Accurate diagnosis of implant-related infections in orthopaedics, particularly in orthopaedic trauma, has long been encumbered by difficulties in identifying microorganisms.1 Culture methods of diagnosis have been the standard of care and comprise most of the FDA’s approved diagnostic instruments used in hospitals in the United States. Molecular methods have been reserved for situations in which specific pathogens are difficult to grow in vitro. The sensitivity and accuracy of traditional culture methods have come under increasing skepticism because of the number of studies suggesting that these techniques are insensitive and inaccurate. Phenotypic differences among bacterial strains of the same species may account for the poor sensitivity of cultures.2,3 Technologic advances in diagnostics have been applied to other fields of orthopaedics, such as prosthetic joint infections, and are a vital aspect of diagnostic algorithms.4 However, implementation of these tests in the setting of orthopaedic trauma has been slow to nonexistent.
Bacteria are one of the most successful forms of life on earth. Although the genetic heterogeneity of bacterial species is extensive, their ability to respond phenotypically to environmental stimuli has led to their success.5 Biofilm and planktonic bacteria account for two main phenotypes that are encountered in medicine. Planktonic bacteria are free floating, have a high metabolic rate, and divide rapidly. As a result, they are susceptible to the surrounding hostile environment, including systemic and local antibiotics and phagocytes6 (Figure 1).
In contrast to the planktonic phenotype, the biofilm phenotype is characterized by a protected mode of growth in which bacteria undergo a metamorphosis into a multicellular organism that provides for its survival in an unfavorable environment as an adherent form at the expense of individual cells. Specific structures that allow cross-talk between different regions form in the biofilm, and channels allow for the transport of nutrients.6 Biofilm bacteria also produce an extracellular matrix that provides a protective coating to withstand shear stress and guard against an external biocidal environment.7 This protective environment allows them to be sessile, with reduced metabolic activity and relatively low cell division rates. Within this structure, microbes are resistant to antibodies, phagocytes, and antibiotics; they are even capable of releasing planktonic bacteria to disseminate to neighboring environments. This can lead to tissue damage around the zone of infection because of the activation of phagocytic enzymes. As a result, the opportunistic biofilm can now expand and infect the weakened surrounding tissue.
Acute bacterial diseases caused by planktonic bacteria cells have been largely controlled by the development of antibiotics.8 The success of antibiotics relies on accurate identification of bacteria. Planktonic, free-floating bacteria replicate and form colonies on agar plates with appropriate nutrients. This relatively easy means of replication facilitates the detection and identification of numerous bacterial species and measurement of antibiotic resistance.9 Although culture methods continue to be the standard of care for diagnosing infection, these methods have been shown to be inaccurate and lack sensitivity for diagnosis of biofilm-related infection.4,10,11
Internal fixation implants used in orthopaedic trauma serve as an inert surface for the propagation of biofilms. Not only are biofilm bacteria difficult to diagnose, they are also challenging to treat. Biofilm infections have been shown to be resistant to systemic antibiotic therapy, even at 100 to 1,000 times higher doses than that needed to treat planktonic infections of the same strain.12
Overview and Risk Factors
The diagnosis of postoperative infection in trauma patients requires a combination of clinical examination, laboratory data, plain radiography, advanced imaging, and intraoperative findings such as Gram stains and cultures. Infections can lead to nonunion and become a source of patient disability and pain and often require multiple revision procedures, resulting in significant costs to patients and the healthcare system.13 Risk factors for infection after orthopaedic trauma include patient-related factors (eg, smoking, diabetes mellitus, poor nutrition, homelessness, compromised immune status, vascular disease) and injury-related factors such as open traumatic wounds, soft-tissue loss, and vascular damage.14 Large et al15 found that surgical fixation of lower extremity fractures in the setting of systemic infection had no impact on the rate of surgical site infection, and the most important factor in predicting infection was the presence of an open injury.
Clinical Examination, Laboratory Testing, and Basic Imaging
Trauma patients with postoperative infections may present with any combination of surgical site erythema, pain, swelling, or drainage. Presentation can be delayed or appear in the immediate postoperative period. Systemic manifestations may include fever, night sweats, malaise, anorexia, and weight loss.
Laboratory values are often elevated in patients with infections. Routine laboratory evaluation includes acute phase reactants such as erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) level, white blood cell count, and peripheral blood cultures. The ESR and CRP level are nonspecific indicators of inflammation, with ESR primarily reflecting the concentration of fibrinogen and immunoglobulin in the plasma. CRP is an acute phase protein made in the liver in response to inflammation, infection, injury, and surgery.16 The serum levels rise rapidly 4 to 6 hours after stimulation, doubling every 8 hours after that and peaking 36 to 50 hours later.17 The levels then decline as long as there is no ongoing tissue destruction or inflammatory source. The half-life of serum CRP has been calculated to be 2.6 days and elimination follows first-order kinetics.18 ESR is slower to rise after physiologic insult and can take weeks to normalize in spite of clinical improvement.17 The use of trending CRP levels has been more reliable than trending ESR for diagnosis of many orthopaedic infections. Although elevation of ESR and CRP level is normal in the acute postoperative period, continued or recurrent elevation can be an indication of infection. Evaluation of these parameters has improved diagnostic sensitivity to approximately 98%19 but only for acute phase pathogens. These parameters are not useful for diagnosis of infections associated with slow-growing anaerobes and treponemes. However, these values can be trended over time to monitor infection improvement and resolution. The literature on pediatric orthopaedic infection suggests that CRP level may be more sensitive than ESR.13 Less frequently, patients may present with leukocytosis.19 Bacteremia can be identified with peripheral blood cultures, but these cultures are not specific and bacteremia may be caused by myriad other conditions, including urinary tract infection and pneumonia.
Radiography is a common part of the diagnostic workup. For trauma patients with infection, plain radiographs may show associated soft-tissue swelling or evidence of joint effusion, which may indicate an infection, or the direct osseous effects of the infectious process such as bone destruction, implant loosening, or failure. Radiographic changes often reflect the destructive aspect of this process. Focal changes, such as osteopenia and periosteal thickening and/or elevation, may be delayed days to weeks because at least 50% to 75% of the bone matrix must be destroyed before radiographs show lytic changes.20 The lack of radiographic changes in plain film does not rule out infection. Mok et al21 reviewed 22 cases of neonatal osteomyelitis over 10 years and found that initial plain radiographs were abnormal in 5% of cases, radiographs obtained at week 1 and week 4 demonstrated an increase in the number of abnormal results (33% and 90% of cases, respectively). Implant loosening, implant failure, and nonunion may also indicate infection.10,22 All patients with nonunions should be evaluated for infection.
CT and MRI
Advanced imaging modalities such as CT and MRI provide detailed anatomic projections and have a role in diagnosis of postoperative infection. CT with contrast can aid identification of soft-tissue abscesses. MRI can delineate soft-tissue structure and highlight inflammation and fluid collection. Osteomyelitis remains difficult to diagnose even with these advanced imaging modalities, which are significantly impaired by metal implant artifact, a common problem in orthopaedic trauma patients.23 Moreover, the normal metabolic activity of bone healing after fracture can result in false-positive findings that indicate infection at the healing fracture site.
Nuclear Medicine Studies
Nuclear medicine techniques, which are less prone to metal artifact, are used to diagnose musculoskeletal infections. Bone scintigraphy (ie, bone scan) has high sensitivity but lacks specificity for the diagnosis of osteomyelitis.24 Technetium tc-99m phosphate is the most common isotope used for this study because it is inexpensive and has a relatively short half-life. After infusion, the isotope is distributed through the body and is either excreted in the urine or accumulates in bone, favoring areas of increased metabolic activity. Typically, there are three phases to the bone scan study when evaluating for infection. The blood flow phase is the first phase of the scan; it is a series of dynamic images that focus on the area of concern immediately after infusion. In the second phase, a static image is obtained, demonstrating bone pooling. The third phase is performed 3 hours after the second phase, when there is less isotope activity in the soft tissues, and an area of increased signal at the infection location should be identified. Although infection induces increased bone metabolism, resulting in increased isotope uptake, malignant foci and fracture sites are also areas of increased metabolic activity and isotope uptake. Bone scans are also commonly used as a diagnostic tool for stress fractures. Recently, the use of technetium tc-99m–labeled ceftriaxone was proposed to improve the specificity of bone scans in detecting infections by increasing the concentration of the agent at the site of bacterial colonization.25
Gallium-67 citrate (67Ga) also has been used with technetium in bone scan studies to increase specificity. It rapidly binds to proteins in the serum, especially transferrin, and is taken up into cells (particularly leukocytes). In the setting of infection, bone scans have increased specificity because of the increased concentration of leukocytes at the infection site caused by inflammation and increased capillary permeability. Bacteria at the site of infection have high metabolic needs and perform rapid uptake of iron and 67Ga, leading to increased signal at the infection site. Images are routinely taken 48 to 72 hours after infusion; however, because the study is a reflection of increased metabolism, other physiologic processes, such as fracture healing, cannot be differentiated with this study.
Indium In-111 (111In)– or technetium tc-99m–labeled leukocyte scans are preferred for diagnosis of osteomyelitis,23 with each scan having excellent diagnostic accuracy and reported sensitivities as high as 95%. Technetium tc-99m–labeled leukocyte studies have superior imaging characteristics and have widely replaced 111In–labeled leukocyte scans. Both depend on leukocyte influx into infected tissues and detect the concentration gradient between tissues. Merkel et al26 compared the efficacy of 111In–labeled leukocyte and technetium-Ga scans for diagnosis of low-grade osteomyelitis and found the indium scans to be 26% more accurate than technetium-Ga scans. The authors also reported that the indium scans had an accuracy rate of 94% in patients with implanted hardware, but found a 14% false-positive rate in cases of fracture. The drawbacks of labeled leukocyte scans include laborious preparation, the need for expensive equipment, and the need to handle and process potentially pathogenic blood products.
F-18 fluorodeoxyglucose (FDG)–positron emission tomography (PET) has an emerging role in osteomyelitis detection. FDG is a glucose analog metabolite that is transported into lymphocytes, neutrophils, and macrophages where it is phosphorylated and accumulates within the cells.23,27,28 Similar to labeled leukocyte scans, the immunologic cells preferentially influx into infected tissue. FDG-PET is used to identify loci of soft-tissue inflammation and osteomyelitis and has a sensitivity of 94% to 100% and a specificity of 87% to 100%.23 This technology has been applied to the diagnosis of prosthetic joint infection, but polyethylene particulate debris can lead to increased FDG uptake and may hinder specificity.29 In patients with metal implants, CT and MRI are limited by metal artifact; FDG-PET has improved lesion-to-background contrast and can be used to detect soft-tissue or bone infection in patients with implants.30 The half-life of FDG is short (approximately 2 hours), which allows future nuclear imaging studies to be performed without concern over residual tracer altering subsequent studies.31 However, because of its short half-life, an onsite facility is often required to produce and handle the substance, which may be cost-prohibitive.
Schiesser et al32 prospectively evaluated the efficacy of FDG-PET for diagnosis of infection associated with metal implants in trauma patients. The authors reviewed 29 scans in 22 patients and compared the findings to microbiologic reports from surgical specimens and intraoperative findings. They reported that FDG-PET had a sensitivity, specificity, and accuracy of 100%, 93.3%, and 97%, respectively and concluded that this imaging technique was a sensitive and specific method for detecting infection in trauma patients with metal implants.
When concern for infection exists in the setting of orthopaedic trauma, multiple intraoperative wound swabs and/or tissue specimen cultures are obtained. Although no set standard exist for obtaining cultures, a minimum of three specimens (preferably five or six samples) should be sent for testing.3,33 A diagnosis of infection is presumed when more than one of three cultures is positive or when three or more of five or six cultures is positive. Although histopathology does not aid in identification of the offending organism, obtaining frozen intraoperative samples should be considered to differentiate true infection from contamination.34 The presence of five or more stromal neutrophil granulocytes in a single high-powered field constitutes a positive result for infection. The accuracy of histopathology findings relies on proper sampling of inflamed tissues and the experience of the pathologist. The use of microbiologic culture techniques and histopathology aids in accurate diagnosis of infection. Specific cultures for zoonotic and fungal organisms as well as mycobacteria are not routinely obtained but should be considered based on environmental and epidemiologic factors, including community exposure, travel history, previous infection, immunocompromised status, and a prior history of splenectomy.
Recently, Aggarwal et al35 examined the efficacy of swab and tissue cultures for diagnosis of periprosthetic joint infection. They found that tissue swabs were less effective than tissue culture (70% versus 93%). They also reported on the sensitivity, specificity, and positive and negative predictive values for tissue culture versus swabs (93% versus 70%; 98% versus 89%; 93% versus 68%; and 98% versus 90%, respectively).35 These culture-based techniques are ubiquitous in medicine and have remained largely unchanged since their introduction by Koch36 in 1884. However, these techniques have many limitations. Contamination can lead to false-positive results, and inadequate sample size, incorrect agar composition, and variable atmospheric conditions and temperatures may prevent identification of bacterial species.37,38 Rare pathogens (eg, Mycobacterium tuberculosis) require special culture media and growth conditions. Other pathogens, such as Propionibacterium acnes, require up to 13 days of incubation, significantly more time than the traditional practice of 3 to 5 days.22
Identification of pathogens that form biofilms using traditional culture techniques is challenging. Veeh et al3 reported that traditional culture techniques isolated Staphylococcus aureus in only 15 of 44 positive vaginal specimens (polymerase chain reaction [PCR] and DNA-sequencing techniques were also used to confirm diagnosis). In contrast, advanced molecular techniques (ie, fluorescence in-situ hybridization) detected the pathogen in all 44 specimens; in 37 specimens, the pathogen was in biofilm form. This suggests that the preferred culture technique is not effective for identifying bacteria in this form.
Furthermore, in the setting of orthopaedic trauma, traditional culture-based methods have been shown to be ineffective for the diagnosis of implant-related infections caused by biofilm.9 In a study of the efficacy of intraoperative cultures for diagnosis of infection in the setting of nonunion, Palmer et al10 found that intraoperative culture results were positive in 8 of 34 patients with “aseptic” nonunions, whereas molecular techniques detected bacteria in 30 of 34 patients, which calls into question the reliability of traditional culture techniques. The use of intraoperative Gram stain has been found to be highly specific but has poor sensitivity (as low as 25%) for detecting infection.19 Molecular techniques are more sensitive than traditional techniques for identification of bacterial pathogens, but the exact role of this technology has yet to be defined. Although the use of molecular techniques is being explored, the role that sonication plays in increasing the efficiency of current culture techniques is gaining ground.
Sonication of Removed Implants
Placement of implants (eg, screw, plate, intramedullary nail) within the wound area is known to increase the chance for bacterial colonization. The foreign body effect refers to the increased susceptibility to infection and morbidity that can occur when an intracorporeal foreign body is present. The surfaces of implanted devices are the eminent domain of microbes rather than tissue or host defense cells.39 The “race to the surface” concept described by Gristina et al39 is useful for understanding the relationship between tissue integration and bacterial colonization. The “surface” is a biomaterial implant and “the race” is between the host’s tissue-cell integration versus bacterial adhesion to the implant. If the contest is won by tissue then the surface is occupied, defended, and less available for bacterial colonization. The difference between bacterial colonization and clinically significant infection, however, has yet to be defined, and the implications of treating isolated microbes or their byproducts at the surface of implanted devices remain unclear. It should be noted that the presence of microbes does not always imply infection, and treatment is not always required. For most orthopaedic implants (eg, screw, plate, or intramedullary nail), host tissue cells arrive at the implant first and form a cohesive bond. Consequently, bacteria are confronted by host immune cells and will be less likely to colonize and form a biofilm. Infections resulting from bacterial colonization on biomaterials are difficult to eliminate because the bacteria are in a biofilm mode of growth. Treatment typically requires removal of the device.
Detecting the presence of bacteria in patients with orthopaedic implants can be challenging. Trampuz et al2 found that samples obtained through sonication of explanted hip and knee prostheses were shown to be more sensitive than tissue cultures for detecting infection. The process involves removal of the implant and the addition of Ringer solution followed by vortexing and sonication. Sonication itself uses low-frequency and low-intensity ultrasonography to cause cavitation bubbles that disaggregate biofilm. This causes microorganisms to be dislodged from the implant so they can be cultured. Sampedro et al40 also found this approach to be effective in patients with suspected spine implant infections. Compared with cultures of peri-implant tissue, sonicate fluid cultures had better sensitivity (73% versus 91%; P = 0.046) and specificity (93% versus 97%). In a study of sonication of intramedullary nails implanted to manage diaphyseal fractures of the femur, tibia, and humerus, cultures obtained from the nails detected infection in 15 of 31 cases compared with just 8 cases detected in unsonicated controls.41
The culture-based methods used today were developed in Germany in the late nineteenth century.8 Although these methods are effective for identifying and characterizing planktonic bacteria that cause acute systemic bacterial infections, cultures are poor at identifying bacteria within a biofilm. Virtually all fields of microbiology, except medical microbiology, have abandoned culture methods in favor of molecular diagnostics (MDx).
In MDx, DNA or RNA is extracted from a sample and purified. Short nucleic acid primers are used to amplify specific DNA sequences with PCR to a quantity great enough to measure. Notably, sequences of the 16S ribosomal DNA that encode the 16S ribosomal RNA gene often are used because this gene is universal to prokaryotes and can be used to identify organisms at the level of species, genus, or phylum.9 For species-specific (or higher taxa-specific) PCR-based approaches, primers with base sequences that match a target region in prokaryotic DNA will produce amplicons when that region is present. The issue is that specific target species and primers must be selected in advance. Thus, if the clinical laboratory makes the wrong selection, it will miss the infecting organism. These, narrowly based MDx methods provide excellent positive predictive value, but meaningless negative predictive value. Using more primers can become expensive and cumbersome, with most broad-based approaches taking too long for clinical applications.9
Hannigan et al42 recently showed that high throughput genomic sequencing of 16S prokaryotic RNA can be used to identify the bacteria present in traumatic open wounds. Genomic sequencing is not constricted by the need for a priori primers, which are required for traditional PCR-based techniques, and this technology may be a powerful tool in future diagnostics for infection.
The IBIS T5000 is an exciting new technology that can rapidly (within 6 hours) identify and discriminate among any and all bacterial species at the domain level of life; it is now available for research studies but has yet to be adopted for widespread clinical use. This technology combines multilocus PCR with electrospray ionization/time-of-flight mass spectrometry to provide rapid, high-throughput, and precise analyses of bacteria.43
Sample processing begins with PCR amplification of bacterial or fungal DNA. The amplified nucleic acids (ie, amplicons) are next introduced into a mass spectrometer by electrospray ionization and each intact strand of each amplicon is “weighed,” by its time of flight. The weight of the amplicon is then deconvoluted into an exact base composition that provides the number of A, C, G, and T nucleotides. This composition is compared with known bacterial and fungal base compositions to provide the corresponding taxonomic information. The database is constantly updated, so the system will be able to identify infecting organisms and distinguish them from contaminants. The domain-based amplification procedure uses 16 different primers that target sequences conserved at various taxonomic levels. Not all primers will support amplification of DNA from all species within the domain, but all species will support amplification with multiple primers. Through a triangulation process, it is possible to obtain a species-specific diagnosis in almost all cases by cross-referencing the individual amplicon base compositions. A similar approach is used for the detection and speciation of fungi.43 In addition, a molecular antibiogram can be obtained by including primers for important antibiotic-resistant genes (eg, mecA for methicillin-resistant Staphylococcus aureus, vanA for vancomycin-resistant Enterococci).
The major advantage of using the IBIS T5000 is that all bacterial DNA is amplified by the primer cocktail, rather than just organisms specifically selected, as with cultures and conventional PCR. This makes this technology not only a highly effective diagnostic method, but also an important research tool because new unexpected etiologies will arise from its routine use. These attributes make for a profoundly different diagnostic method compared with other approaches (eg, culture-based, MDx). An example of the results from clinical samples from infected orthopaedic hardware can be seen in Figure 2. The IBIS T5000 is an example of the emerging technologies available for orthopaedic surgeons. Recently, a molecular diagnostic system has been cleared by the FDA for blood, respiratory, and gastrointestinal infections.44 This approach uses panels for the intended application. For example, the blood culture identification panel tests for 24 pathogens (8 gram positives, 11 gram negatives, and 5 yeasts) and three antibiotic-resistant genes. Although the number of pathogens the panel detects is limited, this approach makes up for it by the relatively short amount of time it takes; only 2 minutes of preparation time is required and results are received in approximately 1 hour. Clinical comparison of the gastrointestinal FilmArray (BioFire) to another molecular system demonstrated clinical utility and both were better at detecting mixed infections than standard cultures.45 MDx approaches such as this could provide valuable information to surgeons.
Current culture-based diagnostic techniques are poor predictors of infecting organisms because bacteria within a biofilm cannot be identified using standard, culture techniques. Advanced imaging modalities such as FDG-PET, can increase the accuracy of identifying implant-related infections. Molecular diagnostics can also potentially provide promising new data that can be used to identify biofilm bacteria and select an appropriate antibiotic therapy.
The authors would like to thank Garth D. Ehrlich, MD for graciously providing the data for Figure 2 and for his critical reading of the manuscript.
Evidence-based Medicine: Levels of evidence are described in the table of contents. In this article, references 2, 10, 34, and 35 are level I studies. References 9, 13, 14, 18, 28, 40-42, and 45 are level II studies. References 19, 22, 24, and 29 are level III studies. References 3, 4, 15, 16, 21, 26, and 32 are level IV studies.
References printed in bold type are those published within the past 5 years.
1. Arnold WV, Shirtliff ME, Stoodley P: Bacterial biofilms and periprosthetic infections. J Bone Joint Surg Am 2013;95(24):2223–2229.
2. Trampuz A, Piper KE, Jacobson MJ, et al.: Sonication of removed hip and knee prostheses for diagnosis of infection. N Engl J Med 2007;357(7):654–663.
3. Veeh RH, Shirtliff ME, Petik JR, et al.: Detection of Staphylococcus aureus biofilm on tampons and menses components. J Infect Dis 2003;188(4):519–530.
4. Dowd SE, Wolcott RD, Sun Y, McKeehan T, Smith E, Rhoads D: Polymicrobial nature of chronic diabetic foot ulcer biofilm infections determined using bacterial tag encoded FLX amplicon pyrosequencing (bTEFAP). PLoS One 2008;3(10):e3326.
5. Costerton JW, Lewandowski Z, Caldwell DE, Korber DR, Lappin-Scott HM: Microbial biofilms. Annu Rev Microbiol 1995;49:711–745.
6. Costerton JW, Stewart PS, Greenberg EP: Bacterial biofilms: A common cause of persistent infections. Science 1999;284(5418):1318–1322.
7. Hoffman LR, D’Argenio DA, MacCoss MJ, Zhang Z, Jones RA, Miller SI: Aminoglycoside antibiotics induce bacterial biofilm formation. Nature 2005;436(7054):1171–1175.
8. Costerton JW, ed: The Biofilm Primer. Heidelberg, Springer, 2007.
9. Costerton JW, Post JC, Ehrlich GD, et al.: New methods for the detection of orthopedic and other biofilm infections. FEMS Immunol Med Microbiol 2011;61(2):133–140.
10. Palmer MP, Altman DT, Altman GT, et al.: Can we trust intraoperative culture results in nonunions? J Orthop Trauma 2014;28(7):384–390.
11. Wolcott RD, Ehrlich GD: Biofilms and chronic infections. JAMA 2008;299(22):2682–2684.
12. Patzakis MJ, Zalavras CG: Chronic posttraumatic osteomyelitis and infected nonunion of the tibia: Current management concepts. J Am Acad Orthop Surg 2005;13(6):417–427.
13. Pääkkönen M, Kallio MJ, Kallio PE, Peltola H: Sensitivity of erythrocyte sedimentation rate and C-reactive protein in childhood bone and joint infections. Clin Orthop Relat Res 2010;468(3):861–866.
14. Manson TT, Perdue PW, Pollak AN, O'Toole RV: Embolization of pelvic arterial injury is a risk factor for deep infection after acetabular fracture surgery. J Orthop Trauma 2013;27(1):11–15.
15. Large TM, Alton TB, Patton DJ, Beingessner D: Does perioperative systemic infection or fever increase surgical infection risks after internal fixation of femur and tibia fractures in an intensive care polytrauma unit? J Trauma Acute Care Surg 2013;75(4):664–668.
16. Hariharan P, Kabrhel C: Sensitivity of erythrocyte sedimentation rate and C-reactive protein for the exclusion of septic arthritis in emergency department patients. J Emerg Med 2011;40(4):428–431.
17. Jaye DL, Waites KB. Clinical applications of C-reactive protein in pediatrics. Pediatr Infect Dis J 1997;16(8):735–746; quiz 746-747.
18. Mok JM, Pekmezci M, Piper SL, et al.: Use of C-reactive protein after spinal surgery: Comparison with erythrocyte sedimentation rate as predictor of early postoperative infectious complications. Spine (Phila Pa 1976) 2008;33(4):415–421.
19. Stucken C, Olszewski DC, Creevy WR, Murakami AM, Tornetta P: Preoperative diagnosis of infection in patients with nonunions. J Bone Joint Surg Am 2013;95(15):1409–1412.
20. Butt WP: The radiology of infection. Clin Orthop Relat Res 1973;96:20–30.
21. Mok PM, Reilly BJ, Ash JM: Osteomyelitis in the neonate: Clinical aspects and the role of radiography and scintigraphy in diagnosis and management. Radiology 1982;145(3):677–682.
22. Butler-Wu SM, Burns EM, Pottinger PS, et al.: Optimization of periprosthetic culture for diagnosis of propionibacterium acnes prosthetic joint infection. J Clin Microbiol 2011;49(7):2490–2495.
23. van der Bruggen W, Bleeker-Rovers CP, Boerman OC, Gotthardt M, Oyen WJ: PET and SPECT in osteomyelitis and prosthetic bone and joint infections: A systematic review. Semin Nucl Med 2010;40(1):3–15.
24. Howie DW, Savage JP, Wilson TG, Paterson D: The technetium phosphate bone scan in the diagnosis of osteomyelitis in childhood. J Bone Joint Surg Am 1983;65(4):431–437.
25. Kaul A, Hazari PP, Rawat H, Singh B, Kalawat TC, Sharma S, Babbar AK, Mishra AK: Preliminary evaluation of technetium-99m-labeled ceftriaxone: Infection imaging agent for the clinical diagnosis of orthopedic infection. Int J Infect Dis 2013;17(4):e263–270.
26. Merkel KD, Brown ML, Dewanjee MK, Fitzgerald RH Jr: Comparison of indium-labeled-leukocyte imaging with sequential technetium-gallium scanning in the diagnosis of low-grade musculoskeletal sepsis: A prospective study. J Bone Joint Surg Am 1985;67(3):465–476.
27. Kubota R, Kubota K, Yamada S, Tada M, Ido T, Tamahashi N: Microautoradiographic study for the differentiation of intratumoral macrophages, granulation tissues and cancer cells by the dynamics of fluorine-18-fluorodeoxyglucose uptake. J Nucl Med 1994;35(1):104–112.
28. Rini JN, Bhargava KK, Tronco GG, et al.: PET with FDG-labeled leukocytes versus scintigraphy with 111In-oxine-labeled leukocytes for detection of infection. Radiology 2006;238(3):978–987.
29. Chacko TK, Zhuang H, Stevenson K, Moussavian B, Alavi A: The importance of the location of fluorodeoxyglucose uptake in periprosthetic infection in painful hip prostheses. Nucl Med Commun 2002;23(9):851–855.
30. Stumpe KD, Strobel K: 18F FDG-PET imaging in musculoskeletal infection. Q J Nucl Med Mol Imaging 2006;50(2):131–142.
31. Love C, Tomas MB, Tronco GG, Palestro CJ: FDG PET of infection and inflammation. Radiographics 2005;25(5):1357–1368.
32. Schiesser M, Stumpe KD, Trentz O, Kossmann T, Von Schulthess GK: Detection of metallic implant-associated infections with FDG PET in patients with trauma: Correlation with microbiologic results. Radiology 2003;226(2):391–398.
33. Datz FL: Indium-111-labeled leukocytes for the detection of infection: Current status. Semin Nucl Med 1994;24(2):92–109.
34. Trampuz A, Widmer AF: Infections associated with orthopedic implants. Curr Opin Infect Dis 2006;19(4):349–356.
35. Aggarwal VK, Higuera C, Deirmengian G, Parvizi J, Austin MS: Swab cultures are not as effective as tissue cultures for diagnosis of periprosthetic joint infection. Clin Orthop Relat Res 2013;471(10):3196–3203.
36. Koch R: An address on cholera and its bacillus. Br Med J 1884;2(1236):453–459.
37. Perry JL, Ballou DR, Salyer JL: Inhibitory properties of a swab transport device. J Clin Microbiol 1997;35(12):3367–3368.
38. Wilson ML, Winn W. Laboratory diagnosis of bone, joint, soft-tissue, and skin infections. Clin Infect Dis 2008;46(3):453–457.
39. Gristina AG, Naylor P, Myrvik Q: Infections from biomaterials and implants: A race for the surface. Med Prog Technol 1988-1989;14(3-4):205–224.
40. Sampedro MF, Huddleston PM, Piper KE, et al.: A biofilm approach to detect bacteria on removed spinal implants. Spine (Phila Pa 1976) 2010;35(12):1218–1224.
41. Esteban J, Sandoval E, Cordero-Ampuero J, et al.: Sonication of intramedullary nails: Clinically-related infection and contamination. Open Orthop J 2012;6:255–260.
42. Hannigan GD, Hodkinson BP, McGinnis K, Tyldsley AS, Anari JB, Horan AD, Grice EA, Mehta S:Culture-independent pilot study of microbiota colonizing open fractures and association with severity, mechanism, location, and complication from presentation to early outpatient follow-up. J Orthop Res 2014;32(4):597–605.
43. Ecker DJ, Massire C, Blyn LB, et al.: Molecular genotyping of microbes by multilocus PCR and mass spectrometry: A new tool for hospital infection control and public health surveillance. Methods Mol Biol 2009;551:71–87.
45. Khare R, Espy MJ, Cebelinski E, et al.: Comparative evaluation of two commercial multiplex panels for detection of gastrointestinal pathogens by use of clinical stool specimens. J Clin Microbiol 2014;52(10):3667–3673.