Screws, plates, staples, pins and wires are the hardware that the foot and ankle surgeon uses to fixate fractures, fusions and/or osteotomies. An infection involving hardware may jeopardize the bone healing process and is a precarious situation for both the patient and the surgeon. In some situations, the infection may be easily managed yet it can be limb threatening in other situations. Like any infection, early diagnosis is paramount.
Hardware is necessary to stabilize osseous segments until one achieves complete bone healing, a process that typically takes six to eight weeks. After that, the implants are theoretically unnecessary and one may remove them. Generally, most surgeons will not recommend hardware removal for at least six months after the index operation, mostly for soft tissue and bone remodeling purposes. However, when the hardware becomes infected, most attempts focus on immediate removal but this often depends on whether or not the underlying bone has healed.
The term “infected hardware” is often loosely used in clinical practice but is probably an inappropriate designation. The hardware, being an inanimate object, cannot become infected. Rather, it becomes coated with bacteria and may secondarily infect its associated bone.
Staphylococcus aureus and epidermidis are the most frequent infecting bacteria of orthopedic implants and prostheses, accounting for approximately 50 to 70 percent of infections.1 These bacteria adhere to the implant surface and produce an extracellular glycocalix (slime layer or biofilm) that protects the organism from antibiotics as well as the host immune responses.2 As a result, retained hardware that has been exposed to bacteria may become a nidus for persistent infection if it is not treated or removed.
How To Identify Infected Hardware
Identifying infected hardware may be a challenging task. In some situations, the diagnosis is straightforward. Subtle infections are frequently more difficult to diagnose and require a detailed workup, and the presence of infection may only be presumptive.
A thorough timeline detailing the clinical circumstances before and after hardware implantation is crucial. Hardware may become hematogenously infected so one should investigate any bacteremic episode, such as recent dental procedures, upper respiratory or urinary tract infections.3 In stable joint prosthesis, a Staphylococcus aureus bactermia has been associated with a 34 percent implant infection rate.4 In particular, intravenous drug users may be at risk for a hematogenous spread of infections. Patients who are biologically or pharmacologically immunocompromised are theoretically susceptible as well. A septic joint adjacent to retained hardware may secondarily seed the hardware.
Any information from the patient history that may help identify the infecting agent will help tailor the course of action and treatment program. Though Staphylococcus aureus and epidermidis are most commonly identified, other bacteria (gram negatives, streptococci, enterococci and anaerobes) are common enough and one should consider them.5 Exposure of the implant to bacteria during the index operation is thought to be a leading cause for hardware infection.
Researchers have reported the incidence of infection following clean orthopedic surgery to be as high as 6.5 percent although certain procedures may have a greater risk for postoperative infection.6 Surgeries that involve percutaneous fixation with pins exiting the skin, like one would see with hammertoe surgery, are at risk for pin tract infections. A study involving distal radius fractures treated with percutaneous exposed Kirschner wires versus those buried deep to the skin demonstrated a significantly greater infection rate with percutaneous wires.7
In 1988, Halpern and Trepal identified a low Kirschner wire associated infection rate of 1.8 percent in 40 patients (54 Kirschner wires) undergoing elective forefoot surgery when no preoperative antibiotics were administered. Interestingly, 73 percent of all wires were culture positive for Staphylococcus epidermidis or diphtheroids, but did not produce a clinical infection.8 One may remove percutaneous wires in the office and surgeons should culture the tip of the wire if an infection is suspected. With deep retained plates, one may attempt to aspirate purulence adjacent to the hardware prior to a formal incision and drainage. This may be particularly useful to help diagnose subtle infections.
Large incisions that dehisce in the face of implanted hardware may be at risk for contiguous infection due to disruption of the skin barrier. Patients developing aggressive cellulitis in the early postoperative period may be at risk for seeding the hardware as well. A late infection may occur several months after the index operation and present only with pain.
If one suspects an infection, obtain a white blood cell (WBC) count, erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP). An elevated WBC count is more beneficial with helping to detect acute infections with cellulitis and/or a wound. In late infections, the WBC count may be normal. Both the CRP and ESR may be elevated with inflammatory conditions (i.e, rheumatoid arthrosis) and the absence of their elevation is helpful in excluding infection.
A study of 179 patients after orthopedic surgery with or without implants demonstrated a 12-fold increase in CRP after four weeks postoperatively, a sensitivity of 100 percent, a specificity of 83.6 percent and a negative predictive value of 100 percent for infection.9 In comparison to the ESR, the CRP is more sensitive and specific for implant infection.4 Nonetheless, one should correlate all lab work to the clinical scenario.
Radiographs are necessary to evaluate the bone healing process and help identify any findings consistent with a bone infection. However, keep in mind that it may be difficult, if not impossible, to distinguish osteomyelitic signs radiographically from normal healing, stress fracture or nonunion. Infection can occur concomitantly and accordingly confuse the radiographic findings further. Peri-hardware lucencies are suspicious for an underlying nonunion or may represent septic or aseptic loosening.
Similarly, broken hardware is suspicious for an underlying process. Gas in the tissue is a clear indicator of infection but the organism responsible for producing gas, Clostridia, typically does not infect hardware.10 Periosteal reaction is a common finding with a bone infection but may be difficult to distinguish from normal bone healing callus or stress fracture callus. Primary bone healing should not heal with callus and the presence of a periosteal reaction surrounding a stable well-fixated osteotomy may indicate a loss of hardware stability and/or infection. Clinicians may also confuse radiographic avascular necrosis with osteomyelitis. Accordingly, one should correlate it with the clinical scenario.
Standard nuclear imaging with technetium may be useful when identifying infections more than a year after the index operation because a negative scan strongly suggests the absence of infection. White blood cell labeled nuclear scans with focal uptake strongly suggest infection with joint prostheses. However, one should correlate this study with a technetium scan.11 Magnetic resonance imaging may be useful when patients have titanium hardware.
The presence of bacteria on retained implants without clinical signs of infection is an area of significant investigation.12 It has been postulated that standardized culture techniques are inadequate in identifying bacteria and, in some cases, may leave an infection untreated or unrecognized. Newer techniques, such as immunofluoresence, PCR and sonication, are being developed to identify the presence of bacteria associated with orthopedic implants. However, remember that the presence of bacterial DNA does not necessarily indicate an infectious process and may simply represent bacterial remnants.12
Pertinent Treatment Considerations
Infected hardware is often synonymous with long-term parental antibiotic therapy, typically six to eight weeks. In the absence of a septic patient and if the clinical situation permits, it is preferable not to begin any antibiotics until one has obtained cultures, preferably cultures of deep bone and tissue. Very often, clinicians will preemptively start patients on oral antibiotics before the infection has declared itself. This practice may lead to antibiotic resistance as well as selecting out certain organisms. In general, antimicrobial therapy should target Staphylococcus aureus and epidermidis. If you suspect other bacteria, expand the coverage beyond gram-positive organisms. Once you have obtained preliminary and final culture results, alter the antibiotics as necessary.
Cefazolin (Ancef, GlaxoSmithKline) is the empiric drug of choice and is dosed 1 to 2 g IV q 8 hours. In beta-lactam allergic patients, clindamycin or vancomycin are good alternatives. In areas or institutions with significant methicillin resistant Staphylococcus aureus (MRSA) penetration, consider vancomycin until the cultures demonstrate Staphylococcus with susceptibility to cephalosporins. For aggressive limb-threatening infections, consider empiric treatment with vancomycin as well as expanded coverage for gram negatives and anaerobes.
Treatment of infected hardware does not stop with just long-term antibiotic therapy. One would often perform surgical measures such as incision and drainage, hardware removal and bone debridement in conjunction with the pharmacologic treatment. The amount of bone debridement depends on the index operation, location, radiographic and intraoperative findings. Surgeons should debride necrotic, nonviable bone and pinpoint bleeding is a good indicator of bone viability.
Infected fusion sites often result in bone loss that may require subsequent bone grafting to fill the defect. Surgeons may use an antibiotic methylmethacrylate spacer to provide local antibiotics as well as maintain the dead space until the infection has resolved.
Biomechanical stability is the best predictor for bone healing.13-16 As such, hardware removal often depends on whether the underlying bone has healed and if the hardware is grossly purulent. If the fusion, fracture or osteotomy has healed and the soft tissue envelope is intact, then one should remove the hardware and start the patient on long-term antibiotics. Due to the glycocalyx deposited on the implant, retained hardware may be a nidus for persistent infection and one should remove it if the clinical situation will permit.
It is possible to achieve healing of septic bone by leaving the infected hardware untouched and providing concomitant parental antibiotics. However, consider this on a case by case basis and do not initiate it in patients with an abscess or purulent hardware. Grossly infected implants with purulence typically require removal. The presence of a deep abscess requires an incision and drainage regardless of the underlying healing bone. If the bone has not healed and removing the hardware will result in instability, one should strongly consider removing the hardware and applying an external fixator.
The ability to utilize an external fixator depends on the location of the infection and remaining bone stock. Ritcher, et. al., demonstrated an 86.6 percent septic ankle and/or subtalar fusion rate in 45 patients who were treated with a combination of internal fixation and external fixation.15 In a retrospective study detailing the use of circular wire external fixation of salvage ankle arthrodesis, Rush and Schuberth achieved a fusion or stable nonunion in nine of 12 patients who underwent a septic fusion.17 Other studies also demonstrate the ability to obtain septic fusions.18,19 However, be aware that these cases are difficult and involve a prolonged treatment course with multiple revisions, which may ultimately go on to catastrophic loss.
In cases of severe infection and an osseous defect of a septic ankle joint, Cierny, et. al., recommend a staged reconstruction utilizing antibiotic beads as a temporary spacer along with an Ilizarov external fixation.20 One would often attempt bone grafting once the infection is controlled and/or eradicated. In my opinion, autograft is preferable over allograft for its osteogeneic properties, especially in the face of previous infection and possible repeat nonunion where achieving bone healing may be difficult. In cases of nonunion without bone loss, I still augment the revision site with autologous cancellous bone, which one may obtain from the calcaneus or tibia. Be careful not to seed the donor site when obtaining the graft. The advances in orthobiologics may lessen, if not eliminate, the need for autologous bone grafts in the future.
Retrieving broken hardware may be challenging because deeply embedded implants are sometimes inaccessible. Depending on the type of hardware and location, one may use a large bore cannulated drill to overdrill on top of deeply retained portions of hardware. Surgeons may leave irretrievable hardware in place but should monitor it cautiously for recurrent infection.
When VAC Therapy Can Be A Useful Adjunct
Exposed hardware is considered infected simply because it has been exposed to the environment and bacteria. However, exposed hardware does not necessarily need to be removed as it is possible to form granulation tissue over hardware. Only attempt this if the hardware is still required for bone stabilization and if you can control the infection with parental antibiotics with or without surgical washouts. One may use vacumm assisted closure (VAC® Therapy, KCI) as an intermediate procedure until performing a definitive closure or the wounds heal through secondary intention.13 Pelham, et. al., retrospectively reported the ability of the VAC to form granulation tissue in a variety of clinical situations with exposed hardware (screws, plates, total knee arthroplasty, intramedullary nail and wire) in 10 patients who went on to definitive closure with free or local tissue flaps.21 One should consider a local or free tissue flap for acute traumatic wounds or defects.22
Attinger and Cooper have used the VAC to treat dehisced calcaneal wounds of less than 2 cm after open reduction internal fixation.23 One may retrieve hardware once the underlying bone has healed. Mullner, et. al., removed three of 11 lower extremity implants due to septic loosening after using the VAC to close wounds successfully while the underlying bone was healing.24
When Should You Stop Antibiotic Treatment?
The decision to stop the antibiotics depends on several factors — overall duration, laboratory values, clinical appearance and radiographic findings. Monitor weekly lab tests (CBC & Chem 7) for the response to treatment as well as any untoward effects (such as bone barrow suppression or renal damage) of the antibiotics.
Theoretically, the ESR and CRP should decrease with treatment and may be good indicators of successful treatment.
When bone changes become worse on radiographs, this development is worrisome for persistent infection and the patient may need further treatment or a repeat surgical debridement. A successful septic fusion is a good indicator that the infection has resolved. Most patients require six to eight weeks of antibiotics but some situations require a longer duration. In select cases, suppression antibiotics may be indicated.
Steps You Can Take To Prevent Hardware Infection
Using prophylactic antibiotics prior to clean orthopedic type surgery when implanting hardware is common practice. There should be bactericidal levels of antibiotic in the tissue at the time of incision.25,26 The antibiotic should target gram-positive cocci, specifically Staphylococcus aureus and Staphylococcus epidermidis. Cefazolin or cefuroxime (Ceftin, GlaxoSmithKline) are used most frequently. In penicillin allergic patients, clindamycin or vancomycin are the most likely alternatives.
One should administer the antibiotic within one to two hours of surgery. If using vancomycin, anticipate the proper administration time because of the well-known “red man” reaction from rapid infusion. With regard to primary total joint arthroplasty, a 2004 advisory statement by the American Academy of Orthopedic Surgeons recommends, “antibiotics should be administered within one hour prior to skin incision” and “must be completely infused prior to inflation of the (proximal) tourniquet.”27
Zgonis and Jolly performed a retrospective review of 555 patients and evaluated the efficacy of preoperative antibiotics in elective outpatient foot and ankle surgery.28 Approximately half of the patients in the study received antibiotics, which was associated with a 1.6 percent infection rate.
This was in comparison to a 1.4 percent infection rate among those patients who did not receive antibiotics. There were no infections involving bone or hardware. The authors demonstrated that “preoperative antibiotic use, age, gender, type of surgical procedure, operative time, tourniquet use, past medical history and internal fixation were not predictive of or associated with postoperative wound infection or complication.”28
The study authors suggest that routine use of prophylactic intravenous antibiotics in routine elective foot and ankle surgery is not warranted.28 However, because this study was a retrospective review, antibiotic usage was utilized more frequently when hardware was implanted or external fixation was in use and involved more completed surgical procedures.
Further studies, especially randomized controlled double-blind prospective studies, are necessary to further clarify the role of preoperative antibiotic prophylaxis with clean elective foot and ankle surgery. Due to the potential morbidity (“prolonged and repeat hospitalization, sepsis, persistent pain, device replacement and possible death”) as a consequence of postoperative hardware infection, there remains a bias to continue the use of preoperative antibiotic prophylaxis.26
Moreover, Zgonis and Jolly point out that postoperative infection is largely related to other factors besides preoperative antibiotic administration. These include postoperative hematoma, tissue trauma, surgical inexperience, nonsterile instruments and additional personnel in the operating room.28
The prevailing thinking is that prolonged surgeries increase infection risk and one should accordingly try to minimize operating room time. Obviously, an open surgical wound exposed to the environment for a prolonged period of time is at greater risk for bacterial contamination from airborne bacteria. It is generally thought that surgeries longer than two hours are at risk. Surgical technique may also play a role in postoperative complications and infections. Poor technique and handling of the tissues may result in dehiscence or skin necrosis, which may create a portal for bacterial entry.
Addressing infected hardware is an involved process that requires attention to detail and close monitoring. Adhesion of bacteria to implants makes treating these infections particularly complicated. Both surgical and pharmacologic measures are needed to eradicate the infection.
It is important to identify the infection early and begin aggressive treatment. Utilizing parental antibiotics directed against Staphylococcus aureus and epidermidis is the first line measure to treatment. Surgical debridement with or without hardware removal often depends on whether the underlying bone has healed as well as the aggressiveness of the infection. Infections involving hardware/ implants may be limb-threatening situations.
Dr. Blitz is an attending podiatric surgeon within the Department of Orthopedics at the Kaiser Permanente Medical Center in Santa Rosa, Ca. He is a Fellow of the American College of Foot and Ankle Surgeons.
Editor’s note: For related articles, see “Win The Battle Against Postoperative Infections” in the August 2002 issue, “A Closer Look At Diabetic Foot Infections” in the July 2005 issue, or “Current Concepts In External Fixation” in the April 2003 issue.
Also be sure to check out the archives at www.podiatrytoday.com.
1. De Lalla F. Antibiotic prophylaxis in orthopedic prosthetic surgery. J Chemother. 2001;13 Spec No 1(1):48-53. 2. Stoodley P, Kathju S, Hu FZ, et al. Molecular and imaging techniques for bacterial biofilms in joint arthroplasty infections. Clin Orthop Relat Res. 2005; (437):31-40. 3. Sia IG, Berbari EF, Karchmer AW. Prosthetic joint infections. Infect Dis Clin North Am. 2005; 19(4):885-914. 4. Bernard L, Lubbeke A, Stern R, et al. Value of preoperative investigations in diagnosing prosthetic joint infection: retrospective study and review of the literature review. Scan J Infect Dis 2004; 36:410-6. 5. Steckelberg J, Osmon D. Prosthetic joint infections. In: Waldvogel F, Bisno A, editors. Infections associated with indwelling medical devices. 3rd Edition. Washington: American Society for Microbiology Press; 2000. 6. Zgonis T, Jolly GP, Garbalosa JC. The efficacy of prophylactic intravenous antibiotics in elective foot and ankle surgery. J Foot Ankle Surg. 2004 Mar-Apr;43(2):97-103. 7. Hargreaves DG, Drew SJ, Eckersley R. Kirschner wire pin tract infection rates: a randomized controlled trial between percutaneous and buried wires. J Hand Surg [Br]. 2004; 29(4):374-6. 8. Halpern FP, Trepal MJ, Hodge W. Contamination and infection rate of percutaneous Kirschner wires in foot surgery. J Am Podiatr Med Assoc. 1990; 80(8):433-7. 9. Codine P, Barbotte E, Denis-Laroque F, et al. C-reactive protein, leukocyte count and D-dimer monitoring after orthopedic surgery: early diagnosis of infectious or thromboembolic complications. Part one: C-reactive protein and leukocyte count as an aid in diagnosing postoperative infection. Ann Readapt Med Phys. 2005; 48(8):590-7. 10. Cohen RF, Yourofsky LA. Gas gangrene: a postoperative complication. J Foot Surg. 1980 Winter; 19(4):202-6. 11. Love C, Tomas MB, Marwin SE, et al. Role of nuclear imaging in diagnosis of the infected joint replacement. Radiographics 2001; 21:1229-38. 12. Nelson CL, McLaren AC, McLaren SG, et al. Is aseptic loosening truly aseptic? Clin Orthop Relat Res. 2005; (437):25-30. 13. Blitz NM. Vacuum Assisted Closure in Lower Extremity Reconstruction, In Lower Extremity Soft Tissue & Cutaneous Plastic Surgery, Edited by Dockery GL & Crawford ME, Elsevier Science LTD. Jan 2006. 14. Defranzo AJ, Argenta LC, Marks MW, et al. The use of vacuum-assisted closure therapy for the treatment of lower extremity wounds with exposed bone. Plast Reconstr Surg 2001; 108(5):1184-91. 15. Richter D, Hahn MP, Laun RA, et al. Arthrodesis of the infected ankle and subtalar joint: technique, indications, and results of 45 consecutive cases. J Trauma. 1999; 47(6):1072-8. 16. Muhr G, Kayser M. The infected elbow fractures. Orthopede. 1988; 17:279-286. 17. Zarutsky E, Rush SM, Schuberth JM. The use of circular wire external fixation in the treatment of salvage ankle arthrodesis. J Foot Ankle Surg. 2005 Sep-Oct;44(5):420. 18. Kolling E, Esenwein SA, Muhr G, et al. Fusion of the septic ankle with 15 cases using hybrid external fixation. J Trauma. 2003; 55(4):685-91. 19. Richter D, Hahn MP, Laun RA. Arthrodesis of the infected ankle and subtalar joint: technique, indications and results of 45 consecutive cases. J Trauma. 199; 47(6)1072-8. 20. Cierny G, Zorn KE. Arthrodesis of the tibiotalar joint for sepsis. Foot Ankle Clin. 1996; 177-197. 21. Pelham FR, Kubiak EN, Sathappan SS, et al. Topical negative pressure in the treatment of infected wounds with exposed orthopaedic implants. J Wound Care. 2006;15(3):111-6. 22. Gopal S, Majumder S, Batchelor AG, et al. Fix and flap: the radical orthopaedic and plastic treatment of severe open fractures of the tibia. J Bone Joint Surg Br. 2000; 82(7):959-66. 23. Attinger C, Cooper P. Soft tissue reconstruction for Calcaneal fractures or osteomyelitis. Orthop Clin North Am. 2001; 32(1):135-70. 24. Mullner T, Mrkonjic L, Kwasny O, et al. The use of negative pressure to promote the healing of tissue defects: A clinical trial using the vacuum sealing technique. Br J Plast Surg. 1997; 50(3):194-9. 25. Fitzgerald Jr RH, Thompson RL: Cepalosporin antibiotics in the prevention and treatment of musculoskeletal sepsis. J Bone Joint Surg 1983; 65A:1201-1205. 26. AAOS American Society of Health-System Pharmacists ASHP Therapeutic Guidelines on Antimicrobial Prophylaxis in Surgery. American Society of Health-System Pharmacists. Am J Health Syst Pharm 1999 15; 56(18):1839-88. 27. AAOS Advisory Statement, Recommendations for the Use of Intravenous Antibiotic Prophylaxis in Primary Total Joint Arthroplasty. June 2004. 28. Zgonis T, Jolly GP, Garbalosa JC. The efficacy of prophylactic intravenous antibiotics in elective foot and ankle surgery. J Foot Ankle Surg. 2004; 43(2):97-103.