Antibiotic prophylaxis in foot and ankle surgery: a systematic review of the literature

Abstract

Background

With the advent of bacterial resistance, it is important now more than ever to evaluate use of antibiotic chemoprophylaxis in foot and ankle surgery. Within this area of the body there may be less dissection, surgery time with smaller incisions and importantly smaller sizes of implanted fixation as compared to other bone and joint procedures. Our objective was to systematically evaluate the quality of evidence behind existing guidelines.

Methodology

A systematic literature search was performed: MEDLINE, CINHAL, EMBASE and the Cochrane library from 1990 up to March 2018. To avoid omitting any studies on the subject, Google Scholar was also used. The inclusion criterion were studies exploring perioperative antibiotic use, postoperative infection rates in elective foot and ankle surgery and studies associated with this subject evaluating antibiotic use in clean elective foot and ankle surgery. The exclusion criterion being studies upon contaminated or dirty surgery or those which were inclusive of procedures proximal to the foot and ankle.

Results

Overall 11 studies met the inclusion criteria. From the grading of evidence, 2 level one and 4 level two studies were recognised. These studies ranked relatively highly in comparison to 5 studies that were graded as level three and level four tiers of evidence. Results of SSI rates found within this systematic review ranged from 0% to 9.4% of overall postoperative infections encountered after foot and ankle surgery in the studies analysed.

Conclusion

Whilst fragmented, aspects of antibiotic chemoprophylaxis are established fields in elective surgery with a growing body of evidence. Evidence for antibiotic use however, specifically within elective foot and ankle surgery, is lacking. This systematic review is a seminal paper which delivers an impression of the most influential literature within the field of foot and ankle surgery, with the aim being to entice conclusions and guide future research.

Background

Antimicrobial stewardship is central to attaining optimal clinical outcomes in surgery, diminishing adverse drug events, reducing cost burdens associated with Surgical Site Infection (SSI) and curtailing the pathogenesis of antimicrobial resistant strains. The vast expenditure and misapplication of antibiotic agents acts to hasten the advent of antibiotic resistant pathogens, rapidly becoming a global public health issue [1].

Patients infected with antibiotic resistant organisms are more likely to have lengthier, more costly inpatient stays and a greater frequency of morbidity and mortality [2,3,4]. It is documented that patients who do not receive appropriate antibiotic prophylaxis are 2.32 times more likely to acquire a SSI as compared to those who receive antibiotic chemoprophylaxis [5].

In orthopedic surgery, routine antibiotic prophylaxis is considered the standard of care, particularly within total hip and knee arthroplasty, trauma procedures and open fracture repair [6]. A Cochrane Review of SSI rate and antibiotic prophylaxis in bone & joint surgery, confirmed the effectiveness of prophylaxis in the prevention of both superficial and deep postoperative infection [7]. Data from 8447 participants in 23 studies evaluating closed fracture fixation, found a bolus dose of chemoprophylaxis significantly reduced SSI rate (risk ratio 0.40, 95% CI 0.24 to 0.67).

Comparatively, a recent review article examined the current evidence for antibiotic prophylaxis in hand surgery (where incision and fixation size may be comparable to the foot) with the recommendations that antibiotic prophylaxis should not be routinely used with patients without autoimmune disease, not taking steroid medication, in clean and hygienic patients or in procedures without hardware, of low haematoma risk and performed in an outpatient or ambulatory setting [8]. It has been suggested that in clean elective hand surgery, sterile prepping and draping technique is more pertinent to reducing the risk of postoperative infection and has fewer adverse effects than administration of antibiotic prophylaxis [9].

Nonetheless, antibiotic prophylaxis is adopted in clean elective hand surgery. An international survey revealed 49% of American surgeons and 13% of international surgeons gave antibiotics perioperatively for carpal tunnel release surgery, a procedure which does not require the use of fixation [10].

Within clean elective foot and ankle surgery the accepted rate of SSI quoted in the literature ranges between 0.26 to 2% [11] and the latest National Institute for Health and Clinical Excellence (NICE) guidelines for SSI state at least 5% of patients will suffer a postoperative infection following surgery [2]. The many heterogeneous studies with varying research methodologies, patient selection, procedure technique and local policy may account for the difference in SSI rate, though the importance of antibiotic choice and dosing may also play a significant role when equating these surgical outcomes [12].

The NICE Clinical Guideline 74 [2] states that for bone and joint surgery involving implantation of foreign material such as joint replacement prostheses and metalwork fixation, pharmacological choice for chemoprophylaxis should follow local policy, which are developed via national guidelines. This should be in line with patient ASA grade (American Society of Anaesthesiologists); current best practice; local resistance patterns; and type of surgery undertaken.

Comparably, the Society for Healthcare Epidemiology of America (SHEA), the Infectious Diseases Society of America (IDSA), the Surgical Infection Society (SIS), the American Society of Health-System Pharmacists (ASHP) advise several important factors to consider in SSI risk [3]:

• The health status of the patient, a ASA score of more than 2 denotes increased risk.

• Implantation of prosthetic fixation or device and case duration.

• Class of surgery; dirty contaminated or clean.

The NICE Clinical Guideline 74 [2] is based on a 1–10 panel rating system, conferred by medical professionals. In the process of development, proposals receiving a score of over 7.5, were published as central recommendations within the NICE guidelines. Eleven health professionals comprised the team of raters, however a surgeon undertaking any type bone and joint surgery was not represented on the panel of raters.

Nonetheless, the 2008 SIGN 104 Guidelines [4] and the 2013 ASHP report [3] advocate very similar practice as the NICE Clinical Guideline 74 [2] and have both critically appraised an abundance of literature through peer review in the formation of their respective recommendations. Methodical analyses of the available literature were utilised in order to apply findings to chemoprophylactic antibiotic use, in a wide variety of surgical specialties [3, 4].

Within foot and ankle surgery in the United Kingdom the PASCOM-10 database is used within the Podiatric Medicine profession to audit clinical performance, practice and outcomes [11]. Statistics reported from the database across 75 centres from 2010 to 2014 yielded an infection rate of 2.34% (suspected) and 0.67% (proven) in 11,673 patients undergoing scarf osteotomies with internal fixation. Interestingly, only 37.4% of patients were prophylaxed with antibiotics [13]. Based on the national database records therefore, it can be deduced that it is common practice in UK Podiatric Surgery to not use antibacterial prophylaxis in clean elective foot and ankle surgery where internal fixation is utilised, and/or where uninfected bone is involved [13]. These statistics remain consistent with what the national guidelines and literature advise as acceptable SSI rates and yet over 60% of patients in this cohort did not receive antibiotic prophylaxis.

The preponderance of literature concerning bone and joint surgery recounts procedures more proximal to the ankle and often with more substantial hardware, larger incisions and longer exposure times [2,3,4]. The SIGN and NICE guidelines particularly have not referenced any studies pertaining to surgery of the foot and ankle in their recommendations [2, 3]. Therefore, these recommendations are based on heterogeneous data that is extrapolated from other surgical procedures unrelated to the foot and ankle and as such the perceived strength of these recommendations is low.

It can be concluded that there is a necessity for new high level data related to antibiotic prophylaxis in foot and ankle surgery. A pre-requisite being analysis of the existing data, of which an exhaustive systematic review does not exist to date. Thus serving to increase not only the amount but also quality of available evidence on which to make practice recommendations with greater confidence [2,3,4].

Methodology

To undertake this review a systematic literature search utilising the following electronic bibliographic databases was performed: MEDLINE (Medical Literature Analysis and Retrieval Online, Bethesda, MD), CINHAL (Cumulative index to Nursing and Allied Health Literature, Ipswich, MA) and EMBASE (ExerptaMedica Database, Amsterdam, Netherlands) the Cochrane library without date restriction up to 1st March 2018. To avoid omitting any studies on the subject, Google Scholar search engine was also used, as well as a hand search of all references listed at the end of each of the selected studies for review.

The keywords and medical subject heading (Mesh) terms used are shown in Table 1 with the Boolean operators being AND or OR.

Table 1 Search terms for formal literature search

The following (Table 2) represents the selection inclusion and exclusion criteria for the purposes of this systematic review, with careful thought to what variables would help to satiate the scarcities in the current guidelines outlined and how best those disputes raised could be tackled.

Table 2 Selection criteria for review

The preliminary search yielded a total of 135 journal articles conducted by the primary author (RKM) and repeated by a secondary author (MS). Publication dates ranged from 1990 to 2018. The two screeners (RKM & MS) then reviewed the titles, abstracts and full texts of relevant articles for their eligibility based on the defined inclusion and exclusion criteria. Any disputes or cases of ambiguity were settled by a third screener (JL). This process led to the exclusion of 96 studies which clearly did not meet the inclusion criteria of the review. From the remaining 41 studies, 30 were excluded after examination and assessment of eligibility, as they too did not meet the inclusion criteria as first thought. The primary reasons being, that the majority included hip or knee surgery within their reviews and were not exclusive to surgery of the foot and ankle. Nine studies were deemed to meet the eligibility criteria. The references of inclusive studies were screened and a further 2 full text articles were identified. Eleven studies were included in the final review, which were then reviewed using the PICO tool [14, 15] to extract the relevant information needed to answer the research question in a systematic fashion thus reducing error and bias [16]. All studies were then collated and classified, with recommendations then graded A to C based on the level of associated evidence formulated by two internationally accepted hierarchies of evidence (see Tables 3, 4) [2].

Table 3 Classification of evidence

Table 4 NICE hierarchy of evidence and recommendation grading scheme

The studies were further rated based on their scientific methodology and robustness using a Modified Coleman Methodology Scoring Tool on a scale of 0–100 [17] Fig. 1.

Fig. 1

– Flow Diagram of screening process

A quality assessment is vital within a systematic review in order to substantiate reliability and both internal/external validity of its findings and this tool is specifically designed to score a published study based on its methodological strengths [16]. All studies were automatically allocated the maximum score for follow-up due to the nature of the intervention/study outcomes, in line with NICE definitions of Surgical Site Infection [2]. This was deemed appropriate as with such antibiotic chemoprophylatic studies, there is no scientific and or clinical benefit for an extended long-term follow-up of more than 2 months, as an infection outside this date range would not be classed as an SSI [2].

Results

Overall 11 studies met the inclusion criteria. From the grading of evidence, 2 level one and 4 level two studies were recognised. These studies ranked relatively highly in comparison to 5 studies that were graded as level three and level four tiers of evidence [2] Table 5.

Table 5 Classification of evidence based on NICE protocol [2]

Further data analysis summarised in Tables 6, 7. was undertaken using the PICO tool [15], evidenced a trend toward heterogeneity amongst the studies meeting the inclusion criteria, meaning it was not entirely possible to undertake pooled statistical analyses of the overall findings

Table 6 Modified Coleman Methodology Scoring Tool analysis of studies meeting inclusion criteria further demonstrated that over half of the studies included in the review scored 70 or over

Table 7 PICO Analysis for data extraction of the studies meeting the inclusion criteria [15]

Infection incidence

Results of SSI rates found within this systematic review ranged from 0 to 9.4% of overall postoperative infections encountered after foot and ankle surgery in the studies analysed Table 8.

Table 8 Comparison of SSI rates within all studies utilised for this systematic review

A multicentre prospective cohort study of 4238 patients [18] demonstrated that without antibiotic prophylaxis infection rate was 2.6% which was 1.5% higher than the group receiving chemoprophylaxis which was 1.2% incidence. Interestingly, they also reviewed a separate cohort which received both pre-& postoperative antibiotics who demonstrated an SSI rate of 2.1%.

After collating the information upon those studies which compared chemoprophylaxis vs no chemoprophylaxis the total pooled data demonstrated postoperative infection rates of 1.2% (antibiotic chemoprophylaxis) and 2.3% (no antibiotic chemoprophylaxis) (Table 9).

Table 9 Overview of studies which compared SSI rates in cohorts which received and did not receive antibiotic prophylaxis

Incidence of adverse drug event

From analysis of all studies, there was no reported incidence of anaphylaxis, toxicity or any other adverse drug event and/or serious side effects related to the use of antibiotics in foot and ankle surgery Table 10.

Table 10 Pooled data from the 4 studies which drew causal relationships between antibiotic chemoprophylaxis and subsequent SSI rate

Pneumatic tourniquet application and antibiotics

Three studies evaluated antibiotic chemoprophylaxis and the influence of tourniquet application [19,20,21]. One RCT prospectively evaluated SSI rates and the delivery of antibiotics both 5 min prior to tourniquet inflation and 1 min after inflation [19]. This study reported infection rates of 14.8% (antibiotics before inflation) and 3.9% (antibiotics after inflation) respectively within the groups studied.

Two studies prospectively evaluated antibiotic bone penetration, using the minimum inhibitory concentrations (MIC90) as a quantitative marker to demonstrate effectiveness against 90% of the most common causative organisms found in vitro at the site of surgery [20, 21].

A prospective cohort study of 25 patients reported that IV administration of 1 g Cefazolin 60 min prior to tourniquet inflation led to mean bone concentration levels of 2.39 ± 1.19 μg/g (sd).

This translates to approximately 2 to 4 times the required amount to exert a bacteriostatic effect at the site of surgery (0.5–1.0 μg/g required MIC90) [20] raising the question of whether this standard dose of IV Cephalosporin is unnecessarily high and in the course of good antimicrobial stewardship whether there would be any benefit to reducing this dose where chemoprophylaxis is required Tables 11 and 12.

Table 11 Use of Antibiotics in studies meeting the inclusion criteria and their administration protocol where defined

Table 12 List of pathogens encountered in those patients who developed a postoperative SSI in the study conducted by Zgonis et al. [22]

Causative organisms

Zgonis et al. [22] examined bacterial isolates of patients with laboratory confirmed SSI following elective foot and ankle surgery, concluding that the most commonly encountered pathogens were coagulase negative and positive Staphylococcus. Interestingly, all patients with preoperative chemoprophylaxis and postoperative wound infection developed a pathogen unfailingly resistant to Penicillin and/or Ampicillin. These findings are consistent with similar studies which have noted a proliferation in difficult to treat bacterial isolates of Methicillin-esistant Staphylococcus aureus and epidermidis (MRSA/MRSE) and Vancomycin resistant Staphylococcus aureus and Enterococcus (VRSA/VRSE) in clean elective surgery [23] [24,25,26]. Thus validating the thought that unremitting overuse of the same prophylactic antibiotic(s) will result in the evolution of tremendously virulent pathogenic isolates, which will arduously be a challenge to treat postoperatively [18, 22, 27].

Duration and type of foot and ankle surgery

Zgonis et al. demonstrated a statistically significant association (P < 0.001) between antibiotic use and surgical category and fixation choice [22]. However, as patients were not randomised prospectively owing to the nature of the study, patients undergoing more complex surgery with higher amounts of both internal and external fixation were more likely to receive antibiotics due to surgeon preference.

This could therefore explain the difference in infection rates between the 2 groups (antibiotic vs no antibiotic) as a longer duration of surgery with higher amount of implanted fixation may indeed lead to a higher rate of infection and it would be these patients who received preoperative antibiotics, had they not been prophylaxed it could be argued that the infection rate could have been higher in this group.

Risk & predictive factors for SSI

No studies appraised type and duration of foot and ankle surgery as independent variables influencing SSI rates, with the exception of Tantigate el al. postulating via the findings in their study, that independent predictors of postoperative infection were ASA Grade, non-ambulatory surgery and a longer duration of surgery, with almost 92% of the risk of a postoperative infection being predicted by these factors (P > 0.05) alone [28].

Summary of studies

The characteristics of the research represented seems to be split into a mixture of largely retrospective or prospective cohort designs, with more recent studies adopting a prospective approach and recognising the need for better quality research into the area. All studies reviewed were either published in the 1990s or post 2010, indicating a lag-period of nearly 20 years where the topic was not well researched. These studies were international and no one country seems to be leading upon this research field.

Discussion

Infection incidence

The results of the systematic review demonstrate inconsistent postoperative infection rates, ranging from 0 to 9.4% of overall SSI’s encountered after foot and ankle surgery.

The highest-level study enrolled 100 patients [27]. With this type of prospective RCT a sample size between 425 and 1145 could be used to determine a reduction of infection from 4 to 1% and 4% to 2% assuming 80% power and statistical significance of 5%, although it may be possible that even with a larger sample size the results may not differ significantly. However, this study was unique in that it exclusively reviewed lesser toe surgery and not other types of foot and ankle surgery; moreover, with the use of percutaneous k-wires in all cases, the application or generalisability of these findings to studies which do not utilise external fixation methods may be limited.

A multicentre prospective cohort study of 4238 patients [18] demonstrated significantly higher infection rates when patients received both pre and postoperative antibiotics in comparison to those who received preoperative antibiotics alone, concluding that postoperative antibiotics are counter-productive for clean elective foot and ankle surgery.

Influence of pneumatic tourniquet application

Pragmatically, in the manifestation of transitory surgical limb ischaemia, the MIC90 of an antibiotic may diminish at the operative site, a hypothesis which would suggest that prophylaxis would then be less effectual. There is however, deficient evidence to sustain this rationale within the extant latitude of literature.

A prospective RCT of 106 consecutive patients undergoing ‘clean, elective, foot and ankle surgery’ demonstrated SSI rates of 14.8% (Group 1) and 3.9% (Group 2) in 2 patient cohorts, both of whom received IV antibiotics. Group 1 received IV Cefuroxime 5 min prior to tourniquet inflation and Group 2 were administered chemoprophylaxis 1 min after inflation in an attempt to discern thigh tourniquet influence upon postoperative infection rate [19]. The researchers, however, admitted that over 74% of procedures were trauma cases e.g. open fractures requiring ORIF and up to a 5 day inpatient stay as standard practice at this medical facility. This would suggest that the internal validity of this RCT is compromised as it is no longer evaluating infection rates within ‘clean, elective, foot and ankle surgery’ as the study had outlined in its design and aims.

Antibiotic regimen

All studies within this systematic review which described their antibiotic regimen utilised exclusively IV antibiotics, with the majority of studies utilising Cephalosporin antibiotics [19,20,21, 28].

However, there has been a national drive within the UK particularly, to move away from this type of antibiotic for the use of surgical prophylaxis not only to mitigate the critical risk of bacterial resistance but also to subdue the rates of Clostridium difficile (C.diff) infections in the hospital environment [29, 30]. According to the NICE guidelines (2015) the results of 3 meta-analyses established that the most prominent antibiotics associated with nosocomial C.diff infection were second and third generation Cephalosporin’s, Clindamycin and the Fluoroquinolone group [30]. This makes the development of local guidelines challenging due to the existing body of evidence being largely based on the use of a class of antibiotics which are believed to be broad not narrow spectrum drugs.

Moreover, the Cephalosporin group has been shown to be associated with higher rates of C.diff infection and are consequently being discouraged as a primary agent for surgical prophylaxis in clean elective surgery by some health authorities, therefore going against the national UK guidelines it seems on not one - but two - fronts [4, 29, 30]. It would seem pertinent in light of this issue to undertake further prospective studies evaluating the use of narrow spectrum antibiotics which are not associated with the reported higher rates of C.diff infection.

Oral vs. IV

The most prominent authority guidelines on the use of antibacterial chemoprophylaxis counsel on the use of the IV [2,3,4]. Oral antibiotic drugs must first be metabolised via the bodies first pass mechanism, increasing the length of time taken to reach MIC90, which should be reached by the first incision in order for chemoprophylaxis to be most effective [3, 20]. In the case of oral administration of antimicrobials, there is some high level evidence to suggest oral antibiotic prophylaxis is as effective as IV prophylaxis [31], though no studies specific to the foot and ankle have been published to date. It is suggested that antibiotic chemoprophylaxis take place within 60 min prior to tourniquet inflation [32, 33]. The significant lack of relevant research into oral antibiotic prophylaxis is noted and once again, prospective study of the RCT design is warranted in order to validate the widely held hypothesis that IV administration of antibiotics for surgical prophylaxis is more effective than oral or intramuscular (IM) routes [2,3,4].

Bolus vs. multiple doses

Evidence yielded from this systematic review further proposes that a solitary dose of prophylactic antibiotics, with a suitable half-life, is appropriate in diminishing postoperative infection rate [2]. Re-dosing is recommended in surgical procedures lasting more than 2 half-lives of the antibiotic used for surgical prophylaxis with consideration to blood loss (> 1.5 L), fluid replacement and patient factors e.g. both liver and renal function [2, 3]. It is noted however that from conducting this review no studies in foot and ankle surgery have independently reviewed the influence of re-dosing at various time intervals intra-operatively and ensuing outcomes related to SSI.

Narrow Spectrum antibiosis

It should be noted that Staphylococcus aureus and Staphylococcus epidermidis are common sources of infection in foot and ankle surgery and are particularly associated with infections of foreign material due to their propensity to form biofilms [2,3,4, 22].

The choice of suitable antibiosis is of paramount importance to safeguard absolute efficiency while curtailing the risk of adverse drug events, which was interestingly not reported in any of the studies in this systematic review. If antibacterial prophylaxis is used, the agent(s) chosen should be narrow spectrum and avoid any unnecessary broad-spectrum activity [2, 4].

Conclusion

Summary of recommendations

• In clean, uncomplicated and elective soft tissue surgery of the foot and ankle antibiotic prophylaxis is not routinely warranted [2,3,4, 18,19,20,21,22, 27, 28, 34,35,36,37].

  Grade of recommendation: B.

• In clean, uncomplicated and elective bone and joint surgery of the foot and ankle antibiotic prophylaxis is not routinely warranted [2,3,4, 18,19,20,21,22, 27, 28, 34,35,36,37].

  Grade of recommendation: B.

• Where the use of metallic hardware is necessitated, the decision to provide prophylaxis must be agreed locally as a conclusive recommendation on whether or not to chemoprophylax cannot be made due to the paucity of the literature in favour of chemoprophylaxis within foot and ankle surgery. With respect to this, the decision should ultimately be based upon surgeon preference, patient and surgical risk factors with pertinence to the amount and type of hardware and surgery required. It is the personal view of the author that until better research is established national guidelines be followed and chemoprophylaxis in this situation be provided [2,3,4, 18,19,20,21,22, 27, 28, 34,35,36,37]

  Grade of recommendation: C.

• Where the use of joint replacement prostheses are necessitated, antibiotic prophylaxis is recommended due to their propensity to form biofilms [2,3,4, 18,19,20,21,22, 27, 28, 34,35,36,37].

  Grade of recommendation: A.

• Antibiotic prophylaxis should be delivered to the patient within 60 min of inflation of pneumatic tourniquet to allow sufficient plasma concentrations of the drug to be utilised [2,3,4, 18,19,20,21,22, 27, 28, 34,35,36,37].

  Grade of recommendation: A.

• There is insufficient current evidence to suggest that IV antibiotics are more effective than oral or IM with respect to chemoprophylaxis in foot and ankle surgery, it is therefore advised that local policy considers drug class, local resistance patterns, drug penetration and bioavailability as well as being administered within a sufficient time frame to allow for adequate serum tissue concentrations to be reached (usually within 60 min) [18,19,20,21,22, 27, 28, 34,35,36,37].

  Grade of recommendation: C.

Strengths & Limitations

The strengths of this systematic review are the clear definition of the research question, which eliminates bias in the selection of the studies. Furthermore, the strengths of our study are the stringent adherence to an unambiguous research protocol that was developed prior to the analysis. As well as this, the broad nature of the literature search and consensus between the two reviewers throughout the screening process add validity and reliability to the research findings.

Nevertheless, despite the strengths of the review process, the authors concede that the absence of high level studies within the literature is apparent. The primary limitation of this review is that a direct comparative meta-analysis of infection rates within foot and ankle surgery, with and without antibiotic prophylaxis was not possible because there are few head-to-head trials.

Additionally, the variability of the outcomes measures, limited the number of studies which could be directly compared. Indeed, dissimilarities in patient cohorts, surgical procedures and variances in outcome assessment tools will be in some measure responsible for heterogeneity among these studies and prospective researchers should aim to address these issues with high level prospective studies.

Recommendations for further research

There is a need for more prospective randomised studies within the field of foot and ankle surgery where drug route (IV, IM and oral) is utilised as an independent variable, currently no such study exists to inform the decision as to which route would be superior for foot and ankle surgery and we are therefore reliant on studies specific to more proximal (to the ankle) surgery.

Moreover, prospective randomised studies must also evaluate the influence of antibiotic chemoprophylaxis upon SSI rate in clean elective foot and ankle surgery where metallic hardware is implanted as internal fixation, as the available studies, national guidelines and UK national audit data all appear to represent a conflicting representation of views and warrant review.