Improved management of open fractures is essential to reduce FRIs. The incidence of FRI following an open fracture is high in all settings, up to 52 % in some case series (Natoli et al., 2025; Schade et al., 2020). There are elements of open-fracture management that may be particularly important, such as early treatment in an appropriate centre (Kock et al., 2025) and engagement with community leaders and traditional bone setters (TBSs) may need to be prioritized to reduce delays.
Surgery should prioritize soft-tissue-friendly approaches, which decrease infection risk, as well as implementation of WHO recommendations such as the use of the surgical safety checklists, which has been shown to significantly reduce surgical-site infection (SSI) rates (Haynes et al., 2009).
FRI is an infection arising at any time following any fracture. Following the publication of the consensus papers (Metsemakers et al., 2018), the term FRI supersedes terms such as post-traumatic osteomyelitis/osteitis, fracture-fixation-associated infection (FFAI), and SSI. An SSI following a fracture is considered to be an FRI as it is impossible to differentiate between deep and superficial infections after fractures. FRI should be used as the diagnostic term to bring clarity to the clinical entity, facilitate use of appropriate treatment principles, and allow for robust research. Fractures occurring through sites of chronic haematogenous osteomyelitis are relatively common, and their management follows similar principles to those applied in FRI.
FRI should be suspected in all fracture cases, at any time point, where there is delayed bone healing following open fractures, surgically treated closed fractures, or fractures initially treated by TBSs. Patients and carers should be informed about the symptoms and signs of FRI to facilitate early review. A documented history and clinical examination should focus on the confirmatory and suggestive criteria of FRI (Govaert et al., 2020) to make the diagnosis (listed in Table 1). With late presentations and under-treatment of early FRI, clinical confirmatory criteria of sinus tracts and purulence may be particularly relevant in low-resource settings (Fonkoue et al., 2024). Any wound breakdown or dehiscence should be considered to be an FRI. In the absence of clinical confirmatory criteria, in any case of non-union or delayed union, or in the presence of suggestive criteria, FRI should be suspected and investigated accordingly.
Clinical symptoms and signs of fracture-related infection.
Blood tests, microbiological analysis, histopathology, and imaging investigations aid in the diagnosis of FRI and support management. Sinus tract samples or wound swabs for microbiological culture are not recommended as they are likely to grow skin commensals that may not reflect the causative pathogen. Intra-operative sampling is the microbiological diagnostic test of choice.
Table 2 summarizes key investigations. Blood tests help assess the general health of the patient and highlight areas that can be optimized. Serum inflammatory markers have limited use in the diagnosis of FRI (Trenkwalder et al., 2024).
Summary of diagnostic work-up for fracture-related infection.
Routine X-rays are advocated for in all cases of suspected infection and should be prioritized over more complex and costly imaging modalities (Govaert et al., 2020). Serial radiographs should be used to assess the following: fracture union (or potential for union), signs associated with possible infection such as the presence of necrotic or ischemic bone, loosening of implants, or subtle lucent lines around bone fragments or implants. Where available, CT, MRI, and nuclear imaging may be helpful in diagnosis and to guide surgical management, but they are not essential (Govaert et al., 2020). CT scans are useful to assess fracture configuration, bone healing, and the presence of sequestra to plan debridement. MRI is helpful to delineate the pattern of any sinus tracts and extent of soft tissue involvement. 18FDG-PET-CT has high sensitivity for diagnosing FRI, particularly in chronic and complex cases without the presence of confirmatory criteria, and may be helpful in differentiating between localized and segmental bone involvement (Lemans et al., 2019).
The Cierny and Mader classification (Cierny and Mader, 1984) and BACH classification (Hotchen et al., 2019) for osteomyelitis have historically been helpful in guiding the management of bone infection and are validated with regard to outcome. The new FRI classification (Alt et al., 2024) (see Fig. 1) is yet to be fully implemented.
Fracture-related infection classification. Used with permission from © Alt et al. (2024).
There should be a low threshold for instituting prompt appropriate treatment in any case of suspected FRI (Metsemakers et al., 2020). Patients with signs of systemic sepsis require immediate assessment and urgent treatment. For suspected FRI that presents early, prompt intervention is important to ensure good outcomes (Marais et al., 2024a). The treatment goals are fracture consolidation, eradication of infection (or suppression of infection until fracture consolidation is achieved), healing of the soft tissue envelope, restoration of function, and prevention of residual chronic infection.
Factors such as low vitamin D, anaemia, thyroid dysfunction, diabetes mellitus, malnutrition, liver function, and renal function should be addressed. HIV management should be optimized (Nieuwoudt et al., 2020). Smoking cessation, alcohol intake reduction and stopping illicit drug use or uncontrolled traditional treatment should be encouraged but should not necessarily preclude treatment. Patients should be involved in the decision-making process, and patients, their family members, and caregivers benefit from being well informed. FRI has a significant associated psychological burden, and patients benefit from a supportive treatment environment and expert input when necessary.
FRIs are complex and require input from different areas of expertise; often best delivered in the setting of an multidisciplinary team (MDT). The composition of the MDT should be adapted to local resources and expertise; many specialities are available in larger urban centres, and an MDT should, ideally, be set up in facilities managing FRI, as is currently the case for cancers. Studies in high-resource settings have demonstrated better outcomes with MDTs and centralization of services (Ferguson et al., 2021), but further work is needed to demonstrate how this might be applied in low-resource settings.
Antibiotic treatment should not be started before a diagnostic workup (unless the patient is systemically ill) and should be guided by local prevalence patterns and individual risk factors. Generic broad-spectrum cover should be provided in the peri-operative period and should be rationalized when microscopy, culture, and sensitivity (MC&S) results from intra-operative samples are available. Immediately after surgical sampling, generic intravenous broad-spectrum antibiotics should be given based on local microbial susceptibilities following an agreed-upon protocol and should be continued until definitive culture results are available.
In high-resource settings,
For empirical antibiotic treatment (EAT), the association of glycopeptide and an agent against Gram-negative bacteria is recommended (for example: vancomycin and amikacin or meropenem; if these are not available, cefazolin plus gentamicin may be a reasonable alternative) (Fonkoue et al., 2024). Definitive antimicrobial therapy should be culture specific. Drug choice should ideally be made with microbiology and/or infectious-disease expertise and follow principles of good antimicrobial stewardship. In culture-negative FRI, an agreed-upon policy should define the empiric antimicrobial choice.
The choice of antibiotic agent is dependent on sensitivity patterns, bone penetration, oral bioavailability, side effect profile, and biofilm activity. Bacteriocidal agents, with high oral bioavailability and good bone penetration, are generally preferred: rifampicin, fluoroquinolones, clindamycin, co-trimoxazole, tetracyclines (doxycycline or minocyclin), lincomycin, and fusidic acid. Cephalosporins are not recommended as they have lower oral bioavailability. In cases of implant retention, antibiotics with sufficient anti-biofilm activity are recommended. The association of rifampicin (anti-Staphylococcal biofilm activity)
Oral antibiotic therapy was shown to be non-inferior in relation to intravenous antibiotic therapy (Li et al., 2019). The applicability of the OVIVA to low-resource settings where drug availability, patient expectations, and cost implications are different has not yet been demonstrated.
Antibiotic duration is controversial. Traditionally, 6 weeks of systemic therapy is employed after implant removal and up to 12 weeks when implants are retained. However, recent evidence suggests that shorter and predominantly oral regimens may be sufficient when debridement is adequate and when local antibiotic carriers are used.
Antibiotic carriers help deliver antibiotics to the site of the infection, providing targeted delivery and avoiding side effects of systemic administration. In a study of 433 confirmed FRIs, the use of local antibiotics in the definitive management of FRI reduced the rate of failure from 18.3 % to 10.3 % (McNally et al., 2022).
Polymethylmethacrylate (PMMA) is readily available and affordable in African settings. It can be used in spacer or bead form or for implant coating. Disadvantages include the need for multiple-stage surgery to remove the PMMA (which may itself become a foreign body and drive infection) and a poorer elution kinetics profile compared to absorbable carriers, which may lead to its colonization with drug-resistant bacteria. It is usually safe to add the equivalent of a normal daily intravenous or oral dose to a local carrier.
Examples of biodegradable absorbable carriers include calcium sulfate (e.g. Stimulan©) and calcium sulfate with hydroxyapatite ceramics (e.g. Cerament©, PerOssal©). Composite ceramics offer the advantage of being osteoconductive and have an improved drug elution profile. It may be that local aminoglycoside monotherapy is sufficient due to the high local antimicrobial concentrations achieved (Unsworth et al., 2024). Other local antibiotic carriers include intraoperatively made antibiotic-impregnated bone autograft or locally made carriers with calcium sulfate. The use of local antibiotic powder without a carrier is not recommended in the management of FRI as the antibiotic is likely to quickly dissipate into the interstitial fluid, negating its action in established infection (Saka et al., 2024).
Results presented from the SOLARIO trial (Dudareva et al., 2019b) suggest that up to 7 d of systemic antimicrobial therapy is non-inferior in relation to prolonged courses when local antibiotic carriers have been used. This will have significant benefits in compliance with antibiotic therapy and will play a role in improved antimicrobial stewardship. There are, however, no reliable supply chains for these products in most low-resource settings.
The primary objective of surgical management is to restore function, with a healed fracture and eradication of infection. Key surgical steps are illustrated in Table 3.
Key surgical steps in fracture-related infection management.
Intra-operative sampling is the investigation of choice in FRI, for diagnosis in the absence of clinical confirmatory criteria, and to guide management. For stable patients, antibiotic therapy should be stopped at least 2 weeks prior to sampling (Hoffmann et al., 2024), but a longer period without antibiotics is preferable to improve diagnostic yield (Li et al., 2014). Diagnostic sampling should be carried out as early as possible during surgery, and separate, clean instruments should be used for each sample to avoid contamination. The samples should be sent to an accredited laboratory as soon as possible to improve diagnostic yield. Samples should be sent for microbiology analysis (Dudareva et al., 2021) and histology analysis (Morgenstern et al., 2018). Five samples is the gold standard as the diagnostic yield rises from 84 % for three samples to 97 % for five samples (Dudareva et al., 2021). Where there are resource constraints, three samples is acceptable.
Two or more histology samples of bone and soft tissue should be taken and analysed in an accredited laboratory. In resource-constrained settings, this recommendation could be limited to cases of suspected FRI without clinical confirmatory criteria, and a single histology specimen may be acceptable. The presence of more than five neutrophils per high-power field (NPs/HPF) is significant, and it is a modality that is readily available in low-resource settings, particularly where laboratories carry out blood films for malaria testing.
All non-viable tissue and sequestra should be excised in a meticulous fashion. Viable bone should be preserved to avoid creating a critical bone defect. Bone should be assessed for bleeding, and excision should be progressive until punctuate bleeding (Paprika sign). It may be necessary to create a small elliptical window, with rounded edges to reduce fracture risk, in order to access the intramedullary space. Intramedullary reaming is recommended for extensive medullary infections and the management of infected intramedullary nails.
Copious irrigation with normal saline at low pressure is recommended to decrease the bacterial load and remove debris. The volume of fluid is less important than ensuring all loose fragments are flushed away. Irrigation is deemed to be complete only when the irrigation is flushing through clear with no remaining loose fragments seen in the fluid. The use of additives is not advised as they may add to cell toxicity, and the evidence base is weak regarding their efficacy.
Following sampling and excision, the bone and fracture should be assessed for healing and stability. This may be done under direct vision of the fracture or using fluoroscopy. If there is any movement detected, the bone should be stabilized as fracture stability is of upmost importance to achieve union and infection eradication.
In FRIs occurring in the first 2–3 months, where there is adequate stability, satisfactory reduction, and good potential for the fracture to heal, prior fixation may be retained. This may be done as part of a debridement and implant retention (DAIR) procedure (Buijs et al., 2022). DAIR may be a particularly useful approach in the low-resource setting where implant availability is restricted.
In late and/or chronic FRI, implants should be exchanged (unhealed fracture) as part of a debridement and implant exchange (DAIEX) procedure or removed (healed fracture). In low-resource settings, when the patient is unable to purchase a new implant, a relevant option is debridement and implant sterilization (DAIS): the implant is removed at the beginning of the surgical procedure, immediately sent for re-sterilization, and reused for fracture fixation after thorough debridement and irrigation during the same procedure.
The exchange of infected intramedullary nails should always be considered due to the difficulty of adequately debriding an infected nail. DAIS may be appropriate. Initial intramedullary nail retention and antimicrobial suppression may be a viable option, particularly if healing is progressing well; however, higher failure rates of infection eradication must be considered. Fracture stabilization with fine-wire circular frames is a good option, but its availability remains limited in many contexts.
Bone resection may create a bone defect which is poorly perfused. Approaches to filling bone defects include bone cement with or without antibiotics as part of the induced membrane Masquelet's technique, biodegradable ceramics delivering local antibiotics, antibiotic-impregnated bone grafts (AIBGs), bone allografts, bone transport (distraction osteogenesis) using an external fixator (including Ilizarov frame), and soft tissue transfer such a hemisoleus flap of a vascularized bone graft (fibula free flap).
For stable (contained) bone defects and non-critical unstable conical (uncontained) defects that are healing or expected to heal, normal autologous bone grafting may be a reasonable option in resource-constrained settings. When available, synthetic bone graft substitutes impregnated with antibiotics would also be an option. For critical unstable defects that are not expected to heal and complete (cylindrical) defects, the options are compression alone, compression with distraction at a second site, bone transport, or free fibula grafting. The vascularized fibula flap is often available to be used pedicled in the ipsilateral leg and can be harvested with additional soft tissue such as the flexor hallucis muscle and or a skin island. Other techniques such as the Papineau technique or Lautenbach technique may be reasonable options for healed FRIs in certain settings.
Definitive good-quality soft tissue closure is paramount to improve blood flow, improve antimicrobial delivery, aid fracture healing, and prevent recurrence. A good soft tissue envelope will also help contain local antibiotic carriers. Furthermore, definitive closure of wounds prevents further wound contamination. Local or free flaps may require additional expertise, and it is recommended to include plastic surgical expertise early, as part of surgical planning, if this is available.
There is limited evidence to support specialist dressings, including vacuum-assisted (VAC) dressings, in the management of open fractures and to show that prolonged VAC dressings are detrimental when used to treat FRI (Sweere et al., 2022).
Regular follow-up is vital to ensure that wounds heal, the fracture unites, and patients continue with their antibiotic therapy (including monitoring for side effects) and to identify early signs of recurrence. Patients should be reviewed by clinicians with FRI management experience for a minimum of 12 months. All patients with proven or suspected FRI should be discussed at regular bone and joint infection MDT meetings wherever possible. There may be particular barriers to follow-up such as geographical distance with poor transport links, cost, and time away from work. Other mechanisms may be considered such as involving community health workers and using messaging systems. Outcomes should be evaluated regularly; measures include re-operation rates, non-union, infection recurrence, amputation, and death.