce solutions of equal tonicity" wound irrigated for 2 min; then 10 mL instilled through a quill after closure of the skin

Cointerventions: antibiotics and drains as required

Outcomes

Primary outcome: SSI (wound sepsis) defined as a wound discharging pus

Group I (taurolin): 18/99

Group II (placebo): 29/93

Secondary outcome: length of stay

Group I (taurolin): 6.4 d (mean, no SD)

Group II (placebo): 6.6 d (mean, no SD)

Notes

Funding: NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "the taurolin and placebo being randomly allocated to sequential numbers 1 to 200"

Comment: no information on how the randomisation sequence was generated

Allocation concealment (selection bias)

Unclear risk

Quote: "the taurolin and placebo being randomly allocated to sequential numbers 1 to 200"

Comment: no information on whether allocation was adequately concealed

Blinding of participants and personnel (performance bias)
All outcomes

Low risk

Quote: "neither solution was distinguishable to users"

Comment: it appeared that personnel (and participants) were blinded

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Quote:"During the hospital stay the wound was observed be a member of the medical staff participating in trial"

Comment: it was unclear whether the individual who assessed the outcomes was blinded to treatment group

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Comment: the 8 participants who were not included in the analyses were clearly documented. Although 7 of these were placebo group‐allocated participants it appears unlikely that these exclusions would have affected the results.

Selective reporting (reporting bias)

Unclear risk

Comment: the primary outcome was specified but it was unclear which other outcomes were planned to be recorded

Other bias

Unclear risk

Comment: there is no evidence of other sources of bias but reporting was insufficient to be certain.

Methods

2‐arm RCT

Setting: multicentre trial: 4 UK hospitals in the "South of England"

Study period: 18‐months

Follow‐up: 30 d post surgery or discharge from unit (1 assessor in each hospital reviewed participants' wounds twice a week until discharge)

Participants

356 participants with a displaced intracapsular fractured neck of femur, due to be treated with a hemiarthroplasty, were randomised into 2 groups

Inclusion criteria: displaced intracapsular fractured neck of femur, due to be treated with a hemiarthroplasty

Exclusion criteria: NR

Interventions

Group I: the ‘pulse lavage’ group had a 2‐L normal saline wash delivered via pulsatile lavage in stages throughout the procedure (164 participants)

Group II: the control group had a 2‐L normal saline wash delivered by a jug or a syringe according to the surgeon's preference with 1 L being given before prosthesis insertion and 1 L after insertion (192 participants)

Co‐interventions: NR

Outcomes

Primary outcome: SSI

Wound infections were diagnosed using criteria from the Nosocomial Infection National Surveillance Survey and graded as superficial or deep.

Group I (pulse lavage): 9/164 (3/164 'deep')

Group II (control): 30/192 (10/192 'deep')

Secondary outcome: occurrence of infections with antibiotic resistance

No group data but quote: "Half of the deep space infections were due to methicillin‐resistant Staphylococcus aureus"

Secondary outcome: mortality

NR by group. There were 25 deaths within the study period (7%); 18 of these were associated with American Society of Anaesthesiologists (ASA) scores of 3 or below

Notes

Funding: no records

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "...all patients... were randomized into two groups"

Comment: no details about method

Allocation concealment (selection bias)

Unclear risk

Comment: no mention of allocation concealment

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Comment: no mention of blinding, but nature of intervention and control means personnel would not be blind to treatment

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Quote: "One assessor in each hospital reviewed patients' wounds twice a week until discharge"

Comment: no mention of blinding and it is unclear if the assessor would have been aware of treatment

Incomplete outcome data (attrition bias)
All outcomes

High risk

Quote: "the difference in size between the two groups was due to 'start up' problems within the hospitals where pulse lavage had not been previously associated with hemiarthroplasty operations" "In cases where hemiarthroplasties were due to have pulse lavage but this was forgotten, the cases were struck from the study"

Comment: the authors describe issues with implementing the pulse lavage intervention and although they describe participants being excluded at some point for this reason (presumably post‐randomisation) there are no details about them. It is probable that this introduced bias to the study.

Selective reporting (reporting bias)

High risk

Comment: data are fully reported for some outcomes but others are not reported by group

Other bias

Unclear risk

Comment: there is insufficient information to judge

Methods

2‐arm, parallel‐group RCT

Setting: single centre in USA

Follow‐up: NR

Participants

196 women undergoing caesarean delivery. 94 were elective repeat procedures, age 27.5 vs 28.2 years

Inclusion criteria: women presenting with term (> 37 weeks) singleton pregnancies undergoing routine caesarean delivery for arrest of dilation, arrest of descent, foetal malpresentation or as an elective repeat procedure

Exclusion criteria: women diagnosed with chorioamnionitis, type I diabetes, placenta previa, placenta accreta, maternal coagulopathy, multiple gestation, HIV–positive status, prior severe gastrointestinal disease, or non‐reassuring fetal monitoring requiring immediate delivery

Interventions

Group I: irrigation with 500‐1000 mL warm saline after closure of the uterine incision but before closure of the abdominal wall (97 participants)

Group II: no irrigation (99 participants)

Outcomes

Primary outcome: SSI (undue tenderness, erythema, discharge, or separation of the incision accompanying maternal fever)

Group I (saline irrigation): 1/97

Group II (no irrigation): 2/99

Secondary outcome: length of stay (d)

Group I (saline irrigation): 2.9 (1.0)

Group II (no irrigation): 2.8 (0.9)

Secondary outcome: adverse events (postpartum complications including SSI)

Group I (saline irrigation): 14/97

Group II (no irrigation): 13/99

Notes

Funding: NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Quote: "Assignment was performed by pulling sequentially numbered opaque envelopes containing computer‐randomized individual allocations."

Comment: an appropriate method of random sequence generation was reported

Allocation concealment (selection bias)

Low risk

Quote: "Assignment was performed by pulling sequentially numbered opaque envelopes containing computer‐randomized individual allocations."

Comment: an appropriate method of allocation concealment was reported

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Quote: "This randomization was carried out by research staff before initiation of the study, and the patients were blinded to treatment once assigned."

Comment: although participants were blinded to treatment allocation it is unclear whether personnel were also blinded, the nature of the intervention groups suggests that they were not.

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

Quote: "Postoperative care providers were blinded to group assignment to minimize potential bias. .... The randomizing physician collected the initial data. Data entry was performed by data technicians who did not participate in the design or execution of the study; these technicians also reviewed the charts of each randomized patient to assess the accuracy of information provided by the treating physician. The senior investigator performed periodic reviews of data entry to ensure completeness and accuracy of information in the computer database. The data analysis was performed by an investigator blinded to group assignment."

Comment: blinded outcome assessment was conducted

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Comment: all randomised participants were included in the analyses

Selective reporting (reporting bias)

Low risk

Quote: "The primary outcome measure was the incidence of maternal morbidity, defined as the presence of at least one of the following:..."

Comment: primary and secondary outcomes were clearly specified and fully reported.

Other bias

Low risk

Comment: there were no other sources of bias evident and reporting was sufficient to be reasonably confident that this was the case.

Methods

2‐arm RCT

Setting: single hospital in Slovakia

Follow‐up: 7.8 months (mean) 2‐14 months (range): follow‐up at 2 weeks, 6 weeks and then 3‐monthly

Participants

162 children (undergoing 182 surgical procedures on soft and bone tissues in the proximal femur, hip and pelvic regions. mean age was 7.9 vs 7.5 years Types of procedures: adductor tenotomy, femoral or pelvic osteotomy, extraction of metal materials, open reductions, epiphysiodesis, resection or biopsy. Children had the following long‐term conditions: developmental dysplasia of the hip, cerebral palsy, tumours, Perthes disease

Inclusion criteria: children undergoing surgery of the femur, hip or pelvis

Exclusion criteria: NR

Interventions

Group I: lavage with 3.5% Betadine solution (0.35% povidone iodine) diluted in 30 mL sterile saline

Group II: lavage with 30 mL sterile saline

Cointerventions: antibiotic prophylaxis begun preoperatively in participants with femoral or pelvic osteotomy or massive surgery of soft tissues and continued for 48‐72 hs postoperatively (dose determined by weight). Drains left in until second postoperative day where necessary

Outcomes

Primary outcome: SSI (positive bacteriological examination)

Group I (Betadine lavage): 0/89

Group II (saline lavage): 2/73

Notes

Funding: NR

Slovak; data entry and risk of bias based on information provided by translator

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Comment: no information about the sequence generation process but stated that the participants were allocated to 2 groups randomly

Allocation concealment (selection bias)

Unclear risk

Comment: no information on whether the allocation was adequately concealed

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Comment: no information on whether personnel or participants were blinded to the interventions

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Comment: no information on who assessed the presence of SSI or whether they were blinded to treatment allocation

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Quote "In the first group (89 patients) [we] found no peri‐ or post‐operative infection. In the second group of patients (73) [we] brought to light two surface infection[s] (2.7%)"

Comment: it appears that all participants were included in the analysis

Selective reporting (reporting bias)

Unclear risk

Comment: outcomes were not specified in the methods section so difficult to be certain whether all planned outcomes were assessed

Other bias

Unclear risk

No evidence of additional bias but reporting insufficient to be certain

Methods

2‐arm RCT

Setting: NR, appears to be single centre in Japan

Follow‐up: NR

Participants

16 children (aged 2‐12 years) undergoing appendectomy for perforated appendicitis

Inclusion criteria: generalised peritonitis or nonlocalised abscess

Exclusion criteria: NR

Interventions

After appendectomy, the peritoneal cavity was lavaged with 100 mL/kg (1500‐4000 mL) of the following warmed lavage solutions:

Group I : normal saline (8 participants)

Group II: acidic oxidative water (AOPW), a strong acidic water produced by the electrolysis of tap water containing 10% W/V sodium chloride (8 participants)

Co‐interventions: antibiotics moxalactam [reported as LMOX] (100 mg/kg/d) or cefazolin [reported as CEZ] (50 mg/kg/d) were given in both groups for 5 d or until serum C‐reactive protein was at a normal level

Outcomes

Primary outcome: SSI

No definition given for wound infection

Group I (saline): 4/8

Group II (APOW): 1/8

Secondary outcome: adverse events: abscess formation

Group I (saline): 1/8

Group II (APOW): 0/8

Secondary outcome: length of hospital stay (mean (SD) d)

Group IS (saline): 22.7 (11.1)

Group II (APOW): 12.1 (5.1)

Notes

Funding: NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "They were randomly divided into two groups"

Comment: no details of randomisation method

Allocation concealment (selection bias)

Unclear risk

Comment: no mention of allocation concealment

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Comment: no mention of blinding

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Comment: no mention of who performed the assessment of outcomes or whether blinding occurred

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Comment: all participants are included in the results

Selective reporting (reporting bias)

High risk

Comment: outcomes were not prespecified beyond "effectiveness and safety"

Other bias

Unclear risk

Comment: there is not enough methodological detail to judge

Methods

2‐arm RCT

Setting: appears to be multiple centres, Japan ("our affiliated hospitals")

Follow‐up: 30 d

Duration of study: 2008‐2012

Participants

44 children aged 3‐14

Group I: 16 boys and 4 girls, ranging in age from 4‐11 years

Group II: 12 boys and 12 girls, ranging in age from 3‐14 years

Inclusion criteria: children (age not defined) appendectomy for perforated appendicitis with extensive or panperitonitis

Exclusion criteria: pre‐operative antibiotics or requirement of antibiotics due to massive abscess formation

Interventions

After appendectomy, the peritoneal cavity was lavaged with 100 mL/kg saline or SAEW (strong acid electrolysed water, generated by electrolysis of tap water containing 0.2% NaCl), in Groups I and II, respectively. After closure of the fascial layer, the wound was washed out with 200 mL same solution before skin suture

Group I: 100 mL/kg saline (20 participants)

Group II: 100 mL/kg SAEW (24 participants)

Co‐intervention: cefmetazole, 100 mg/kg/d, was given initially to both groups, which was replaced by the most sensitive antibiotics after identification of causative pathogens for 5 or 7 d depending on response. The abdominal wall was disinfected with povidone iodine, and laparotomy was performed via a pararectal incision, saving the muscle layers, followed by appendectomy, carried out in the same manner in both groups.

Outcomes

Primary outcome: SSI

Defined as infection at the operation site, occurring up to 30 d after surgery, with confirmed causative pathogen(s) identical to those of the appendicitis.

Group I (saline): 4/20

Group II (SAEW): 0/24

Secondary outcome: adverse events (intraperitoneal abscess)

Group I (saline): 1/20

Group II (SAEW): 1/24

Secondary outcome: length of hospital stay (mean (SD) d)

Group I (saline): 9.4 (4.7)

Group II (SAEW): 8.7 (4.0)

Notes

34 participants were excluded from the study because some had received antibiotics before the operation and some had required antibiotics for massive abscess formation to be resected primarily with appendectomy

Funding: NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "Patients were allocated randomly to one of two treatment groups"

Comment: no details about method of randomisation

Allocation concealment (selection bias)

Unclear risk

Comment: no mention of allocation concealment

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Comment: no mention of blinding

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Comment: no mention of who performed the outcome assessment or whether blinding occurred

Incomplete outcome data (attrition bias)
All outcomes

High risk

Quote: "34 patients were excluded from the study..."

Comment: it is unclear if the exclusions were before or after randomisation

Selective reporting (reporting bias)

Low risk

Comment: all outcomes of interest appear to be reported

Other bias

Unclear risk

Comment: there is not enough methodological information to judge

Methods

3‐arm RCT

Setting: Kaiser‐Permanente Medical Center ‐ Santa Clara, USA

Follow‐up: at least 8 weeks

Participants

128 women undergoing cesarean section for various indications including repeat, breech and cephalopelvic disproportion. (132 entered study but 4 were excluded for irrigation protocol deviation and data are only presented for 128)

Inclusion criteria: undergoing cesarean section

Exclusion criteria: fever or other evidence of infection in labour, history of sensitivity to cephapirin or cefoxitin

Interventions

Group I: following delivery of the placenta, the uterine cavity and incision, bladder flap, pelvic gutters, and subcutaneous tissue were irrigated with 2 mg cephapirin in 1000 mL normal saline (44 participants)

Group II: irrigation with 2 mg cefoxitin in 1000 mL normal saline (41 participants)

Group III: irrigation with 1000 mL normal saline only (43 participants)

Outcomes

Primary outcome: SSI (defined as purulent wound discharge with or without wound separation)

Group I (cephapirin): 0/44

Group II (cefoxin): 0/41

Group III (saline): 3/43

Secondary outcome: endometritis (defined as persistent fever, uterine tenderness, foul‐smelling lochia, with no other obvious source of infection)

Group I (cephapirin):4/44

Group II (cefoxin): 1/41

Group III (saline): 5/43 (one of these also had wound infection)

Secondary outcome: length of hospital stay (mean (SD) d)

Group I (cephapirin): 4.8 (1.2)

Group II (cefoxin): 4.9 (1.9)

Group III (saline): 5.2 (2.1)

Notes

4 excluded from the analysis for protocol deviation

Funding: NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "... bags were sequenced randomly by a lottery method and used in numerical order"

Comment: the method is not explained in enough detail to know whether it was appropriate

Allocation concealment (selection bias)

Unclear risk

Quote: "... bags were sequenced randomly by a lottery method and used in numerical order"

Comment: there is not enough detail about the method of randomisation and allocation concealment to judge

Blinding of participants and personnel (performance bias)
All outcomes

Low risk

Quote: "...bags of irrigant were prepared by pharmacy personnel" "One millilitre of multivitamin infusion was added to create an identical appearance of all solutions" "Patients, physicians, operating room personnel, and data collectors were thus blinded to the group assignment."

Comment: steps were have been taken to ensure blinding of participants and personnel

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

Quote: "Patients, physicians, operating room personnel, and data collectors were thus blinded to the group assignment."

Comment: steps appear to have been taken to ensure blinding of outcome assessors

Incomplete outcome data (attrition bias)
All outcomes

Unclear risk

Quote: "One hundred thirty‐two patients were entered in the study. Four patients were eliminated from the statistical analysis because of deviations from the protocol of irrigation technique"

Comment: there are only a few participants lost during the study but no details of these participants are reported and since numbers of events are small this could potentially have an impact.

Selective reporting (reporting bias)

Low risk

Comment: the outcomes of interest appear to be fully reported

Other bias

Low risk

Comment: there is no evidence of other bias

Methods

2‐arm RCT

Setting: NR; appears to be single hospital in USA

Follow‐up: NR

Participants

200 participants undergoing elective and emergency gastrointestinal surgery (procedures on biliary tract 63/100 vs 57/100; gastroduodenal 14 vs 19 and colon 23 vs 24). Mean age was 61.7/61.6 years (range 17‐93). Malignancy present in 29/100 versus 25/100; diabetes 6/100 vs 8/100; obesity 30/100 vs 18/100

Inclusion criteria: elective or emergency procedures on the gastrointestinal tract

Exclusion criteria: parenteral antibiotics had been administered preoperatively or intraoperatively; a colostomy was required; if frank pus was encountered at operation

Interventions

Group I (kanamycin sulphate and cephalothin sodium): operative site was irrigated intermittently from the beginning to completion of the operation with a solution containing 1 g of kanamycin sulphate and 1 g of cephalothin sodium in 1000 mL of normal saline solution (100 participants)

Group II (saline): operative site was irrigated intermittently from the beginning to completion of the operation with a solution of normal saline (100 participants)

Cointerventions: the average volume of irrigant used for each operation was 750 mL. No cointerventions were reported

Outcomes

Primary outcome: SSI (postoperative wound infection)

Group I (kanamycin sulphate and cephalothin sodium): 3/100

Group II (saline): 9/100

Secondary outcome: mortality (postoperative deaths‐ last recorded 43 d post‐op)

Group I (kanamycin sulphate and cephalothin sodium): 5/100

Group II (saline): 3/100

Note: 2 participants who died had wound infections but sepsis was not the cause of the deaths

Secondary outcome: antibiotic resistance

Details of individual species recovered from wounds of participants with SSI were reported together with their sensitivity or resistance where this was tested for. Multiple species reported for each participant

Secondary outcome: adverse events

Postoperative peritonitis:

Group I (kanamycin sulphate and cephalothin sodium): 0/100

Group II (saline): 1/100

Secondary outcome: antibiotic resistance

Resistance of specific organisms to kanamycin sulphate and cephalothin sodium reported for each type of surgery. Large numbers of samples reported as not tested

Notes

Funding: NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Quote: "Designation was made by computer‐generated listing using a standard table of random numbers."

Comment: sequence generation used an appropriate method

Allocation concealment (selection bias)

Unclear risk

Quote: "Designation was made by computer‐generated listing using a standard table of random numbers."

Comment: no indication as to whether allocation was adequately concealed

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Quote: "a prospective, randomized, double‐blind study"

Comment: it was unclear who was blinded; while participants would probably be blinded it is unclear if personnel were

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Quote: "a prospective, randomized, double‐blind study"

Comment: it was unclear whether the double‐blinding referred to outcome assessors

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Comment: all randomised participants were included in the analyses

Selective reporting (reporting bias)

Unclear risk

Comment: outcomes were not defined in the methods section so it is difficult to be sure whether all planned outcomes were fully reported.

Other bias

Unclear risk

Comment: no evidence of other sources of bias but reporting insufficient to be certain

Methods

Parallel‐group RCT (factorial)

Setting: single hospital in the USA

Follow‐up: NR

Participants

100 women undergoing caesarean section

Inclusion criteria: women undergoing caesarean section; indications for surgery included elective repeat caesarean, failed trial of labour after prior caesarean, abnormal presentation, failure to progress, cephalopelvic disproportion, and severe pre‐eclampsia without thrombocytopenia or coagulopathy
Exclusion criteria: chorioamnionitis at caesarean, emergency caesarean for foetal distress with inadequate time for skin preparation

Interventions

Group I: saline irrigation (500 mL) of pelvis and subcutaneous tissue at uterine and fascial closure (50 participants)

Group II: cefazolin irrigation (1 g in 500 mL saline) of pelvis and subcutaneous tissue at uterine and fascial closure (50 participants)

Cointerventions: factorial randomisation to 2 alternative skin preparations: povidone iodine 7.5% scrub followed by povidone iodine 10% solution (standard skin preparation) versus 5‐min scrub with parachlorometaxylenol followed by povidone iodine scrub and solution (special skin preparation). No additional interventions were reported.

Outcomes

Primary outcome: SSI (hyperemic skin incision and fluctuant mass which when opened contained purulent material)

Group I (saline): 4/50 (3/25 with standard skin preparation; 1/25 with special skin preparation)

Group II (cefazolin): 2/50 (2/25 with standard skin preparation)

Secondary outcome: adverse events (endometritis)

Group I (saline): 30/50 (16/25 with standard skin preparation; 14/25 with special skin preparation)

Group II (cefazolin): 11/50 (8 with standard skin preparation; 3/25 with special skin preparation)

Notes

Funding: supported in part by the Vicksburg Hospital Medical Foundation

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Quote: "Random assignment was achieved by card selection from sealed opaque envelopes with group appointment derived from a random number table"

Comment: it appeared that an appropriate method was used to generate the randomisation sequence

Allocation concealment (selection bias)

Unclear risk

Quote: "Random assignment was achieved by card selection from sealed opaque envelopes with group appointment derived from a random number table"

Comment: it was not clear that enough measures were taken to ensure adequate concealment of allocation

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Comment: there was no information on whether personnel and participants were blinded to treatment allocation

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Comment: there was no information on who performed the outcome assessment or whether they were blinded to treatment allocation

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Comment: all randomised participants were included in the analyses

Selective reporting (reporting bias)

Unclear risk

Comment: outcomes were not clearly prespecified so difficult to determine if all planned outcomes were fully assessed

Other bias

Unclear risk

Comment: no obvious additional sources of bias but reporting insufficient to be certain

Methods

2‐arm RCT

Setting: single hospital in Australia

Follow‐up: letter or text message 4 weeks after surgery (following discharge from hospital)

Participants

3270 women undergoing caesarean section. Of those followed up 1508 had elective surgery and 1519 had surgery during labour. Mean age was 28.5 years in the Betadine group vs 28.6 years in the no Betadine group. 13.8% versus 12.9% had diabetes

Inclusion criteria: women undergoing caesarean section either elective or during labour (stratified randomisation)

Exclusion criteria: suspected or known allergy to iodine

Interventions

Group I: wound irrigation with 50 mL of 10% aqueous povidone iodine (Betadine) solution just before skin closure (1634 participants randomised; 1634 received allocated intervention; 1520 analysed)

Group II: no irrigation (1636 participants randomised; 1636 received allocated intervention; 1507 analysed)

Cointerventions: alcoholic povidone iodine for skin preparation unless an allergy to iodine present, in which case chlorhexidine was used. Prophylactic cephalothin administered to all women soon after spinal anaesthesia

Outcomes

Primary outcome: SSI (wound abscess or wound draining pus or sero‐sanguinous fluid, or redness, induration, warmth and tenderness or if woman’s general practitioner had seen her and prescribed antibiotics for presumed infection)

Group I (povidone iodine irrigation): 144/1520

Group II (no irrigation): 147/1507

Secondary outcome: return to theatre

Group I (povidone iodine irrigation): 7/1520

Group II (no irrigation): 9/1507

Secondary outcome: hospital readmission

Group I (povidone iodine irrigation): 39/1520

Group II (no irrigation): 30/1507

Completed case analyses reported here: ITT population 1634 vs 1636 ‐ used in analyses for SSI

Notes

Funding: states "we had no funding for this study..."

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Quote: "The allocation was prepared using computer generated list of random numbers using a variable block of 10 and performed by a staff member not part of the clinical team"

Comment: acceptable method of sequence generation; randomisation also stratified by elective versus non‐elective procedure

Allocation concealment (selection bias)

Low risk

Quote: "Women were randomised to ‘Betadine’ or ‘no Betadine’ group using sequentially numbered sealed opaque envelopes that contained the allocation..... After all layers were sutured and just prior to starting skin closure, the envelope with the allocation was opened by one of the theatre nurse[sic]"

Comment: appropriate measures appear to have been taken to ensure allocation concealment.

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Quote "In the control group (no Betadine group), layers would be sutured in exactly the same manner right up to the point of skin closure as it was only at this point that the allocation was revealed to the surgical team."

Comment: surgical personnel were aware of group allocation; unclear whether participants were blinded

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

Quote: "Information on outcome measures was obtained by the research team blinded to the allocation and also not involved in clinical care of the women"

Comment: blinded outcome assessment for all outcomes

Incomplete outcome data (attrition bias)
All outcomes

Unclear risk

Quote: "Of the total number randomised, 243 women were inadequately followed up either due to change of address, wrong telephone number or just not receiving the text messages"

Comment: 7% of women were lost to follow‐up; this was balanced between the arms (114 versus 129) and between the types of surgery (elective versus non‐elective) undertaken. Although this is close to the wound infection rate it does not appear likely to have impacted on the risk ratio of infection.

Selective reporting (reporting bias)

Low risk

Quote "Primary outcome was the incidence of SSI as a whole but also specifically readmission for intravenous antibiotics and/ or return to theatre for wound infection"

Comment: The outcomes were specified in the methods section and then fully reported

Other bias

Low risk

Comment: no specific quote but no evidence of other bias and reporting sufficient to be reasonably confident.

Methods

Three‐arm RCT

Setting NR; appears to be single hospital in Switzerland

Follow‐up: not clear beyond 4 d/discharge from hospital

Participants

162 participants with appendicitis

Inclusion criteria: people undergoing appendectomy carried out through an incision at McBurney's point, without the use of drains and without the use of pre‐, peri‐ or post‐operative systemic antibiotics

Exclusion criteria: people with perforated or gangrenous appendices or with peritonitis

Interventions

Group I (saline): irrigation with 500 mL saline (0.9%) after the closure of the peritoneum. The liquid was re‐aspirated and the wound was not swabbed. The skin was then closed.

Group II: (epicillin): irrigation with 500 mL saline with 1 g epicillin (Spectacilline) in solution as for Group I

Group III: (lincomycine): irrigation with 500 mL saline with 600 mL lincomycine (Lincocin) in solution as for Group I

Cointerventions: NR

Outcomes

Primary outcome: SSI (septic complications with spontaneous or induced purulent discharge): results only reported for all groups together (1/162) compared with a non‐randomised group without irrigation (7/158)

Secondary outcome: adverse events including abscess: results only reported for all groups together: 1/162 abscess from one of the antibiotic groups

Notes

Funding: NR

Paper in French; data extracted by one review author, checked by a fluent speaker

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Quote: "162 appendicectomies were randomised to treatment by blindly drawn lots"

Comment: acceptable method of sequence generation

Allocation concealment (selection bias)

Unclear risk

Quote: "162 appendicectomies were randomised to treatment by blindly drawn lots"

Comment: unclear whether there was adequate concealment of allocation

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

No direct quote; no information on whether any personnel were blinded after the treatment allocation

Blinding of outcome assessment (detection bias)
All outcomes

High risk

Quote "In the irrigated groups a questionnaire was sent to the treating physician to establish whether postoperative wound infections developed"

Comment: no information as to whether the physicians were aware of treatment allocation

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Comment: all randomised participants appeared in the analysis

Selective reporting (reporting bias)

High risk

Comment: the results for the three randomised groups were reported together and contrasted only with a non‐randomised comparison group

Other bias

Unclear risk

No other sources of bias were apparent but reporting insufficient to be certain

Methods

2‐arm RCT

Setting: a single hospital in Iran

Follow‐up: 1, 2, 4 and 6 weeks post surgery

Participants

102 participants (mean age: Group I: 50.63 years, Group II: 50.28 years) undergoing open cholecystectomy surgery

Inclusion criteria: cholecystitis diagnosed by surgeon

Exclusion criteria: age > 80 years, diabetes, immunosuppression (acquired or hereditary), history of immunosuppressive therapy, use of antibiotics during referral time for other reasons, history of recurrent cholecystitis, laparoscopic cholecystectomy, limitation for follow up [sic]

Interventions

Group I: open cholecystectomy, then before wound closure irrigation with 1 g cefazolin IV [sic] antibiotics (51 participants)

Group II: open cholecystectomy with no antibiotic irrigation before wound closure (51 participants)

Co‐interventions: same surgery and general anaesthesia

Outcomes

Primary outcome: SSI

Signs of infection included erythema, induration, tenderness, warmth, suppurative discharge

Group I: 6/51

Group II: 6/51

Notes

Funding: none reported

Reported in Persian. Data extraction and 'Risk of bias' assessment performed by translator

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Comment: they used a random number table

Allocation concealment (selection bias)

Unclear risk

Translator did not identify any information

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Translator did not identify any information

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Translator did not identify any information

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Translator judged all randomised participants included in analysis

Selective reporting (reporting bias)

Unclear risk

Translator did not identify any information

Other bias

High risk

Comment: had a potential source of bias related to the specific study design used, they did not have well‐defined outcome, the evaluation of some participants by telephone!

Methods

2‐arm RCT

Setting: NR; appears to be single hospital in Malaysia

Follow up at 2, 4 and 6 weeks post‐operatively for wound infection and adverse events

Participants

190 participants (178 analysed) undergoing CABG surgery. Mean age was 61.6 (7.6) years. Comorbidiities documented were diabetes (44.4%), chronic obstructive pulmonary disease (37.1%), end stage renal failure (18%) and obesity (11.2%)

Inclusion criteria: scheduled for elective CABG

Exclusion criteria: emergency cases, those who underwent other surgical procedures in addition to CABG, those allergic to Dermacyn, and those who had infective or other skin lesions over anterior chest wall area

Interventions

Group I (Dermacyn): Dermacyn wound irrigation (15‐min soak) upon chest closure and after insertion of sternal wires before subcutaneous tissue and skin closure (88 participants analysed)

Group II (povidone‐iodine): povidone‐iodine would irrigation (15‐min soak) upon chest closure and after insertion of sternal wires before subcutaneous tissue and skin closure (90 participants analysed)

Cointerventions: IV prophylaxis with 1.2 g Augmentin (amoxicillin and clavulanate) at induction. 2 drains were normally left in the mediastinal cavity.

Outcomes

Primary outcome: SSI (sternotomy wound infection, which was defined according to the Centers for Disease Control and Prevention system. Wound infections were graded as superficial (involving the skin and subcutaneous tissue of the incision), deep (involving fascia, muscle layers, and sternum), or deep organ space.

Group I (Dermacyn): 5/88 (5 superficial; 0 deep)

Group II (povidone‐iodine): 14/90 (10 superficial; 4 deep)

Secondary outcome: need for reoperation

Group I (Dermacyn): 0/88

Group II (povidone‐iodine): 4/90 (sternal dehiscence requiring surgical debridement and repair)

Secondary outcome: mortality

12 participants were described as having "dropped out"; 4 owing to postoperative mortality and 8 dropped for re‐opening of chest due to bleeding

Group I (Dermacyn): 7/95

Group II (povidone iodine): 5/95

Notes

Funding: NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "patients were consecutively randomized into 2 groups"

Comment: no information on how the randomisation sequence was generated

Allocation concealment (selection bias)

Unclear risk

Quote: "patients were consecutively randomized into 2 groups"

Comment: no information on whether allocation concealment was adequate

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Comment: no information on whether the participants or personnel were blinded

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Quote: "The sternotomy wounds were inspected on postoperative day 2 and daily until discharge. Patients were then followed up at 2, 4, and 6 weeks postoperatively to assess for the presence of wound infection and Dermacyn side effects."

Comment: no information on whether the inspections and follow‐up were carried out by blinded assessors

Incomplete outcome data (attrition bias)
All outcomes

High risk

Quote: "We recruited 190 patients for this trial, 95 patients in each group. Twelve patients, however, dropped out owing to postoperative mortality (4 cases, 2 deaths due to poor left ventricular function of < 20% and 2 deaths due to cerebrovascular accident) and chest re‐opened for bleeding (8 cases)"

Comment: the number of dropouts (exclusions) was close to the number of primary outcome events. The outcome of mortality can be assessed using the ITT population so is at low risk but the SSI outcome is at high risk.

Selective reporting (reporting bias)

Unclear risk

Quote: "The primary outcome was the presence of sternotomy wound infection,"

Comment: only the primary outcome was prespecified and it is difficult to be certain whether other planned outcomes were fully reported.

Other bias

Unclear risk

Comment: no evidence of other bias but reporting insufficient to be certain

Methods

2‐arm RCT

Setting: single hospital in USA

Follow‐up: carried out daily until discharge

Participants

260 randomised participants undergoing enterotomy during abdominal surgery. No further information on participant characteristics.

Inclusion criteria: enterotomy during abdominal surgery

Exclusion criteria: NR

Interventions

Group I (kanamycin): lavage prior to closure of the abdominal incision with 100 mL of 1% kanamycin; excess allowed to enter the peritoneal cavity (124 analysed participants; number randomised unclear)

Group II (saline): lavage prior to closure of the abdominal incision with 100 mL of saline; excess allowed to enter the peritoneal cavity (116 analysed participants; number randomised unclear)

Cointerventions: concomitant systemic antibiotics in approximately one fifth of wounds

Outcomes

Primary outcome: SSI (not defined)

Group I (kanamycin): 12/124

Group II (saline): 23/116

Secondary outcomes: mortality, reoperation

20 participants were excluded for early postoperative death, reoperation or delayed primary closure. Group allocations and numbers excluded for each reason were not reported

Secondary outcome: antibiotic resistance to kanamycin in wound culture

Group I (kanamycin): 12/12

Group II (saline): "over half" of 23

Secondary outcome: adverse events

Respiratory depression was only event reported on:

Group I (kanamycin): 2/124

Group II (saline):4/116

Notes

Funding: NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "The solution, either 1% kanamycin or saline, was administered according to an established randomized schedule"

Comment: no information on how the randomisation schedule was generated

Allocation concealment (selection bias)

Unclear risk

Quote: "The solution, either 1% kanamycin or saline, was administered according to an established randomized schedule"

Comment: no information on whether the allocation was adequately concealed

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Quote: "an established randomized schedule under a double‐blind protocol"

Comment: unclear who was blinded to the allocations

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Quote: "The appearance of the wound was graded daily until discharge by a single observer and any deviation from optimal healing was documented serially with photographs"

Comment: unclear whether the single observer was blinded to treatment allocation

Incomplete outcome data (attrition bias)
All outcomes

High risk

Quote: "of the 260 patients admitted to the study, 20 were excluded for reasons of early postoperative death, reoperation or delayed primary wound closure"

Comment: the number of exclusions was comparable to the numbers of infections in each group; the group allocation of excluded participants was not reported.

Selective reporting (reporting bias)

High risk

Comment: outcomes were not defined in the methods section so it is difficult to be sure whether all planned outcomes were fully reported. Some outcomes were not fully reported with data for each treatment group.

Other bias

Unclear risk

Comment: no evidence of other sources of bias but reporting insufficient to be certain

Methods

2‐arm RCT

Setting: unclear but appears to be a single centre in Germany

Follow up: NR

Participants

197 participants undergoing elective colorectal resection

Inclusion criteria: NR

Exclusion criteria: NR

Interventions

Group I: irrigation with polyhexanide 0.04% solution before final wound closure (101 participants)

Group II: irrigation with Ringer's solution before final wound closure (concentration not given) (96 participants)

Cointerventions: NR

Outcomes

Primary outcome: SSI (not defined)

Group I (polyhexanide): 19/101

Group II (Ringer's solution): 22/96

Notes

Funding: NR

Abstract only

States "interim analysis was done after 250 patients were screened and is presented here."

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote "This study was conducted as a double blind, randomized, single center study"

Comment: no information on how the randomisation sequence was generated

Allocation concealment (selection bias)

Unclear risk

Quote: "This study was conducted as a double blind, randomized, single center study"

Comment: no information on whether the allocation sequence was adequately concealed

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Quote "This study was conducted as a double blind, randomized, single center study"

Comment: although described as double‐blind it is unclear who was blinded and whether the blinding was adequate

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Quote "This study was conducted as a double blind, randomized, single center study"

Comment: although described as double‐blind it is unclear who was blinded and whether the blinding was adequate

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Quote: "A total number of 197 elective colorectal resections were randomized. 101 patients received verum. Univariate analysis was followed by multivariate analysis where appropriate. Results: There were 41 wound infections in 197 patients (20.8%). 19 in the verum group, 22 in the control group (p=0.478)."

Comment: it appeared that all randomised participants were included in the analysis.

Selective reporting (reporting bias)

Unclear risk

Quote: "Primary endpoint was the rate of SSI in each group"

Comment: there is too little information to be sure if all planned outcomes were assessed and reported.

Other bias

High risk

Comment: an interim analysis. This is an abstract and there is too little information to determine if there were other additional sources of potential bias.

Methods

Setting: 2 tertiary hepatobiliary and pancreatic surgery units in Australia

Follow‐up: at 1 and 2 weeks following surgery, and thereafter as indicated. Minimum follow‐up 1 month after surgery

Participants

137 consecutive participants enrolled, undergoing major elective open abdominal operative procedures; 128 assigned to treatment (median age 63 years, range 18‐86 years)

Inclusion criteria: adults undergoing an elective open abdominal operation anticipated to extend beyond 2 h

Exclusion criteria: those undergoing laparoscopic procedures

Interventions

In all cases, prior to abdominal closure, the peritoneal cavity was irrigated with least 3 L of warm saline without any added antibiotics. participants then received the following treatment after randomisation:

Group I (pulse irrigation): surgical irrigation device (Stryker Instruments, Portage, MI) used after fascia closure to irrigate the surgical wound with 2 L of normal saline at room temperature; pressure close to (but not exceeding) 15 psi delivered through cone‐shaped applicator

Group II (saline): following closure of the fascia in the standard group, 2 L of normal saline at room temperature was poured into the subcutaneous tissue without any agitation

Excess fluid was removed from the subcutaneous tissue with application of a dry pack. Subcutaneous drainage or closure was not undertaken. The skin was reapproximated with continuous subcuticular 3/0 Monocryl sutures. Skin staples were not used in any case. A Duoderm dressing was applied to the wound.

Cointerventions: all participants received dexamethasone phosphate 8 mg IV as part of routine antiemetic prophylaxis.

At induction of anaesthesia all participants received ampicillin 1 g IV, gentamicin IV (2 mg/kg), and metronidazole 500 mg IV. Antibiotics were continued for 24 h postoperatively. In cases of penicillin allergy, vancomycin 1 g IV or cefazolin [cephazolin] 1 g IV was administered according to the particular sensitivity reaction.

Outcomes

Primary outcome: SSI

Wound infection defined as: (1) purulent drainage, with or without laboratory confirmation, from the superficial incision; (2) organisms isolated from an aseptically obtained culture of fluid or tissue from the superficial incision; (3) at least one of the following signs or symptoms of infection: pain or tenderness, localised swelling, redness, or heat and superficial incision is deliberately opened by surgeon, unless the incision was culture‐negative; (4) diagnosis of superficial incisional SSI by the surgeon or the attending physician

Group I (pulse irrigation): 4/66 (all superficial)

Group II (saline): 12/62 (2 required major debridement with prolonged course of dressings; one had partial abdominal wall dehiscence)

Primary outcome: wound dehiscence within 30 d of operation

Group I (pulse irrigation): 0/4

Group II (saline): 1/12

Secondary outcome: participants with postoperative SSI using systemic antibiotics within 30 d of surgery

14/16 participants; no data regarding which treatment group these were associated with

Secondary outcome: antibiotic‐resistant infections

Details of organisms isolated but not of resistance were reported, this was not group data

Secondary outcome: adverse events

Cellulitis (without wound cultures): 1/16; no data regarding which treatment group this was associated with

Any non‐wound‐related complication

Group I (pulse irrigation): 32/66

Group II (saline): 23/62

Secondary outcome: surgical reintervention

Relaparotomy

Group I (pulse irrigation): 3/66

Group II (saline): 3/62

Debridement of wounds with SSI

Group I (pulse irrigation): 0/4

Group II (saline): 2/12

Secondary outcome: length of stay

Group I (pulse irrigation): median 9 (range 4‐71) d

Group II (saline): median 9 (range 5‐45) d

Secondary outcome: hospital readmissions

Group I (pulse irrigation): 9/66

Group II (saline): 6/62

Notes

Funding: supported by a University of Melbourne, Early Career Development Grant, awarded to lead author. No supplementary support was provided by Stryker

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "Grouping allocation was determined by a sealed envelope selection. Blocks of 20 patients were randomized at one time. Diabetic patients were randomized separately to achieve close to even distribution in each group."

Comment: no information on how randomisation schedule was generated.

Allocation concealment (selection bias)

Unclear risk

Quote: "Grouping allocation was determined by a sealed envelope selection."

Comment: unclear whether opaque envelopes were used

Blinding of participants and personnel (performance bias)
All outcomes

High risk

No direct quote; no information on blinding

Comment: it is unclear whether participants and personnel were blinded; personnel unlikely to be blinded after randomisation due to difference in intervention procedures between arms

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Quote: "Patients were monitored by a dedicated acute pain service and reviewed daily for any complications arising from their analgesic regime."

Comment: unclear whether personnel were blind to treatment allocation

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Quote: "A total of 137 patients were enrolled, as 8 cases did not reach the 2 h duration required for randomization."

Comment: study flow diagram indicates that 9 were excluded because they did not conform to inclusion criterion of needing to be > 2 h in duration; there is a discrepancy in the report text stating 8 were excluded for this reason. Attrition bias judged to be low risk given that all others conforming to inclusion criteria were analysed

Selective reporting (reporting bias)

Unclear risk

Comment: unclear whether all prescribed outcomes were reported

Other bias

Unclear risk

Comment: no evidence of other sources of bias but reporting insufficient to be certain

Methods

2‐arm RCT

Setting: single hospital in Switzerland

Follow‐up: NR

Participants

540 randomised participants undergoing surgery

Inclusion criteria: participants undergoing surgery

Exclusion criteria: oto‐rhino‐laryngeal surgical cases; thyroid surgeries; day‐case surgeries

Interventions

Group I (saline): irrigation with saline. 2 rinses were performed; the first of the operative site, and the second performed before skin closure (273 participants ‐ information derived from graph)

Group II (povidone‐iodine): irrigation with Betadine‐R solution (10% PVP‐iodine with 1% available iodine, diluted in a 1/10 solution). 2 rinses were performed; the first of the operative site, and the second performed before skin closure. (267 participants ‐ information derived from graph)

Cointerventions: after 1 min the excess liquid was re‐aspirated. In all participants skin disinfection was carried out using a standard Betadine‐R solution.

Outcomes

Primary outcome: SSI (not defined)

Group I (saline): 15/273

Group II (Betadine): 16/267

Notes

Paper in French; data extracted by one review author, checked by a fluent speaker

SSI data presented graphically across 2 figures; these data extracted using graph‐reader software

Funding: NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote "A prospective randomized study was undertaken......a draw of two groups on the basis of a pre‐established list, was carried out at entrance to theatre"

Comment: not clear how the randomisation sequence was established

Allocation concealment (selection bias)

Unclear risk

Quote "A prospective randomized study was undertaken......a draw of two groups on the basis of a pre‐established list, was carried out at entrance to theatre"

Comment: not clear whether allocation sequence was adequately concealed

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Quote " all information about the participant, the type of interventions, the personnel, the bacteriological findings and the antibiotic therapies was stored on a computer"

Comment: no information as to whether personnel were blinded to interventions

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

No direct quote but no information as to who determined the presence of SSI or whether they were blinded to treatment allocation

Incomplete outcome data (attrition bias)
All outcomes

Unclear risk

No direct quote: no attrition is reported but results are presented only graphically and it is difficult to determine whether all randomised participants were represented in the results.

Selective reporting (reporting bias)

High risk

No direct quote: results are not adequately reported for each group and it is very unclear if all outcome data are fully reported.

Other bias

Unclear risk

No other source of bias was apparent but the reporting was insufficient to be certain.

Methods

Parallel‐group RCT

Setting: single hospital in Denmark

Follow‐up: mean 8 d (5‐16 d)

Participants

33 participants undergoing surgery for perforated appendix

Inclusion criteria: all patients admitted for surgery for perforated appendix with diffuse peritonitis. Visible perforation of appendix with free pus in peritoneum, verified by microbiological culture of extracted peritoneal material. Normal function of kidneys in serum‐creatinin concentration demanded.

Exclusion criteria: pregnant women, < 15 and > 75 years, with malignant disorders and known allergy to penicillin

Interventions

Group I: no irrigation

Group II: postoperative peritoneal flushing with 1 g ampicillin per L flushing fluid

Group III: postoperative peritoneal flushing without ampicillin in the flushing fluid

Cointerventions: all participants had systemic antibiotics: ampicillin 2 g 4/d IV, gentamycin 1.5 mg/kg weight as the first dose – then 1 mg/kg weight 3/d IM, and clindamycin 600 mg 3/d IM

Outcomes

Primary outcome: SSI

Group I (no irrigation): 4/10

Group II (ampicillin irrigation): 3/10
Group III (saline irrigation): 2/10

Secondary outcome: adverse events (intra‐abdominal abscess)

Group I (no irrigation): 1/10

Group II (ampicillin irrigation): 010
Group III (saline irrigation): 0/10

Secondary outcome: length of stay

Group I (no irrigation): 14 d (8‐22)

Group II (ampicillin irrigation): 13 d (9‐20)
Group III (saline irrigation): 13 d (10‐22)

Notes

Funding: NR

Data extraction and 'Risk of bias' assessment performed by translator from the Danish; some aspects discussed with a review author

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Comment: described by one Danish word; the envelope system

Allocation concealment (selection bias)

Unclear risk

Comment: described by one Danish word; the envelope system

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Comment: Groups II and III had additional catheter added for flushing fluids. Group II also had ampicillin added to the flushing fluid. Personnel would therefore be aware of allocation, unclear if participants were also aware

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Comment: no description of whether assessors were the same as the personnel

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Comment: all 30 participants were included in analysis

Selective reporting (reporting bias)

Low risk

Comment: all outcomes mentioned in the methods section were reported. In the discussion they reported that no kidney or liver dysfunction was observed

Other bias

Unclear risk

Comment: none identified

Methods

Three‐arm RCT

Setting: single hospital in Spain

Follow‐up: unclear

Participants

51 women undergoing axillary lymph node dissection for breast neoplasm. Mean age 55.6 years

Inclusion criteria: diagnosis of breast neoplasm and plans to undergo an elective axillary lymph node dissection of Berg’s levels I and II because of axillary metastases determined pre‐operatively by core biopsy or evidence of metastasis in the sentinel lymph node biopsy (SLNB) in the intra‐operative or differential analysis

Exclusion criteria: allergy to any of the antibiotic drugs to be used, chronic renal failure secondary to possible toxicity of gentamicin, and planned modified radical mastectomy

Interventions

Group I: 2 lavages with 500 mL of physiologic saline (17 participants)

Group II: lavage with 500 mL of saline followed by lavage with 500 mL of a 240‐mg gentamicin solution (17 participants)

Group III: lavage with 500 mL of saline followed by lavage with 500 mL of a 600‐mg clindamycin solution (17 participants)

Cointerventions: pre‐operative systemic antibiotics (amoxicillin‐clavulanic acid 2 g IV; a single dose within 30 min of incision) were employed in all groups. Once the dissection was finished, a Redon drain was placed and connected to a low‐pressure vacuum device (primary closure was undertaken)

Outcomes

Primary outcome: SSI (incisional, not further defined)

Group I (2 x saline lavages) 0/17

Group II (saline then 240 mg gentamicin lavages) 0/17

Group III (saline then 600 mg clindamycin lavages) 0/17

Secondary outcome: mortality

Group I (2 x saline lavages) 0/17

Group II (saline then 240 mg gentamicin lavages) 0/17

Group III (saline then 600 mg clindamycin lavages) 0/17

Secondary outcome: length of hospital stay

Group I (2 x saline lavages) median 3 d (range 1‐3)

Group II (saline then 240 mg gentamicin lavages) median 3 d (range 1‐3)

Group III (saline then 600 mg clindamycin lavages) median 3 d (range 1‐3)

Notes

Funding: NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote "The patients were randomized by means of an Internet module into three groups"

Comment: mechanism of the internet module used for randomisation unclear

Allocation concealment (selection bias)

Unclear risk

Quote "The patients were randomized by means of an Internet module into three groups"

Comment: no information as to whether the allocation was adequately concealed

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Quote: "The patients were blinded as to whether they received saline, gentamicin, or clindamycin."

Comment: participants were blinded but unclear whether personnel were also blinded

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Comment: there is insufficient information to determine who performed the assessment of outcomes and whether they were blinded to treatment allocation.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Comment: it appears that all the randomised participants were included in the analyses.

Selective reporting (reporting bias)

Unclear risk

Comment: the outcomes reported were not prespecified in the methods so it is difficult to determine whether all planned outcomes were fully reported.

Other bias

Unclear risk

Comment: there is no evidence of other sources of bias but reporting is insufficiently detailed to be sure.

Methods

2‐arm RCT

Setting: NR but appears to be a hospital in Turkey

Follow‐up: NR

Participants

14 participants with perforated appendix (among 279 undergoing appendectomy for acute appendicitis)

Inclusion criteria: perforated appendix, undergoing appendectomy for acute appendicitis

Exclusion criteria: appendix not perforated

Interventions

Group I: peritoneal lavage with irrigation and aspiration (7 participants)

Group II: aspiration alone (7 participants)

Co‐interventions: NR

Outcomes

Primary outcome: SSI

Wound infection

Group I (lavage): 2/7

Group II (aspiration only): 0/7

Secondary outcome: adverse events

(1 intra‐abdominal abscess, 1 postoperative ileus)

Group I (lavage): 1/7

Group II (aspiration only): 1/7

Notes

Abstract only but further information received via study author correspondence

Funding: NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote (via correspondence): "The randomisation method is envelope method"

Comment: unclear how the randomisation sequence was generated

Allocation concealment (selection bias)

Unclear risk

Comment: insufficient information on use of envelopes to be sure whether allocation was adequately concealed

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Comment: no information on whether participants or personnel were blinded but personnel must have been aware due to differences in intervention

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

Comment: no information on who performed the assessment or whether they were blinded to treatment allocation

Incomplete outcome data (attrition bias)
All outcomes

Low risk

It appears that all randomised participants were included in the analysis.

Selective reporting (reporting bias)

Unclear risk

It is unclear whether all planned outcomes were fully reported; we contacted the study author in order to fully report SSI outcome on a per‐group basis.

Other bias

Unclear risk

Abstract only and reporting insufficient to be certain there were no other sources of bias.

Methods

Parallel‐group RCT

Setting: single medical centre in USA

Follow‐up: at least 2 weeks postoperatively

Participants

207 women undergoing caesarean section

Inclusion criteria: non elective caesarean section

Exclusion criteria: allergic to penicillin or cephalosporins, already on antibiotics, ongoing infection, requiring bacterial endocarditis prophylaxis

Interventions

Group I: cefazolin lavage; 2 g in 1000 mL saline administered to the uterine incision (300 mL), bladder flap (200 mL), abdominal gutters (200 mL) and abdominal incision (remaining fluid)

Group II: cefamandole lavage; 2 g in 1000 mL saline administered to the uterine incision (300 mL), bladder flap (200 mL), abdominal gutters (200 mL) and abdominal incision (remaining fluid)

2 additional groups were included in the trial using the same antibiotics delivered IV with saline lavage. These groups are not relevant to this review and we did not extract any data relating to them.

Cointerventions: normal saline bolus IV after cord clamped

Outcomes

Primary outcome: SSI (presence of cellulitis and/or purulent exudate)

Group I (cefazolin): 2/59

Group II (cefamandole) 0/54

Notes

Funding: NR

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Quote: "Patients were randomly assigned to one of four treatments"

Comment: no information on how the randomisation sequence was generated or whether an appropriate method was used.

Allocation concealment (selection bias)

Unclear risk

Quote: "Patients were randomly assigned to one of four treatments"

Comment: no information on whether there was adequate allocation concealment.