The evidence base for 2017 BOAST-4 guidance on open fracture management: Are we due an update?

Abstract

Introduction

Open fracture management in the United Kingdom and several other countries is guided by the British Orthopaedic Association’s Standards for Trauma Number 4 (BOAST-4). This is updated periodically and is based on the best available evidence at the time. The aim of this study is to evaluate the evidence base forming this guidance and to highlight new developments since the last version in 2017.

Methods

Searches have been performed using the PubMed, Embase and Medline databases for time periods a) before December 31, 2017 and from 01/01/2018–01/02/2021. Results have been summarized and discussed.

Discussion

Several contentious issues remain within the 2017 guideline. Antibiotic guidance, the use of antibiotic impregnated PMMA beads and intramedullary devices, irrigation in the emergency department, time to theatre and the use of negative pressure dressings and guidance regarding the management of paediatric injuries have all demonstrated no clear consensus.

Conclusion

The advent of the BOAST-4 guideline has been of huge benefit, however the refinement and improvement of this work remains ongoing. There remains a need for further study into these contentious issues previously listed.

1. Introduction

Management of open fractures in the United Kingdom is guided by the British Orthopaedic Association’s Standard for Trauma number 4 (BOAST-4). The current BOAST-4 guidelines for treatment of open fractures were published in 2017, as an update to the 2009 and preceding 1997 versions.¹ These guidelines are periodically updated, as their evidence base is constantly evolving, and naturally there are always areas that are derived from stronger or weaker evidence. We aim to summarise the evidence upon which the current BOAST-4 antibiotic guidance is based, and highlight some of the new research that has been presented. By summarizing the evidence pre and post 2017, and highlighting some of the more contentious aspects, we hope to provide direction for future research.

2. Methodology

The 19 recommendations within the BOAST 4 guideline were used as the basis for literature searches, which were performed using PubMed, Embase and the Cochrane database. Evidence was categorized into evidence before or after December 31, 2017, critically evaluated using the CASP tool, tabulated and evaluated systematically. Contentious areas within the literature were identified and have been summarized.

3. Discussion

There are several areas of contention within the BOAST guidance. We will give direct quotes from the guideline, and then discuss key papers relating to this topic.

3.1. Antibiotic timing and delivery method

Patzakis et al. (1974)² conducted the first landmark prospective randomized controlled trial exploring the use of antibiotics in open fractures. This 310 patient RCT demonstrated that antibiotic usage significantly reduced the infection rate from 13.9% to 2.4%. Patzakis took this work further in 1989 by exploring other factors influencing the infection rate in 1104 open fractures (77 infections), which found that the timing of antibiotic therapy was the most important factor in preventing infection. Antibiotic delivery within 3 h conferred a 4.7% infection rate, vs 7.4% infection rate when given after 3 h.³ This was later reinforced by several publications from other groups.^4,5

Gosselin et al. (2004) performed a Cochrane systematic review of RCTs, including 913 adult and paediatric open fractures, aiming to quantify the benefit of antibiotics in this context. Gosselin demonstrated significant benefit with antibiotic usage (reduction in relative and absolute risk reductions with 95% confidence intervals, and a number needed to treat of 13). Gosselin concluded that prompt antibiotic prophylaxis plays such a conclusively important role when combined with surgical management, that any further placebo studies would be unjustifiable.⁴

Jaeger et al. (2006) broadened the scope of their enquiry; making recommendations based on various clinical scenarios including closed, open and periprosthetic fractures, and also those with established infections.⁵

Lack et al. felt that deep infections were the most relevant outcome measure and was a more relevant focus of enquiry, rather than superficial infections.⁶ Lack performed a retrospective observational study of 137 patients with Gustilo-Anderson III fractures and reported an infection rate of 6.8% (antibiotics within 1 h) vs 27.9% (>90 min).

While there have been some studies that showed no impact of early antibiotic administration on infection rate,7, 8, 9 these studies had several shortcomings and have largely been disregarded in the abundance of robust evidence to the contrary.¹⁰

There has been a renewed interest in local delivery of antibiotics to the wound, allowing greater concentrations of antibiotic to be given.^11,12 Antibiotic impregnated devices (eg PMMA beads, cement and implants have been shown to be effective at treating infected non-union.¹³ Morgenstern et al. (2018)¹⁴ performed a systematic review and meta-analysis investigating PMMA bead usage in upper limb open fractures, demonstrated a significantly reduced infection rate (4.6% vs 16.5%, p < 0.001). The antibiotic coating of choice has been variable in studies (fluoroquinolones, vancomycin, tobramycin, aminoglycosides, beta-lactams) and there have been very few studies looking at local antibiotic delivery.¹⁵

Pinto et al. (2019)¹⁶ performed a randomized controlled study of 28 open tibial fractures (Gustilo-Anderson I and II), finding that antibiotic coated nails had a statistically significant reduction in infection rate and improved time to union. Villani et al. (2020)¹⁷ performed a systematic review investigating the use of Gentamicin coated tibial nails, including 114 open fractures, finding that these devices were particularly useful in comorbid patients.

3.2. Antibiotic guidelines

“There should be a readily accessible published network guideline for the use of antibiotics in open fractures.”

2017 BOAST-4

Previous versions of the BOAST-4 guidance have specified first and second line antibiotics choices by name, however the 2017 version instead defers to local hospital guidelines.

Choice of antibiotic therapy has been the subject of several Level 1 studies, comparing traditional 1st generation cephalosporin and aminoglycoside with other classes of antibiotics (3rd generation cephalosporin,^18,19 Piperacillin/Tazobactam,²⁰ Vancomycin and Cefepime^21,22), when treating Gustilo-Anderson type III fractures. There was no role found yet in using Vancomycin prophylactically.²³ These studies have found no statistically significant differences in outcomes between these cohorts. This does suggest that the more dangerous nephrotoxic risks of 1st generation cephalosporins and aminoglycosides could be avoided with equally efficacious antibiotic options. There does appear to be some rationale for choosing one equally efficacious antibiotic over another.

Duration of antibiotic therapy is not addressed in the BOAST guidance specifically. The Eastern Association for the Surgery of Trauma (EAST) provides recommendations on when to stop antibiotic therapy,^24,25 are timing of soft tissue closure and the Gustilo-Anderson grade. This is a commendable effort, and it may be worth BOAST incorporating similar advice.

Chen et al. (2013) found that 10% of open fractures became colonised, and of those 10%; MRSA was present in 25% and gram negative bacteria were present in 55%. The group therefore recommended revising current antibiotic regimens. This may be a little hasty as the study only included 202 patients, further study is indicated.²⁶

Deferring to local guidance has the advantage of giving organisations the freedom to choose antimicrobial agents depending on local common pathogens, and bulk purchasing options. This however, does leave the guidance vulnerable to being affected by factors not specifically pertaining to patient outcome.

3.3. Irrigation in the emergency department

“Prior to formal debridement the wound should be handled only to remove gross contamination and to allow photography, then dressed with a saline-soaked gauze and covered with an occlusive film. ‘Mini-washouts’ outside the operating theatre environment are not indicated.”

2017 BOAST-4

Tscherne et al. (1984) provided the basis for the hyperacute management of open fractures and recommended removal of gross contamination where possible.²⁷ Aside from several other major contributions to modern day open fracture management, Gustilo et al. (1990) also identified several specific forms of gross contamination (as above) that warrant urgent debridement.²⁸

Bhandari et al. (2016) investigated optimal irrigation pressure and fluid type to be used.²⁹ This study was not proposing irrigation in ED but rather in theatre. Bhandari did conclude that there was no difference between high or low pressure irrigation and that normal saline gave better results than soap solution. With that said, it would stand to reason therefore that normal saline at low pressure would be a sensible initial intervention in ED also, even if the process will be repeated in theatre.³⁰

Preliminary irrigation in the emergency department has been proposed before, however opinion remains mixed; with some studies advocating this practice and others suggesting it may do more harm than good, by pushing infection deeper into the wound. No further studies relating to this topic have been identified since 2017.³¹

3.4. Timing of surgery and expertise

Previously debridement had been performed by orthopaedic surgeons, often in small hospitals without a plastic surgery service. Later, the plastic surgeons would be expected to provide coverage. This disjointed effort and lack of communication anecdotally made coverage for the plastic surgeons unnecessarily difficult at times, conferring worse outcomes for patients. Through partnership between the BOA (British Orthopaedic Association) and BAPRAS (British Association of Plastic, Reconstructive and Aesthetic Surgeons), this problem was identified and a partnered approach was suggested. Both teams must be sympathetic and mindful of the other’s priorities and work side by side with active dialogue starting well before the patient is on the operating table.

The 2009 BOAST-4 guidance stipulated that in the absence of severe contamination or neurovascular compromise, it would be best to perform primary debridement on a planned daytime theatre list with consultant orthoplastic surgeons input. The rationale in 2009 being that the traumatic insult has occurred, and that performing such an operation overnight may lead to a less judicious primary debridement.

The multidisciplinary orthoplastic approach to open injuries is arguably one of the most impactful changes brought forward by the 2009 BOAST-4 guidance. The introduction of BOAST-4 has been shown to improve outcomes,^30,32 and this novel MDT and trauma system based approach is thought to be partly responsible.33, 34, 35, 36 The argument for aiming to provide treatment as soon as possible by an expert team during day-time hours, rather than be hampered by a poor first attempt at night by an inexperienced orthopaedic surgeon alone.³⁰

2017 BOAST-4 guidance however takes a step away from this, by setting a new 12 h target for high energy injuries. This may be well-intentioned, however it may inadvertently lead to a step backwards in this respect. Further clarification may be helpful in the guideline.

Upper limb injuries may present an opportunity to delay theatre until a planned list. Warrender et al. (2018)³⁷ conducted a review of 27 studies exploring the effect of delayed antibiotic therapy and operative debridement in upper limb injuries. However, this study did include hand and wrist trauma (which are not included in BOAST-4 guidance. Warrender suggested that for upper extremity open fractures, timing of debridement was much less important than timely administration of antibiotics.

3.5. Negative pressure dressing usage

“Definitive soft tissue closure or coverage should be achieved within 72 hours of injury if it cannot be performed at the time of debridement.”

2017 BOAST-4

The 2009 BOAST-4 guidance gave specific advice in the event of multi-stage or delayed closure procedures, advocating the use of a vacuum foam dressing or antibiotic bead pouch until definitive surgery can be performed.

Historically early primary closure was discouraged due to fears of encouraging clostridial colonisation and gas gangrene, advising that closure should be intentionally delayed.³⁸ Several subsequent studies found that open fractures are often colonised by MRSA (23%) and gram negative bacteria (31%) and crucially that early primary closure was helpful in preventing such infections.³⁹

Cross et al. (2008) performed a literature review of management of open fractures, and specifically addressed time to wound closure.⁴⁰ The team suggested that closure should be achieved by 72 h, but that if adequate decontamination is in doubt, then coverage should be postponed, with vacuum assisted dressings and/or antibiotic bead pouches considered in the interim.

Negative pressure wound therapy (NPWT) is identified by BOAST-4 as a temporary adjuvant in scenarios where delayed wound closure is unavoidable, but does not mitigate any further delay in wound closure.⁴¹

NPWT systems subsequently increased in popularity,⁴² and are well-established practice in other contexts eg. skin grafts. However, it has been argued that if residual debris is still within the wound, that negative pressure could trap these between potential spaces and tissue planes.⁴³ 2009 BOAST-4 advice against NPWT usage before initial debridement, washout and haemostasis was seemingly quite sensible. Removal of this seems to have been a step backwards rather than forwards.

Since 2017, several Level 1 studies comparing NPWT to conventional dressings, but have produced conflicting results. Costa et al. (2018)⁴⁴ conducted a multi-centre RCT comparing NPWT with conventional dressings in open lower limb fractures and found that there was no clinical or economic benefit in using NPWT at 12 month follow up. However, Kim et al. (2018)⁴⁵ analysed 12 RCTs (including the same RCT by Costa et al.) and gave starkly contradictory results; finding that NPWT resulted in lower rates of soft tissue infection, non-union, flap necrosis and revision, when compared with conventional wound management in open tibial fractures. Liu et al. (2019)⁴⁶ also performed a similar systematic review including 909 patients from 8 RCTs and 6 cohort studies, and similarly concluded that NPWT resulted in a significantly reduced infection rate, shorter wound coverage time, healing time, hospital length of stay and lower amputation rate when compared to conventional dressings. Whilst these reviews suggest NPWT is favourable to standard dressings, another recent multicentre RCT by Tahir et al. (2020)⁴⁷ analysing open tibial fracture wound management in 486 patients found no benefit of NPWT over standard dressings.

3.6. Paediatric open fractures

2009 BOAST-4 guidance included a statement that “the wound in open tibial fractures in children is treated in the same way as adults”. This statement has been removed in the 2017 revision, and there is no mention of how to manage paediatric open fractures. This seems to be a step in the wrong direction.

Iobst et al. (2005) investigated 40 paediatric Gustilo-Anderson (GA) grade I fractures, finding that treatment as an adult was not indicated. These patient were managed non-operatively as follows:

• -Early administration of antibiotics
• -Superficial wash/removal of gross debris in the Emergency Department
• -Simple dressing with saline and betadine soaked gauze
• -Closed reduction and plaster
• -Further 24–48 h of IV antibiotics

Iobst reported 2.5% incidence of infection (in a 12 year old with a tibial fracture) and recommended that GA I fractures should be treated non-operatively in patients under 12 years old.⁴⁸

Iobst again in 2014, conducted a similar trial, but this time with several small differences.

• -IV antibiotics were administered immediately, but otherwise received the same treatment in the Emergency Department.
• -Patients were admitted for 3 further doses of IV antibiotics over the following 24 h and then discharged home with an oral course to complete.

This conferred a 0% infection rate and so reinforces the proposal that paediatric cases do not require management as adults.⁴⁹

Baldwin et al. (2009) performed a systematic review investigating the union time and infection risk for different GA grades and made two main findings⁵⁰:

• -No observable difference in infection between GA I or II fractures, but up to a 3.5 fold increase in the chance of infection in GA III fractures.
• -Union time was strongly associated with increasing GA grade.

Eastwood et al. (2020)⁵¹ performed a systematic review and meta-analysis including 1093 patients exploring nonoperative vs operative treatment of GA III fractures in children. This study concluded that operative management of GA III fractures in a paediatric population could be managed non-operatively, however this work has come under criticism more recently.⁵² In the absence of robust prospective trials scrutinizing this point specifically, there unfortunately remains much variability in management strategy and no consensus on the management of GA III fractures in a paediatric population. NICE guidelines are available, but none from BOAST regrettably.

3.7. Recommendations

Specific guidance regarding antibiotic choice would be helpful in expediting and simplifying care in acute situations. Previous guidance has been very clear on this topic and perhaps this may have been a better approach.

In cases where soft tissue coverage is being delayed, routine use of antibiotic impregnated implants and vacuum dressing do seem to be a sensible and pragmatic way of optimising our management of open fractures. This is a promising area for further research and cost-benefit analysis to determine their utility.

Use of antibiotic coated intramedullary nails does seem to confer benefit, however further study may be required to determine if this is a cost effective method of treating such a common injury and whether there are specific patient cohorts who may benefit more from that added cost (eg comorbid patients.

An area for discussion may be to advise swabbing of the wound for microbiology, culture and sensitivity. This should be done before antibiotic administration but under no circumstances should antibiotic delivery be delayed. This is a suggestion that has previously been made, but with several shortcomings identified; including low specificity, the inadvertent culture of skin flora and the risk of further contamination.³

Timing of debridement has always been a contentious issue, and there are no new RCTs that have addressed this issue since the last update. What is clear however, is that timing of debridement is less important than timely administration of antibiotics in management of upper extremity open fractures. We would propose the idea that clarifying the benefits of consultant led management on a planned list should be the preferred option, where possible.

Wound management options in the context of delayed wound closure is still clearly a contentious issue. There is a need for further RCTs exploring short-term as well as long term outcome measures. If there is found to be no discernible difference, then this expensive therapy may be obsolete. At present, there appears to be no harm that comes from the use of NPWT dressings, and may be favourable.

Since 2009, there have been several studies exploring open fracture management of the paediatric population. These studies demonstrate that paediatric fractures are an entirely different pathology to deal with, and in many cases may require significantly less intervention to achieve a satisfactory outcome. Piecing together findings from landmark studies by Iobst, Baldwin and Eastwood in particular, paint a picture that GA I and II patients are low risk for infection and GA III patients require further investigation before any determination can be made regarding the indication for surgical treatment. It is our opinion that a pragmatic way forward for now would be:

• -GA I and II fractures should be treated as per the Iobst 2014 protocol, considering there is no increased infection risk between GA I & II. If there is a requirement for indwelling metalwork to achieve a satisfactory fixation then debridement and washout should be performed as per adult treatment.
• -GA III fractures should be treated as adults, with special attention paid to the fact that growth plates are still present in this population.

4. Conclusion

The advent of BOAST-4 guidance was a bounding leap forward in the management of open fractures and has been well demonstrated in the literature. Subsequent iterations in 2009 and 2017 continually provide incremental improvements upon this work, based on the best available evidence at the time. The guidance has a strong evidence base and has provided a significant stepping stone towards improving care, as demonstrated by the improved outcomes after its implementation.

The process of improving these guidelines will inevitably remain a continuous one and there are some aspects of care that remain pragmatic in nature, rather than strictly evidence based. The advent of the TARN database and improved trauma data collection and research will aim to eventually and gradually fill in these gaps in our knowledge.

The 2017 BOAST-4 version has certainly improved in several arenas, however there are some aspects that one could argue are a step backwards. Guidance could be improved in terms of specific antibiotic choice, timing and delivery, as well as timing of surgery, use of vacuum dressings and antibiotic coated implants in delayed closure cases.

Research since 2009 has shown that there is room for improvement, most notably in the treatment of paediatric cases, which are predominantly Gustilo Anderson Grade I and therefore best treated non operatively. The 2017 BOAST-4 version unfortunately does not address this and a paediatric protocol would be a significant area for improvement.

Declaration of competing interest

None to declare. One COI form has been submitted per author.