What’s New in Musculoskeletal Infection

This update on musculoskeletal infection presents a review of infection-related articles from January 2019 through April 2020, in English-language journals, located using the National Center for Biotechnology Information website, with a special emphasis on periprosthetic joint infection (PJI). The additional sections cover recent salient articles in the areas of spine, trauma, hand, and pediatrics.

Overview

While PJI remains a devastating complication for patient and surgeon alike, progress in diminishing its prevalence may be occurring through preoperative optimization, intraoperative strategies, and postoperative wound therapy. In a recent study using data from the National Surgical Quality Improvement Program (NSQIP), the authors evaluated the incidence of surgical site infections (SSIs) in the first 30 days following total hip arthroplasty (THA) in 2016 compared with the years 2012 to 2015¹. The greatest incidence of combined superficial and deep SSI was in 2012 (1.12%), while the lowest incidence was in 2016 (0.81%), a 28% decrease (p < 0.01). However, the incidence of deep infection was not significantly different between 2012 (0.30%) and 2016 (0.23%) (p = 0.153). Interpretation of these data must be tempered by the fact that they only reflect 30-day outcomes.

Mortality following PJI remains a serious issue. A systematic review of 19,169 patients with an average age of 65 years noted a 5-year mortality rate of 21%. The 1-year mortality risk was significantly higher than that for an aged-matched group of U.S. citizens (p < 0.001; odds ratio [OR], 3.58)².

Treating patients with PJI is resource-intensive, requiring longer hospital stays and multidisciplinary treatment to ensure optimal results. An NSQIP database study compared resource utilization between 16,006 septic revision cases and 11,951 aseptic revisions. The authors found a significantly greater length of stay and rate of readmission, reoperation, and complications in the septic revision group (p < 0.001). As such, health policy-makers need to risk-adjust prospective payments to prevent the creation of disincentives to treat infected patients³.

Defining a successful result following the treatment of PJI remains elusive. In order to clarify this, the Musculoskeletal Infection Society (MSIS) convened a group to help define successful PJI treatment. They proposed a tiered system, with Tier 1 being infection control; Tier 2, infection control with suppressive antibiotics; Tier 3, need for reoperation and/or revision; and Tier 4, death. Use of this system will provide consistency in reporting and help to determine whether failure is directly or indirectly related to PJI⁴.

Prevention

The prevention of PJI after total joint arthroplasty (TJA) was a research topic of considerable interest in 2019. Articles generally focused on preoperative and intraoperative factors that affect PJI⁵. Several studies focused on the optimization of medical conditions prior to surgery to minimize infection risk. Two separate meta-analyses demonstrated a correlation of malnutrition with increased wound-healing and infection after arthroplasty (OR of 2.17 and 2.49 in the 2 studies, respectively)^6,7. Prospective research will be required to determine whether correction of malnutrition normalizes PJI risk. Bedard and colleagues showed an increase in postoperative wound complications (OR, 1.78) and PJI (OR, 2.02) after TJA among tobacco users compared with nonusers in a meta-analysis of 14 studies⁸. In a study analyzing the data of U.S. veterans with hepatitis C in the TJA setting, it was shown that preoperative treatment of hepatitis C reduced the risk of postoperative infection roughly 2 to 3-fold after TJA⁹.

Perioperative antibiotic use is a powerful strategy for the prevention of PJI. In a retrospective study of 20,682 primary TJAs, the authors demonstrated through regression analysis that a single dose of preoperative intravenous (IV) antibiotics is as effective as 24 hours of perioperative dosing in the prevention of postoperative PJI¹⁰. Two other studies showed no benefit from the use of commercially available antibiotic-impregnated bone cement for total knee arthroplasty (TKA)^11,12. Regarding local reduction of bacterial burden, Letzelter et al. presented a thorough review of different skin preparation solutions commonly utilized in orthopaedic surgery. They concluded that no consensus exists and that each area of the body may have its own microbiome, which may display different responses to the same skin antiseptic solution¹³. Goodman and colleagues demonstrated a higher prevalence of nasal methicillin-resistant Staphylococcus aureus (MRSA) carriage among patients with rheumatoid arthritis compared with controls, suggesting a possible explanation for higher infection rates postoperatively¹⁴.

Substantial effort has been dedicated to studying the effect of the operative environment on PJI. Results from the Mayo Clinic database called into question the idea that intraoperative dilute povidone-iodine irrigation is protective against PJI in both primary and revision TJA. In analyses of 11,738 primary and 2,884 revision cases, the authors found no difference in the rate of reoperation for infection at 3 and 12 months, comparing cases performed with and without the use of dilute povidone-iodine^15,16. In a small-scale pilot study, Cook et al. showed a significant reduction in postoperative infection after primary TJA performed in an operating room with a supplemental ultraviolet-light-based air decontamination system¹⁷. In an observational study, Klaber et al. showed a low overall rate of bacterial contamination of surgical gowns during hip arthroplasty procedures, provided the operative time was <120 minutes, highlighting the importance of efficiency in the operating room¹⁸.

Surgical and patient factors have also been studied in relation to PJI after TJA. Taneja and colleagues showed that patients who received an allogeneic blood transfusion had a higher PJI risk compared with those who were not transfused¹⁹. Prolonged duration of surgery in primary TJA was shown to be associated with increased risk of PJI²⁰, with a separate study showing an 18% increase in PJI for every 15 additional minutes in the operating room for primary TKA²¹. In a prospective randomized controlled trial, Newman and colleagues showed that patients undergoing revision TJA whose wounds were dressed with a negative-pressure device had a lower 12-week reoperation rate than those whose wounds were dressed with a standard dressing²². Keeney et al. showed similar early benefits in primary TJA dressed with a negative-pressure device, but there was no difference in longer-term PJI risk²³.

Diagnosis of PJI

Shohat et al. validated an algorithm for diagnosing PJI using defined diagnostic criteria²⁴; they defined PJI cases using only major MSIS criteria²⁵ (n = 684) and, as a control, used aseptic revision cases that did not demonstrate failure for any reason within 2 years (n = 820). The algorithm was validated using a separate cohort (n = 422), showing a sensitivity of 97% and specificity of 99.5%. Unfortunately, the diagnosis of PJI within 6 weeks of the index arthroplasty is still elusive, as current accepted thresholds for synovial cell count and inflammatory markers have low sensitivity (<60%), especially in coagulase-negative Staphylococcus cases²⁶. Likewise, Lazarides et al. showed that using inflammatory markers to diagnose PJI is not suitable for immunosuppressed patients, and newly described thresholds for synovial cell count and differential have better operative characteristics²⁷. Moreover, Saleh et al. described how the evidence that has been used to set some of the diagnostic criteria is poor at best if analyzed using a validated Quality Assessment of Diagnostic Accuracy Studies tool, calling for better design in future studies²⁸.

Joint Aspiration

Li et al. described a method to increase culture-positive yield when ≤1 mL of synovial fluid is aspirated: use of saline solution lavage (10 mL) and reaspiration of the joint. The samples were sent in 2 blood culture bottles for inoculation for 14 days in a BACT/ALERT 3D blood culture system (BioMérieux). The authors reported a sensitivity of 80% and specificity of 96%²⁹. Likewise, measurement of synovial fluid viscosity may be a promising test to aid in PJI diagnosis, as it outperformed C-reactive protein (CRP) level and erythrocyte sedimentation rate (ESR), with a sensitivity of 93.3% and a specificity of 66.7%³⁰.

Serum Biomarkers

D-dimer use for PJI diagnosis is still controversial, as Xu et al. demonstrated in a retrospective study. They found that CRP level, ESR, and interleukin (IL)-6 were diagnostically superior to D-dimer. Using an optimized threshold of 1.02 mg/L, the sensitivity of D-dimer was only 68%³¹.

Synovial Biomarkers

The alpha-defensin (AD) test continues to demonstrate diagnostic utility in particular scenarios. Among immunosuppressed patients, it demonstrated a sensitivity of 93% and a specificity of 100% in a retrospective study³². However, its use in reimplantation cases demonstrated low sensitivity³³. Additionally, using Delphi-based criteria, at 1 year of follow-up³⁴, 19.5% of the cases that had negative results prior to the second stage demonstrated treatment failure. Kleiss et al. also showed that the routine use of AD to evaluate painful hips or knees after arthroplasty is insufficient to accurately diagnose PJI, particularly in low-virulence infections³⁵. However, the leukocyte esterase (LE) test was reported to be useful even in patients who were using antibiotics at the time of aspiration (sensitivity of 78%)³⁶. Additionally, another study showed that the LE assay is better than or equivalent to frozen section histology (sensitivity and specificity of 100% and 93.8%, respectively, compared with 78.3% and 96.9%)³⁷.

Molecular Diagnosis

Torchia et al. did a cost-effectiveness analysis of the use of next-generation sequencing and concluded that such a test should be reserved for clinical contexts with a high pretest probability of PJI³⁸. In a multicenter study, polymerase chain reaction assay using restriction fragment length polymorphism to detect bacteria demonstrated a sensitivity and specificity of 97.4% and 100%, respectively³⁹.

Imaging

The diagnostic accuracy of preoperative computed tomography (CT)-guided hip aspiration showed a sensitivity of 66.7% and a specificity of 98%⁴⁰. Enlarged iliac lymph nodes and aspirated fluid volume of >1 mL were strong predictors of infection.

Treatment

Published research in the period of this review continued to clarify the role of each of the 3 major treatment options for PJI: irrigation and debridement, 2-stage exchange, and 1-stage exchange.

Irrigation and Debridement

Irrigation and debridement with component retention continues to show high failure rates (56%) even in the acute hematogenous setting with <3 weeks of symptoms⁴¹. A study looking at a 2-stage debridement protocol with component retention in 83 patients showed an 86.7% success rate of infection control with an average follow-up of 41 months⁴². This shows considerable improvement when compared with previously reported results of single-stage debridement with component retention.

The reported success rates for 2-stage exchange following a failed irrigation and debridement procedure remain variable. In a series of 291 knees with an average follow-up of 6.2 years, the success rate was 81% (51 of 63) in the direct staged revision group and 72% (159 of 221) in the failed irrigation and debridement group⁴³.

Two-Stage Exchange

Recent studies have called into question the success rates of 2-stage exchange. A lack of infection control may lead to the need for an interim spacer exchange for a number of patients, with a rate of 16.9% (90 of 533) seen in 1 recent study⁴⁴. Some patients may never undergo the second stage; a separate study found that to 18% (111 of 616) did not undergo reimplantation⁴⁵. These groups of patients traditionally show higher rates of treatment failure and poor outcomes. Polymicrobial and fungal infections remain challenging to treat, with higher failure rates. In a recent study, polymicrobial infections were associated with lower treatment success, increased operative times, and longer hospitalizations when compared with single-organism infections⁴⁶. In another study, fungal PJI resulted in a 61.5% failure rate with 2-stage exchange at an average of 33 months of follow-up⁴⁷.

The results of a repeat 2-stage exchange for the treatment of recurrent knee infection was found to be largely dependent on MSIS grade, soft-tissue status, and the presence of an antibiotic-sensitive organism⁴⁸. In this series, all MSIS type-C hosts became reinfected.

Different spacer types were examined in numerous studies. Articulating antibiotic spacers remain the standard for the interim treatment of PJI. However, static spacers showed comparable success rates in situations of severe bone loss, with an infection-eradication rate of 77%⁴⁹. Primary TKA components with high-dose antibiotics are becoming increasingly popular for interim treatment. A multicenter study looking at these spacer types for definitive treatment showed a high rate of success, at 78.9% with minimum 2-year follow-up⁵⁰. In 1 study, the addition of intramedullary dowels to antibiotic spacer constructs did not enhance infection eradication, despite their common use⁵¹.

One-Stage Exchange

While a prospective multicenter randomized study comparing 1-stage and 2-stage exchange is ongoing in the U.S., results are not yet available. However, 1-stage exchange continues to gain enthusiasm as treatment for PJI. A recent study examined 111 patients undergoing single-stage revision with cementless components for the treatment of chronic infection following THA. Ninety-nine (89.2%) of the patients were infection-free at a mean follow-up of 58 months⁵². Another study showed a 10-year rate of infection-free survival of 94% following 1-stage exchange for chronic PJI of the hip⁵³. Which patients are good candidates for 1-stage exchange remains a question. In a study including 91 patients who had infection recurrence following 1-stage exchange TKA, risk factors for failure were identified. The authors found that a previous 1 or 2-stage exchange due to PJI and the isolation of Enterococcus or Streptococcus bacteria were associated with a higher risk of failure with 1-stage exchange⁵⁴.

Antimicrobial Treatment

Vancomycin Powder

The use of vancomycin powder in various surgical procedures remains controversial. Two systematic reviews involving prophylactic vancomycin powder use in spine surgery demonstrated that the use of intrawound vancomycin powder significantly reduced the relative risk of developing a surgical site infection; however, there was significant heterogeneity among the studies reviewed^55,56. Several other studies demonstrated reductions in deep and superficial SSIs with the use of intrawound vancomycin in spine surgery⁵⁷. Caution should be exercised, however, since 1 meta-analysis found that spine patients who received intrawound vancomycin had less gram-positive SSIs but were twice as likely to develop a gram-negative or polymicrobial SSI compared with patients in the control group, and other studies demonstrated similar results^58-60. A study using a rat model of spinal fusion found that local application of vancomycin may have an inhibitory effect on fusion⁶¹. Two systematic reviews and meta-analyses of intrawound vancomycin in extremity surgery as well as primary and revision arthroplasty^58,62 demonstrated decreased rates of PJI; however, both studies recognized that only low-quality evidence exists. They both concluded that, before widespread adoption of this practice, large prospective randomized trials are necessary.

Oral Antibiotics in Primary and Revision Arthroplasty

A retrospective study and meta-analysis evaluated the clinical importance of a positive culture at reimplantation in 2-stage exchange arthroplasty. Pooled data of 11 studies showed a higher risk of failure among patients with a positive culture (OR, 4.58). Six weeks of antibiotic treatment following reimplantation decreased the odds of reinfection; however, the risk of reinfection remained 3 times higher for patients with a positive culture at reimplantation than for patients with a negative culture. The initial infecting organism did not match the organism isolated during reimplantation in 82.6% of the study population. There was also substantial heterogeneity in the antibiotic regimen and duration as well as in the protocols for 2-stage exchange, making interpretation difficult⁶³.

A large randomized controlled trial evaluated oral versus IV antibiotics for the management of bone and joint infections and demonstrated that oral antibiotic therapy was noninferior to IV therapy for the treatment of bone and joint infections. The results of this study challenge the traditional management of bone and joint infections using IV antibiotics and may result in a change of clinical practice⁶⁴.

The utility of an antibiotic-free period prior to reimplantation in 2-stage exchange arthroplasty was evaluated in a nonrandomized study, which demonstrated a higher cure rate in patients at 1 center who did not stop antibiotics prior to reimplantation compared with those at another center who discontinued antibiotics for 2 weeks. However, the methodology may reflect a comparison of treatment centers rather than antibiotic usage, making interpretation of the data difficult⁶⁵.

Antibiotic Stewardship

A single-center study evaluated perioperative antibiotic choices for patients undergoing TJA. The risk of PJI was 32% lower among patients who received cefazolin compared with those who received other antimicrobial agents, emphasizing the importance of preoperative allergy testing in patients with stated beta-lactam allergies to ensure that patients receive optimal perioperative prophylaxis⁶⁶. A review of regional and state antibiograms demonstrated that 75% of methicillin-sensitive S. aureus (MSSA) isolates and 60% of both MRSA and coagulase-negative Staphylococcus isolates were susceptible to clindamycin, whereas 99% of all isolates were susceptible to vancomycin⁶⁷. These data can inform antibiotic choice for perioperative prophylaxis in patients with true penicillin allergies. The appropriateness of antibiotic prophylaxis prior to dental visits was assessed in a large study, using current guidelines as a reference, that included manipulation of the gingiva or tooth periapex. The authors found that >80% of antibiotics prescribed before dental visits for a variety of conditions (including TJA) are unnecessary⁶⁸.

Antibiotic Complications and New Therapies

The incidence of Clostridium difficile colitis during inpatient hospital stays following revision TKA was found to be 1% and was associated with longer length of stay, higher costs, and greater mortality⁶⁹. A case report demonstrated successful treatment of osteomyelitis due to an extensively drug-resistant organism with a combination of antibiotics and bacteriophages⁷⁰.

Spine

Accurate microbiologic diagnosis is important for the management of native vertebral osteomyelitis (NVO). Unfortunately, CT-guided biopsy identified a pathogen in only 33% of cases, on the basis of a systematic review of 10 studies⁷¹. The findings of a separate observational study suggest that a repeat biopsy should be delayed by 3 days after the first, as approximately 75% of pathogens will be identified by that time⁷². There remains debate about the timing and necessity of surgical debridement or abscess drainage for NVO. In 1 observational study, early drainage of an epidural/paravertebral abscess provided a protective effect against disability or pain at 12 months⁷³. A retrospective study that used propensity scoring to account for confounders found a trend toward pain reduction with a medical-surgical strategy compared with medical treatment alone⁷⁴.

A large meta-analysis of SSI after spine surgery suggested an increased risk of infection among patients with diabetes, obesity, hypertension, prolonged operative time, and transfusion⁷⁵. An observational study evaluated the impact of 5 preventive techniques on SSI rates: intrawound vancomycin powder, dilute Betadine (povidone-iodine) irrigation, preoperative chlorhexidine scrubs, preoperative nasal screening and decolonization of S. aureus, and perioperative antibiotic administration⁷⁶. A notable reduction in pooled SSI rate, from 6% to 2%, was observed and was found to be highly cost-effective, despite variability in individual surgeon use of several of these strategies.

Pediatrics

The use of a clinical pathway to prevent SSI after posterior spinal fusion in high-risk children was evaluated in a recent study⁷⁷. A marked decrease in deep SSI was observed, from 8% before to 1% after pathway implementation. Among children with acute hematogenous osteomyelitis (AHO), a clinical care algorithm was developed using formal quality-improvement methods and led to cost savings and improved empirical antimicrobial prescribing⁷⁸. The validation and modification of a previously published severity of illness (SOI) score for AHO correlated well with several clinically meaningful surrogate measures⁷⁹. Among a prospective cohort of children with AHO, a severe SOI score predicted osteonecrosis, chondrolysis, or deformity after 2 years of follow-up but did not predict functional outcomes⁸⁰. Long-term, severe adverse outcomes occurred in 7.9% of children overall, but in only 1.3% of children with a mild SOI score, suggesting a way to tailor counseling and follow-up care.

Trauma

Infection after open fracture causes substantial morbidity in children and adults. However, there are little data directly comparing the frequency of infection between the 2 groups. A single-center cohort study found a significantly lower rate of infection among children than adults in patients with open tibial fracture⁸¹. Hypothesized reasons for the observed differences include the injury type, fixation devices used, or soft-tissue envelope.

The management of open reduction and internal fixation (ORIF)-associated infection is often extrapolated from the management of PJI, but it is a separate clinical entity. A 10-year series of consecutive ORIF-related infections demonstrated an overall failure rate of 11.7%⁸². While implant retention was successful for >85% of infections encountered ≤10 weeks after ORIF, failure occurred one-third of the time when late infections were encountered >10 weeks after ORIF.

Hand

The management of open fractures of the hand has traditionally been extrapolated from lower-extremity injury treatment or small, single-center studies. Utilizing the NSQIP database, open metacarpal, proximal, or middle phalangeal fractures were not associated with greater 30-day infection risk compared with closed fractures⁸³. Furthermore, operative management of open fractures >1 day from presentation was not associated with increased risk of infection.

Conclusions

The prevention, diagnosis, and treatment of musculoskeletal infection remain challenging. Preoperative patient optimization and vigilance with regard to sterile technique are essential components of infection prevention. A high index of suspicion for the possibility of infection is critical when a patient with painful musculoskeletal symptoms presents. The appropriate use of currently available diagnostic tools is important for the clinician to fully understand. Collaborative decision-making between orthopaedic surgeons and infectious disease consultants is critical for success. Prompt intervention using evidence-based treatment guidelines should maximize results.

Evidence-Based Orthopaedics

The editorial staff of JBJS reviewed a large number of recently published studies related to the musculoskeletal system that received a higher Level of Evidence grade. In addition to articles cited already in this update, 6 other articles with a higher Level of Evidence grade were identified that were relevant to musculoskeletal infection. We have provided a brief commentary about each of the articles to help guide your further reading, in an evidence-based fashion, in this subspecialty area.

Evidence-Based Orthopaedics

Calkins TE, Culvern C, Nam D, Gerlinger TL, Levine BR, Sporer SM, Della Valle CJ. Dilute Betadine lavage reduces the risk of acute postoperative periprosthetic joint infection in aseptic revision total knee and hip arthroplasty: a randomized controlled trial. J Arthroplasty. 2020 Feb;35(2):538-543.e1. Epub 2019 Sep 12.

In this prospective randomized study of total hip and knee revisions, 478 patients received either dilute Betadine or normal saline solution lavage at the end of their procedure. There were 8 infections in the saline solution group (3.4%) and only 1 in the dilute Betadine group (0.4%). While these data are encouraging, there are a few caveats concerning their interpretation. A 3.4% infection rate is high and, without MSIS host classification being utilized, this trial is hard to interpret. Additionally, the dilute Betadine group had the skin edges prepared with Betadine at the time of closure, while the normal saline solution group did not. Whether or not this affected the results is not known.

Li M, Zeng Y, Wu Y, Si H, Bao X, Shen B. Performance of sequencing assays in diagnosis of prosthetic joint infection: a systematic review and meta-analysis. J Arthroplasty. 2019 Jul;34(7):1514-1522.e4. Epub 2019 Mar 7.

The authors conducted a systematic review and meta-analysis to assess the performance of multiple molecular diagnostic sequencing techniques in the diagnosis of PJI. The pooled sensitivity was 0.81, with a specificity of 0.94. Sequencing by synthesis seemed to be advantageous; however, receiving antibiotics before sampling decreased the sensitivity. While this technology is encouraging, challenges still exist concerning the frequent identification of falsely positive organisms. Other published studies have identified organisms through sequencing in aseptic synovial fluid taken from primary arthroplasties.

Mercurio M, Castioni D, Iannò B, Gasparini G, Galasso O. Outcomes of revision surgery after periprosthetic shoulder infection: a systematic review. J Shoulder Elbow Surg. 2019 Jun;28(6):1193-203. Epub 2019 Apr 16.

The authors evaluated 34 studies in a systematic review including 754 patients. They found that Cutibacterium acnes was the most common organism. The success rate was 96% for a 1-stage procedure, 93% for a permanent spacer, 86% for a 2-stage procedure, and 65% using irrigation and debridement for treatment. While the 1-stage data are encouraging, results could be influenced by selection bias. Healthier hosts may have undergone a 1-stage procedure, and less healthy hosts or patients with more virulent organisms may have undergone a 2-stage procedure. Therefore, a prospective randomized study with host classification would be necessary to determine superiority.

Mirza SZ, Richardson SS, Kahlenberg CA, Blevins JL, Lautenbach C, Demetres M, Martin L, Szymonifka J, Sculco PK, Figgie MP, Goodman SM. Diagnosing prosthetic joint infections in patients with inflammatory arthritis: a systematic literature review. J Arthroplasty. 2019 May;34(5):1032-1036.e2. Epub 2019 Jan 31.

The authors attempted to differentiate between septic and aseptic failure among patients with inflammatory arthritis. They reviewed 20 articles that met criteria. Five studies had pooled data for analysis. They reviewed 1,499 aseptic cases and compared them with 452 septic cases. While the most sensitive test was the percentage of neutrophils in aspirated material, the authors were unable to identify a biomarker to distinguish between aseptic and septic cases since disease activity was not specified in the available studies. Since serologic markers are an integral part of the diagnostic criteria to determine PJI, the influence of disease flares, which may increase serologic markers, and immune-modulating drugs, which may decrease serologic markers, makes this determination still problematic. Further research needs to be performed to clarify this issue.

Nene AM, Patil S, Kathare AP, Nagad P, Nene A, Kapadia F. Six versus 12 months of anti tubercular therapy in patients with biopsy proven spinal tuberculosis: a single center, open labeled, prospective randomized clinical trial-a pilot study. Spine (Phila Pa 1976). 2019 Jan 1;44(1):E1-6.

The authors of this prospective study attempted to determine the length of time necessary for antituberculous therapy in biopsy-proven tuberculosis of the spine. In this study of 100 patients, the authors concluded that antituberculous treatment of 6 months gave similar outcomes to a 12-month course, at 24 months of follow-up. This is an important study because there are no robust guidelines for the treatment of tuberculosis of the spine. There is only an extrapolation from pulmonary tuberculosis treatment. The results of this study may guide treatment, as prolonged antibiotic treatment may have the potential for the emergence of multidrug-resistant tuberculosis bacilli.

Zhang Z, Swanson WB, Wang YH, Lin W, Wang G. Infection-free rates and sequelae predict factors in bone transportation for infected tibia: a systematic review and meta-analysis. BMC Musculoskelet Disord. 2018 Dec 13;19(1):442.

A systematic review of 22 studies including 518 patients evaluated the infection rate with external fixators versus hybrid fixators (combined internal and external fixators) in the treatment of infected tibial nonunion with bone transport. The infection rate was 14% in the hybrid-fixator group versus only 3% in the external-fixator group. Residual biofilm in internal fixators in the hybrid group after debridement may have played a role in this treatment disparity. However, extended use of external fixators leads to not only patient inconvenience but more joint stiffness and pin-track infections. Whether accelerated bone-transport techniques or antibiotic-coated internal-fixation devices could safely shorten external-fixator time remains to be determined.