Table 7.

PICO Analysis for data extraction of the studies meeting the inclusion criteria [15]

Study Name	Akinyoola et al. (2011) Prospective RCT [19]
Participants	• 106 consecutive patients undergoing ‘clean, elective foot and ankle surgery’ although over 74% were trauma cases e.g. open fracture requiring open reduction and internal fixation (ORIF) thus diminishing the validity of measure as the researchers are no longer measuring what the study sets out or claims. • Surgery performed within five days of admission, meaning the patients were kept in hospital prior to and after surgery, which could raise the infection rates due to the potential to be colonised before and after surgery on the ward, around other unwell or infected patients. • Surgery performed using a thigh tourniquet not ankle.
Intervention	• Perioperative timing of antibiotic delivery. • IV Cefuroxime – did not state specific doses just ‘adjusted for age/weight’. • Researchers also gave patients up to three postoperative doses of IV antibiotics.
Control	• Group 1: Antibiotics received 5 min before inflation of tourniquet (ABT). • Group 2: Antibiotics received 1 min after inflation of tourniquet (AAT).
Outcomes	• Statistically significant postoperative infection rates of 14.8% (ABT) and 3.9% (AAT). • Operating surgeons (main researchers) determined the presence of infection thereby introducing experimenter bias at this stage in the study which may affect the internal validity of the findings.
Study Name	Butterworth et al. (2017) Prospective Cohort Study [18]
Participants	• 4238 patients who underwent foot and ankle surgery. • Prospectively rather than retrospectively reviewed. • Multicentre across Australia therefore high level of generalisability. • No randomisation, patients received or did not receive antibiotics dependant on their medical status, outlined procedure and surgeon’s decision. Therefore, patients who received postoperative antibiotics may have been of a less stable ASA Grading.
Intervention	• Patients receiving no antibiotics. • Patients receiving only preoperative antibiotics. • Patients receiving only postoperative antibiotics. • Patients receiving both pre-& postoperative antibiotics. • No mention of antibiotic type e.g. IV or Oral or agent. • No stratification of procedure location or type e.g. rearfoot or forefoot; fusion, osteotomy or toenail avulsion. • Cannot confirm whether patients actually took postoperative antibiotics.
Control	• No control owing to study design.
Outcomes	• Postoperative infection rates primary measure using an established and validated tool to capture data. • Comparisons amongst ASA Grade, age, gender, surgeon experience as predictors of infection. The latter being the only predictor with less rate of SSI with more experienced surgeons, perhaps likely due to a shorter duration of surgery with more experienced hands. • No antibiotics (2.5%). • Preoperative antibiotics (1.1%). • Postoperative antibiotics (2.6%). • Pre-& postoperative antibiotics (2.1%).
Study Name	Dayton et al. (2015) Consensus Panel (Expert Opinion) [34]
Participants	• 5 member panel of expert foot and ankle surgeons who received an invitation from the American College of Foot and ankle Surgery (ACFAS) to form part of a consensus group.
Intervention	• The panel worked to evaluate via email, telephone, face to face discussions the pertinent literature to antibiotic prophylaxis in foot and ankle surgery and found only 6 studies met their inclusion criteria after their literature search. • A modified Delphi method via a Likert Scale technique was then utilised to reach a consensus upon 13 pressing clinical questions put together by the panel chair.
Control	• No control owing to study design.
Outcomes	Panel Consensus:
	• It is appropriate to give antibiotic prophylaxis within foot and ankle surgery where the procedure involves bone, hardware and prosthetic joints. • It is generally accepted that patients at an increased risk of infection should be considered for antibiotic prophylaxis and that patient factors drive the decision to provide chemoprophylaxis over the type of procedure being performed, unless the type of procedure is considered to be a prolonged case (e.g. over 90 min). • Narrow spectrum antibiotics covering Staphylococcus aureus should be used for antibiotic prophylaxis in patients without a history of resistant infection. • It is not always appropriate to routinely perform preoperative nasal swabs to check for MRSA colonisation. • In cases where chemoprophylaxis is deemed necessary, antibiotics should be administered within 60 min prior to tourniquet inflation and discontinued within 24 h after surgery.
Study Name	Deacon et al. (1996) Prospective Cohort Study [20]
Participants	• 25 ASA 1 patients undergoing ‘bunionectomy procedures’ under the age of 55 with no previous history of ‘bunionectomy’ upon the same foot. • No explanation or stratification of type of procedure i.e. was metalwork used?
Intervention	• IV administration of 1 g of Cefazolin 60 min prior to tourniquet inflation. • Evaluated concentration of drug within 1st metatarsal head and compared to minimum inhibitory concentration required for inhibition of 90% of Staphylococcus taph Aureus in vitro (MIC90)
Control	• No control group utilised.
Outcomes	• IV administration of 1 g of Cefazolin led to mean bone concentration levels of 2.39 ± 1.19 (sd) μg/g. • This translates to approximately two to four times the required amount to exert a bacteriostatic effect at the site of surgery (0.5–1.0 μg/g required MIC90). • This was not a study to assess postoperative infection rates, however no infections were encountered. • Researchers concluded that 1 g of IV Cefazolin 60 min prior to tourniquet inflation will provide bone concentration levels of antibiotic that are adequate in principle to inhibit colonization Staphylococcus taph Aureus. • This study used ASA 1 patients therefore findings are based on the assumption that an unremarkable medical history and peripheral vascular status, with adequate circulation will achieve the required concentration levels as serum concentration was not individually measured in this study.
Study Name	Dounis et al. (1995) Prospective Cohort Study [21]
Participants	• 67 patients undergoing foot and ankle surgery with no hepatic/renal insufficiency, known allergies to cephalosporin/penicillin antibiotics or concurrent steroid/immunosuppressant/antibiotic therapy. • Procedures included: 10 ankle fusions, 7 triple arthrodesis of rearfoot, 12 Keller’s arthroplasties, 15 Chevron osteotomies, 16 Mitchell osteotomies and 7 metatarsal head excisions on patients ranging from 39 to 72 with a mean age of 58.9 years old ±9.2 (sd). • All procedures carried out using thigh tourniquet.
Intervention	• Systemic IV vs Local IV Antibiotics (distal to tourniquet). • Intervals of 10 min, 20 min, 2 h and 4 h prior to tourniquet inflation in the systemic group. • Ceftazidime 2 g vs Ceftriaxone 2 g in each group although in the local group were mixed with 25 ml of 0.9% Saline to “fill empty intravascular space” no details of how they calculated weight/height to reach this figure, negatively affects internal validity making it difficult to draw comparisons as antibiotics preparations not controlled for. • Evaluated bone and soft tissue concentrations of participants 10–15 min after skin incision.
Control	• Group 1: Antibiotics received IV systemic at stratified intervals of 10 min, 20 min, 2 and 4 h prior to tourniquet inflation. • Group 2: Antibiotics received IV local after tourniquet inflation but no record of time was described or which distal vein was utilised within methodology.
Outcomes	• Ceftazidime 2 g & Ceftriaxone 2 g demonstrated similar bone and soft tissue concentrations in each group. • In the systemic IV group the highest concentration of antibiotic in bone & soft tissue was yielded 20 min before tourniquet application. • In the local IV group the concentration of antibiotic in bone and soft tissue was consistently between 4 and 12 times higher than systemic IV administration (P > 0.001).
Study Name	Kurup (2016) Conference Paper Abstract Only [35]
Participants	• 340 patients undergoing elective forefoot surgery where perioperative antibiotics were deemed necessary by the operating clinician were reviewed retrospectively between 2011 and 2015. • Procedures ranged from digital fusion, osteotomy and joint replacement. • No description of ASA grade or antibiotic regimen/timings.
Intervention	• Procedures “involving metal/resorbable implant or prosthesis” were undertaken with patients receiving 400 mg of IV Teicoplanin.
Control	No control owing to study design.
Outcomes	• No cases of deep infection. • 1.7% infection rate with 6 superficial soft tissue infections recorded. • Due to nature of publication i.e. abstract only- high level of detail lacking.
Study Name	Mangwani et al. (2016) Prospective RCT [27]
Participants	• 100 patients undergoing lesser toe fusion in which an external k-wire was to be left in situ post-surgery for 4–6 weeks, with a mean age of 58 (group one) ± 17.5 (sd) and 62.7 (group two) ± 14.7 (sd) years old. • No stratification for ASA grade/co-morbidities.
Intervention	• Stratified random allocation of prophylactic Flucloxacillin (or Teicoplanin where an allergy was present). • No description or detail of timing/dose/route other than administration remained consistent with operating clinician’s standard preferences.
Control	• Group 1: Received prophylactic antibiotics • Group 2: Did not receive prophylactic antibiotics
Outcomes	• Group 1: 3 infections (6.2%) 93 toes, 4 diabetics + 1 immunosuppressed patient on methotrexate for sero-negative spondyloarthropathy (PSA) with a high BMI. • Group 2: 1 infection (1.9%) 78 toes, 2 diabetics. • Results not found to be significant after statistical analysis. • Difference in % of patients more susceptible to infection across 2 groups may account for study findings and may not accurately predict likelihood of infection rate with/without antibiotic prophylaxis for this type of surgery, acknowledged by the researchers.
Study Name	Pace et al. (2016) National Survey [36]
Participants	• 112 Orthopedic clinicians from the US were mailed a survey with respect to foot surgery and the use of antibiotics and percutaneous k-wires. • A total of 64 (57%) completed the survey with a mean of 15.2 years post fellowship acquisition.
Intervention	• Mail survey consisted of three clinical scenarios pertaining to use of percutaneous k-wire fixation and minimising postoperative pin tract infection: non-diabetic, diabetic, diabetic with sensory neuropathy. • Evaluated duration of k-wire placement, whether postoperative antibiotics were considered appropriate and which antibiotics are routinely used. • No questions evaluating the use of preoperative antibiotics.
Control	No control owing to study design.
Outcomes	• First case: 25% would use postoperative antibiotics, mean 9.4 days. • Second case: 28% would use postoperative antibiotics, mean 13.8 days. • Third case: 32% would use postoperative antibiotics, mean 14.5 days. • Majority of clinicians would utilise oral Cephalexin postoperatively. • None of the findings were shown to be statistically significant in this study.
Study Name	Reyes et al. (1997) Retrospective Review [37]
Participants	• 459 cases of foot and ankle surgery were reviewed between 1993 and 1996. • Data was gathered on the use of antibiotics, tourniquet, anaesthetic and patient demographics/medical history and type of surgery in order to explore postoperative infection rate and trends in practice.
Intervention	• The following data was recorded from patient records and used for descriptive analysis: antibiotic prophylaxis, age, gender, medical history, procedure duration, fixation used, tourniquet use and postoperative infection. • Majority of cases used Cefazolin 1 g or Vancomycin in those allergic.
Control	No control owing to study design.
Outcome	• Overall infection rate of 0.65%. • Infection rate with antibiotic prophylaxis was 0.43%. • Infection rate without antibiotic prophylaxis was 0.88%. • Not found to be statistically significant. • 50.7% of cases received IV antibiotics 30–60 min prior to surgery, the remaining did not receive any antibiotics. • 82% of cases involving bone and internal fixation were dosed with preoperative antibiotics. • Average tourniquet time of 68.72 min used at ankle and thigh. • No description of type of surgery/co-morbidities/ASA grade etc.
Study Name	Tantigate et al. (2016) Retrospective Comparative Study [28]
Participants	• Retrospective chart review of 1933 ft and ankle procedures in 1632 patients over a 56 month period.
Intervention	• Demographic data, type of antibiotics/dosage/timing were recorded along with rate of postoperative infection.
Control	No control owing to study design.
Outcomes	• When antibiotics were administered between 15 and 60 min prior to incision, there was a 2.7 – fold, statistically significant higher rate of postoperative infection as compared to the group of patients who received antibiotics< 15 min before incision (P < 0.05). • Independent predictors of postoperative infection were ASA Grade, non-ambulatory surgery and lengthier duration of surgery, with almost 92% of the risk of a postoperative infection being predicted by these factors (P > 0.05). • Suggests host factor may play a bigger role in predicting risk of postoperative infection than timing of antibiotics, though did not control for these host factors in a prospective and/or randomised manner therefore difficult to draw any causal relationship, moreover no description of type of antibiotics/timing/dose/regimen/type of surgery or ASA grade in abstract published.
Study Name	Zgonis et al. (2004) Retrospective Review [22]
Participants	• 555 patients who received elective foot and ankle surgery between 1995 and 2001. • Patients who had prior ulcerations, infection or trauma to the foot and ankle were excluded. • Patients were stratified into 6 categories: soft-tissue, digital, lesser metatarsal, first ray, rearfoot and multiple surgeries (more than one procedure).
Intervention	• The following data was recorded from patient records and used for statistical analysis: antibiotic prophylaxis, age, gender, medical history, procedure duration, fixation used, tourniquet use and postoperative infection.
Control	No control owing to study design.
Outcome	• IV antibiotics for prophylaxis were stated to be ordered 30 min prior to incision but administration varied from between 2 h prior to incision and just before surgery, meaning that either there has been administration before the drug order was put in or an error in the recording. Nonetheless, these are the details documented in the researcher’s findings, meaning that unfortunately this was not standardised or consistent, which is unfortunately a common attribute of the study design utilised. • The overall infection rate was 3.1%. • 55.1% of patients received antibiotic prophylaxis in this study with an infection rate of 1.6%. • 44.9% of patients did not receive antibiotic prophylaxis and rate of infection was found to be 1.4% postoperatively. • This was not deemed to be statistically significant, though to demonstrate this the power of the study would need to be significantly improved with a considerably larger sample size of 3452 patients for each group (receiving/not receiving antibiotics) totalling almost 7000 patients. This would be challenging to undertake both practically and ethically with certain surgical institutions having to follow local hospital policy and therefore recruiting over 3000 patients into one of the study groups which would then receive an intervention which goes against local policy may prove an arduous task to justify.