The Journal of Foot & Ankle Surgery xxx (2014) 1–5
Contents lists available at ScienceDirect
The Journal of Foot & Ankle Surgery journal homepage: www.jfas.org
Review Article
Arthroscopic Synovectomy, Irrigation, and Debridement for Treatment of Septic Ankle Arthrosis: A Systematic Review and Case Series Michael R. Mankovecky, DPM 1, Thomas S. Roukis, DPM, PhD, FACFAS 2 1 2
Postgraduate Year III Resident, Gundersen Medical Foundation, La Crosse, WI Attending Staff, Department of Orthopedics, Podiatry and Sports Medicine, Gundersen Health System, La Crosse, WI
a r t i c l e i n f o
a b s t r a c t
Level of Clinical Evidence: 3
Septic arthrosis of the ankle is a rare, often devastating, infection, with a high potential for morbidity and mortality. Delay in treatment can lead to cartilage erosion, painful synovitis, and osteomyelitis. Septic ankle arthrosis deserves prompt recognition and intervention. However, quality, sound, protocol-directed arthroscopic treatment of septic ankle arthrosis of the ankle has not yet been reported. We performed a systematic review of the electronic databases and other relevant peer-reviewed sources to determine the outcomes and treatment protocols associated with septic ankle arthrosis treated with arthroscopic synovectomy, irrigation, and debridement. Nine studies, involving a total of 15 ankles, met our inclusion criteria. In addition, we present the short-term outcomes of a protocol-driven arthroscopic synovial biopsy, deep culture procurement, synovectomy, irrigation, and debridement approach for 8 ankles (8 patients). To our knowledge, this would be the largest individual case series specific to arthroscopic treatment of septic ankle arthrosis. The most common infectious organism reported in the systematic review and in our case series was methicillin-sensitive Staphylococcus aureus. Arthroscopic synovectomy, irrigation, and debridement represents an acceptable treatment method for septic ankle arthrosis and demonstrated outcomes similar to the more traditional open approach, with fewer complications. Additional, appropriately weighted, randomized controlled studies with long-term follow-up are warranted. Ó 2014 by the American College of Foot and Ankle Surgeons. All rights reserved.
Keywords: arthroscopy bacterial arthritis infection pyarthrosis septic arthritis
Septic arthrosis is a common medical emergency, accounting for 0.2% to 0.7% of all hospital admissions and with a mortality rate of 11% (1–4). Pyarthrosis of the ankle has been reported in 10% to 15% of all pyarthrosis cases in adults and 4% to 13% of cases in children (5,6). Septic arthrosis can affect any joint, with the knee the most common site, followed by the hip, shoulder, and wrist (7,8). The published data are replete with septic arthrosis studies treated with arthroscopic debridement of other major joints, such as the hip and knee, with large patient volumes and long-term follow-up (2,5,6,9–19). Arthroscopic surgical decompression follows the time-honored method of surgical treatment consisting of joint evacuation, decompression, elution, and debridement (16). The arthroscopic approach to septic arthrosis has the advantage of shorter hospital stays, quicker recovery, and avoidance of the complications associated with the more traditional open approaches (14,16–18).
Financial Disclosure: None reported. Conflict of Interest: None reported. Address correspondence to: Thomas S. Roukis, DPM, PhD, FACFAS, Department of Orthopaedics, Podiatry and Sports Medicine, Gundersen Health System, Second Floor Founder’s Building, 1900 South Avenue, La Crosse, WI 54601. E-mail address: [email protected] (T.S. Roukis).
Despite the importance of the clinical presentation, high-quality evidence to guide practitioners in the definitive diagnosis of septic arthrosis is lacking (20). The diagnosis of ankle pyarthrosis can often be difficult and delayed owing to the multiple confounding symptoms, rarity of ankle joint involvement, and comorbidities that are often present (1,7). The common clinical signs and symptoms often reported have included pain without an inciting event, new joint edema, fever, rigors, and diaphoresis (2). Investigations of suspected pyarthrosis have included prompt microscopic analysis and culture of the synovial fluid. Culture has been more sensitive than microscopy alone, because synovial fluid gram staining will be positive in only 50% of the cases (1,7,11). Patients with a synovial white blood cell count of 2
IV
MSSA
90
65 y
1
III
MSSA
2
F
33 y
1
NR
1
F
72 y
1
NR
3
NR
NR
1 (I), 2 (II and I)
2-(I) 1-(II)
Mycobacterium xenopi Mycobacterium tuberculosis 1, MSSA; 2, NR
2-7, 1-14
5
NR
NR
1 (I), >2 (II and III)
NR
NR
1
NR
NR
3 (I), 1 (II), 1 (III) NR
NR
5
1, TAR; 4, no recurrence No recurrence
1
F
68 y
III
Lactococcus
7
No recurrence
Investigator [EBM]
Ankles (n)
Gender
Age
Hagino et al (22), 2011 [IV] Shadrick et al (4), 2011 [IV] Nagaya-Waguri et al (15), 2007 [IV] Salliot et al (12), 2006 [III] Titov et al (14), 2004 [III] Seara et al (18), 2002 [III]
1
M
3 mo
1
M
39 y
1
M
1
Stuz et al (17), 2000 [III] Parisien et al (16), 1992 [III] Boffeli et al (26), 2013 [IV]
Arthroscopic Debridements (n; stage) 1
1 2 (III)
Symptom Duration (d)
60 128
Outcome
Examination Findings
Culture Proven
No recurrence
Edema, calor, WBC count 14,000 Dolor, edema, ESR 60
Yes
No recurrence, converted to open No recurrence No recurrence; arthrodesis No recurrence; TTC arthrodesis No recurrence
Dolor, erythema, edema Dolor, erythema, edema, calor NR
Yes Yes Yes Yes 2, yes 1, no
Febrile, leukocytosis, increased ESR, dolor, erythema NR
NR
NR
NR
Pain, febrile, WBC count, 11.3; ESR, 111; CRP, 60.1
Yes
Abbreviations: CRP, C-reactive protein (mg/dL); EBM, evidence-based medicine (level); ESR, erythrocyte sedimentation rate (mm/h); F, female; M, male; MSSA, methicillin-sensitive Staphylococcus aureus; NR, not reported; TAR, total ankle replacement; TTC, tibiotalocalcaneal; WBC, white blood cell (3.8–10.8 103/mL). * G€ achter criteria for septic staging used (25).
M.R. Mankovecky, T.S. Roukis / The Journal of Foot & Ankle Surgery xxx (2014) 1–5
3
Fig. (A) Initial arthroscopic appearance of ankle joint with copious amounts of acute synovitis viewed from anteromedial portal. (B) Arthroscopic-directed procurement of intra-articular cultures from anteromedial portal. (C) Arthroscopic synovial tissue specimen procurement viewed from anteromedial portal. (D) Intraoperative photograph of synovial tissue sample obtained for culture and histopathologic analysis. (E) Arthroscopic synovectomy and debridement of acute synovitis viewed from anteromedial portal. (F) Arthroscopic image after synovectomy, irrigation, and debridement viewed from anteromedial portal.
agent in the studies that reported it was methicillin-sensitive Staphylococcus aureus (4 of 15). The mean interval before presentation was 32 days. No recurrence was reported, 1 case ended in total ankle arthroplasty and 2 in tibiotalar arthrodesis. Five stage I (33.3%), three stage II (20%), three stage III (20%), and one stage IV (6.7%) were reported in the studies that included staging. Arthroscopic debridement was required only once in 7 (46.7%) of the septic ankles (2 stage I, 1 stage II, and 1 stage III). Additional procedures were required in 6 ankles (40%), with 1 requiring open arthrotomy. Three studies did not include the stage. Retrospective Chart Review Results Of the patients treated at our facility, 8 patients (8 ankles) were included in the present study. The weighted mean age was 54 years (range 17 to 86), including 3 men and 5 women. The most common organism identified was methicillin-sensitive S. aureus (3 of 8). All cases resolved without recurrence after protocol-driven arthroscopic Table 2 Joint sepsis staging Stage*
Intraoperative Findings
I II III IV
Synovitis, cloudy effusion Purulent fluid, hypertrophy of hemorrhagic synovial tissue Villous synovitis, adhesions with multiple pouches Osseous erosions chondrolysis
*
G€ aechter criteria for septic joint staging used (25).
synovectomy and debridement, as described. One treatment culminated in planned arthroscopic tibiotalar arthrodesis. The mean white blood cell count, erythrocyte sedimentation rate, and C-reactive protein at presentation was 12,230 cells/mL, 56.8 mm/hr, and 11.2 mg/ dL, respectively. The interval before presentation was 13 days (range 3 to 30). The source of the infection could not be determined in 4 patients (50%). Iatrogenic infections occurred in 2 patients after elective ankle arthroscopy (25%). One patient developed polyarticular sepsis that included contralateral septic wrist involvement. One patient had sustained a cat scratch puncture wound to her medial ankle joint, leading to infection. All but 1 patient had received intravenous antibiotics through a peripherally inserted central catheter, with the therapy guided by the culture findings and sensitivities, with close follow-up and monitoring by our infectious disease specialists. Discussion Regardless of the approach, the goal of treating septic arthrosis includes the removal of all inflammatory cells, enzymes, debris, and foreign bodies, elimination of destructive pannus, and functional recovery. Surgical evacuation of an ankle joint with extensile incisions, continuous irrigation, and daily needle aspiration have been evaluated and discussed in many published studies (3,5,17,27–30). However, these studies have been inconclusive regarding their recommendations. Advocates for early surgical lavage have noted decreased morbidity with early evacuation. Arthroscopy will allow for a direct magnified view of the intra-articular anatomy, joint lavage,
4
M.R. Mankovecky, T.S. Roukis / The Journal of Foot & Ankle Surgery xxx (2014) 1–5
Table 3 Case series of patients treated at our facility Pt. No.
Gender
Age (y)
Arthroscopic Stage* Infectious Agent Interval to LRS Irrigationy Antibiotic Debridements (n) Presentation (d) (L) Therapy
l
F
34
1
I
MSSA
8
9, plain; 6, Bacitracin
2
M
77
1
IV
MSSA
14
3
M
32
1
I
4
F
66
1
IV
30 Staphylococcus lutetiensis/ Staphylococcus bovis Staphylococcus 12 epidermidis
15, plain; 6, Bacitracin 5, plain; 6, Bacitracin
5
M
56
1
I
MSSA
12
6
F
61
1
II
Pasteurella multocida
16
7
F
17
1
III
MSSA
3
60, plain; 6, Bacitracin
8
F
86
1
II
Beta-hemolytic group G strepoccocus
5
30, plain; 6, Bacitracin
Total Male, 3 54 (17–86) 8 Female, 5
3, 2, 1, 2,
I II III IV
d
13 (3–30)
18, plain; 6, Bacitracin
10, plain; 6, Bacitracin 15, plain; 6, Bacitracin
21, plain (9–60); 6, Bacitracin
WBC/ESR/CRP Outcome
Clindamycin 6 d, 9.1/29/2.1 doxycycline 2 wk Ceftriaxone 9.63/114/23 6 wk Ceftriaxone 21 d, 5.89/5/0.2 Augmentin 20 d
Source
No recurrence
Iatrogenic
No recurrence
Unknown
No recurrence
Unknown
PMMA antibiotic 4.8/29/9.5 beads, IV vancomycin 6 wk Cefazolin 3 wk 17.9/75/12.2
Planned ankle Unknown arthrodesis with external fixation No recurrence Unknown
IV ertapenem 8.76/48/6.8 1 g Q24h, 4 wk
No recurrence. Cat scratch arthrofibrosis puncture wound No recurrence Iatrogenic
9.3/55/7.1 IV ceftriaxone 2 g Q24h, 4 wk, cefadroxil 1000 mg PO, BID, 2 wk 33/99/28.5 IV ceftriaxone 2 g Q24h, 4 wk d
12.23/56.8/ 11.2
No recurrence
Immunocompromised polyarticular sepsis
d
d
Abbreviations: BID, twice daily; CRP, C-reactive protein (mg/dL); ESR, erythrocyte sedimentation rate (mm/h); F, female; IV, intravenous; M, male; MSSA, methicillin-sensitive Staphylococcus aureus; PMMA, poly (methyl methacrylate) with 1.2 g tobramycin and 160 mg gentamycin; PO, orally; Q24h, every 24 hours; TTC, tibiotalocalcaneal; WBC, white blood cell (count) (3.8–10.8 103/mL). * G€ achter criteria for septic staging used (25). y Bacitracin, 50,000 U/3 L.
synovectomy of the septic pannus formation without the morbidity of extensile surgical incisions, quicker recovery, and a lower incidence of iatrogenic injury (9,29,31,32). With the morbidity of septic arthrosis at 31%, we believe that arthroscopy of the ankle can play a significant role in early treatment of septic arthrosis (19). Clear, high-level evidencebased medicine has not yet been reported regarding arthroscopic treatment of septic ankle arthrosis (28). In a systematic review conducted by Glazebrook et al (33), only a grade C recommendation (i.e., lacking direct evidence) was given for the arthroscopic treatment of septic ankles from the data reported by Stuz et al (17) and Seara et al (18). The published data have been only slightly more robust for arthroscopic treatment of septic arthrosis involving other joints, such as the knee and hip. Many of these studies were retrospective observational case series (20). Many results have shown greater morbidity with medical management and joint aspiration alone, although no statistical significance could be reached (3,17,18,20,27,32–36). Additionally, very little information is available regarding the medical and surgical management of pyarthrosis of the ankle. However, considering the success of arthroscopic treatment in other joints, we can infer similar outcomes in relationship to the ankle. When diagnosed appropriately and prompt protocol-driven arthroscopic treatment has been activated, such as we have reported, many of the sequela associated with septic arthrosis can be avoided. Our specific technique included immediate procurement of 3 sets of deep aerobic and anaerobic cultures on entry into the ankle joint, biopsy of the synovial tissues, synovectomy under high-volume antibiotic-impregnated irrigation, and thorough debridement of the
anterior, medial, and lateral aspects of the ankle joint. This approach has provided excellent results, with no recurrence of pyarthrosis, regardless of the stage, after a single debridement in our 8 patients. Intraoperative staging offers guidance on the extent of debridement required to remove all infectious material, eliminating the need for repeat procedures. The most common organism found in our case series was methicillin-sensitive S. aureus, a finding similar to that of other studies (4,7,12,14–18,24,37). Our results are similar to those previously reported, with 100% elimination of the infectious process. Arthroscopic synovectomy, irrigation, and debridement, with concomitant intravenous antibiotics, is an acceptable treatment method for septic ankle arthrosis, demonstrating similar outcomes to the more traditional open approach but with fewer complications. High-volume irrigation has been noted to be consistent with 10 to 15 L of sterile fluid (38). In our series of 8 patients, a mean of 21 L (range 9 to 60) was used, with the stage of infection and extent of debridement required the primary guides for the amount of fluid used. The incidence of arthroscopic complications of the ankle has been reported at 7.7%, most commonly transient cutaneous nerve injury and superficial infection (23); however, we did not encounter these problems. The present study had some weaknesses, including that the search for potential references for inclusion was performed predominantly through electronic databases. Although relevant peer-reviewed journals were also manually searched, it is possible that not all pertinent references were identified. An expanded attempt at searching multiple electronic databases might have yielded additional studies for inclusion. In addition, the included studies were all retrospective,
M.R. Mankovecky, T.S. Roukis / The Journal of Foot & Ankle Surgery xxx (2014) 1–5
with 5 level III and 4 level IV evidence-based, medicine-level studies. Additionally, the operative techniques were not standardized in the included studies, and perhaps this lack of homogenicity could have altered the outcomes. Finally, the included sample size from the systematic review of only 15 ankles and patients was small, with the largest individual study sample size including only 5 ankles. However, larger sample sizes, with long-term follow-up data, have not been reported. Our present description of 8 septic ankles treated with protocoldriven arthroscopic synovectomy, irrigation, and debridement is therefore the largest of its type. However, we acknowledge that the retrospective portion of the present study had weaknesses, including its design. The clinical notes and operative reports were used for information regarding the patients postoperatively, and electronic records are rarely robust. In addition, patients might have been treated at other institutions after completing their care with us and that subsequent treatment information would be unavailable to us. However, this was unlikely because our practice is a closed tertiary referral system, and few providers treat ankle septic arthrosis within the geographic area. Also, any patient seen outside our section, but within the institution, would have had electronic medical records reflecting their care and follow-up that would have been easily accessed by us. The relatively short follow-up at completion of our report was an additional weakness. However, given the organisms cultured, we would anticipate that recurrence of septic ankle arthrosis could be expected to occur in the short term and not years later. We routinely consulted with our infectious disease specialists to assist in determining the duration of antibiotic therapy. The duration varied in the present study according to a number of factors. However, in a systematic review and meta-analysis by Stengel et al (39), little highquality evidence for guiding antibiotic therapy for the treatment of septic arthrosis was found. In conclusion, we have performed a systematic review of published reports of septic ankle arthrosis treated with arthroscopic synovectomy and debridement. Elimination of the infection occurred in all reported cases. Studies reported the need for repeat procedures when a higher stage of infection was encountered. We did not find the need for repeat procedures when adequate arthroscopic synovectomy, irrigation, and debridement had been performed during the index procedure. We have also provided a retrospective analysis of 8 culture-proven septic ankles treated at our facility. The most common infectious organism reported in the systematic review and in our case series was methicillin-sensitive S. aureus. Our experience has demonstrated that with a protocol-driven approach and without regard for stage, a single arthroscopic procedure was required in all 8 cases. Our results are similar to those reported regarding the cure rate, complications, and antibiotic management. To our knowledge, this is the largest case series reporting the treatment of septic ankle joint arthrosis alone. To truly validate our findings multiple, appropriately weighted, well-designed, prospective comparative studies are warranted. The results of our systematic review and case series provide a rationale for a protocol-driven guide for arthroscopic treatment of septic ankle arthrosis and could, in the future, provide information to assist with prospective cohort studies and randomized controlled trials.
References 1. Holtom PD, Lawrence B, Zalavras CG. Hematogenous septic ankle arthritis. Clin Orthop Relat Res 466:1388–1391, 2008. 2. Gupta MN, Sturrock RD, Field M. A prospective 2-year study of 75 patients with adult-onset septic arthritis. Br J Rheumatol 40:24–30, 2001. 3. Manadan AM, Block JA. Daily needle aspiration versus surgical lavage for the treatment of bacterial septic arthritis in adults. Am J Ther 11:412–415, 2004.
5
4. Shadrick D, Mendicinio RW, Catanzariti AR. Ankle joint sepsis with subsequent osteomyelitis in an adult patient. J Foot Ankle Surg 50:354–360, 2011. 5. Kang SN, Sanghera T, Mangwani J, Paterson JM, Ramachandran M. The management of septic arthritis in children. J Bone Joint Surg Br 91-B:1127–1133, 2009. 6. Esterhai JL Jr, Gleb I. Adult septic arthritis. Orthop Clin North Am 22:503–514, 1991. 7. Lee CH, Chen YJ, Ung SWN, Hsu RWH. Septic arthritis of the ankle joint. Chang Gung Med J 23:420–426, 2000. 8. LeDantac L, Maury F, Flipo R, Laskri S, Cortet B, Duquesnoy B, Delcabre B. Peripheral pyogenic arthritis: a study of one hundred seventy-nine cases. Rev Rheum 63:103–110, 1996. €der M, Castro WH. Septic arthritis: arthroscopic 9. Jerosch J, Hoffstetter I, Schro management with local antibiotic treatment. Acta Orthop Belg 61:127–134, 1995. 10. Richard JC, Vilian R. Acute septic arthritis of the fingers: a clinical study of 87 cases. Ann Chir Main 1:214–220, 1982. 11. Kaandorp CJ, Krijnen P, Moens HJ, Habbema JD, van Schaardenburg D. The outcome of bacterial arthritis. A prospective community‐based study. Arthritis Rheum 40:884–892, 1997. 12. Salliot C, Desplaces N, Boisrenoult P, Koeger AC, Beaufils P, Vincent V, Mamoudy P, Ziza JM. Arthritis due to Mycobacterium xenopi: a retrospective study of 7 cases in France. Clin Infect Dis 43:987–993, 2006. 13. Smith MJ. Arthroscopic treatment of the septic knee. Arthroscopy 2:30–34, 1986. 14. Titov AG, Nakonechniy GD, Santavirta S, Serodobintzev MS, Mazurenko SI, Konttinen YT. Arthroscopic operations in joint tuberculosis. Knee 1:57–62, 2004. 15. Nagaya-Waguri Y, Goto H, Kobayashi M, Hisazaki S, Otsuka T, Tsuchiya A. Septic arthritis of the right ankle caused by Staphylococcus aureus infection in a rheumatoid arthritis patient treated with etanercept. Mod Rheumatol 17:338–340, 2007. 16. Parisien JS, Shaffer B. Arthroscopic management of pyarthrosis. Clin Orthop 275:243–247, 1982. € ck F, G€ 17. Stuz G, Kuster MS, Kleinstu achter A. Arthroscopic management of septic arthritis: stages of infection and results. Knee Surg Sports Traumatol Arthosc 8:270–274, 2000. 18. Seara JV, Barthel T, Schmitz H, Eulert J. Arthroscopic treatment of septic joints: prognostic factors. Arch Orthop Trauma Surg 122:204–211, 2002. 19. Weston VC, Jones AC, Bradbury N, Fawthrop F, Dohtery M. Clinical features and outcomes of septic arthritis in a single UK health district 1982–1991. Ann Rheum Dis 58:214–219, 1999. 20. Mathews CJ, Kingsley G, Field M, Jones A, Weston VC, Phillips M, Walker D, Coakley G. Management of septic arthritis: a systematic review. Ann Rheum Dis 66:440–445, 2007. 21. Coutlakis PJ, Roberts WN, Wise CM. Another look at synovial fluid leukocytosis and infection. J Clin Rheumatol 8:67–71, 2002. 22. Hagino T, Wako M, Ochial S. Arthroscopic washout of the ankle for septic arthritis in a three-month-old boy. Sports Med Arthroscop Rehab Ther Tech 3:21, 2011. 23. Deng DF, Hamilton GA, Lee M, Rush S, Ford LA, Patel S. Complications associated with foot and ankle arthroscopy. J Foot Surg 51:281–284, 2012. 24. Valentin A, Winge S, Stark A, Eriksson E. Late follow-up results of operative ankle arthroscopy in patients under local anesthesia. Knee Surg Sports Traumatol Arthosc 2:250–254, 1994. €echter A. Der Gelenkinfekt. Inform Artz 6:35–43, 1986. 25. Ga 26. Boffeli TJ, Thompson JC. Arthroscopic management of septic ankle joint: case report of a stage-guided treatment. J Foot Ankle Surg 52:113–117, 2013. 27. Goldenberg DL, Brandt KD, Cohen AS, Cathcart ES. Treatment of septic arthritis: comparison of needle aspiration and surgery as initial modes of joint drainage. Arthritis Rheum 18:83–89, 1975. 28. Field M. Optimum therapy in septic arthritis: to cut or not to cut? Nat Rev Rheumatol 6:8–10, 2010. 29. Broy SB, Stulber DS, Schmid FR. The role of arthroscopy in the diagnosis and management of the septic joint. Clin Rheum Dis 12:489–500, 1986. 30. Renner JB, Agee MW. Treatment of suppurative arthritis by percutaneous catheter drainage. Am J Radiol 154:135–138, 1990. 31. Ayral X. Arthroscopy and joint lavage. Best Pract Res Clin Rheumatol 19:401–415, 2005. 32. Morgan DS, Fisher D, Merianos A, Currie BJ. An 18 year clinical review of septic arthritis from tropical Australia. Epidemiol Infect 117:423–428, 1996. 33. Glazebrook MA, Ganapathy V, Bridge MA, Stone JW, Allard J. Evidence-based indications for ankle arthroscopy. Arthroscopy 25:1478–1490, 2009. 34. Wirtz DC, Marth M, MIltner O, Schneider U, Zilkens KW. Septic arthritis of the knee in adults: treatment by arthroscopy or arthrotomy. Int Orthop 25:239–241, 2001. 35. Thiery JA. Arthroscopic drainage in septic arthritides of the knee: a multicenter study. Arthroscopy 5:65–69, 1989. 36. Schurman DJ, Smith LR. Surgical approach to the management of septic arthritis. Orthop Rev 4:241–245, 1987. 37. Peters RHJ, Rasker JJ, Jacobs JWG, Prevo RL, Karthaus RP. Bacterial arthritis in a district hospital. Clin Rheumatol 11:351–355, 1992. 38. Kirchhoff C, Braunstein V, Paul J, Imhoff A, Hinterwimmer S. Septic arthritis as a severe complication of elective arthroscopy: clinical management strategies. Patient Safety Surg 3:6, 2009. 39. Stengel D, Bauwens K, Sehouli J, Ekkernkamp A, Porzsolt F. Systematic review and meta-analysis of antibiotic therapy for bone and joint infections. Lancet 1:175– 188, 2001.
Contents lists available at ScienceDirect
The Journal of Foot & Ankle Surgery journal homepage: www.jfas.org
Review Article
Arthroscopic Synovectomy, Irrigation, and Debridement for Treatment of Septic Ankle Arthrosis: A Systematic Review and Case Series Michael R. Mankovecky, DPM 1, Thomas S. Roukis, DPM, PhD, FACFAS 2 1 2
Postgraduate Year III Resident, Gundersen Medical Foundation, La Crosse, WI Attending Staff, Department of Orthopedics, Podiatry and Sports Medicine, Gundersen Health System, La Crosse, WI
a r t i c l e i n f o
a b s t r a c t
Level of Clinical Evidence: 3
Septic arthrosis of the ankle is a rare, often devastating, infection, with a high potential for morbidity and mortality. Delay in treatment can lead to cartilage erosion, painful synovitis, and osteomyelitis. Septic ankle arthrosis deserves prompt recognition and intervention. However, quality, sound, protocol-directed arthroscopic treatment of septic ankle arthrosis of the ankle has not yet been reported. We performed a systematic review of the electronic databases and other relevant peer-reviewed sources to determine the outcomes and treatment protocols associated with septic ankle arthrosis treated with arthroscopic synovectomy, irrigation, and debridement. Nine studies, involving a total of 15 ankles, met our inclusion criteria. In addition, we present the short-term outcomes of a protocol-driven arthroscopic synovial biopsy, deep culture procurement, synovectomy, irrigation, and debridement approach for 8 ankles (8 patients). To our knowledge, this would be the largest individual case series specific to arthroscopic treatment of septic ankle arthrosis. The most common infectious organism reported in the systematic review and in our case series was methicillin-sensitive Staphylococcus aureus. Arthroscopic synovectomy, irrigation, and debridement represents an acceptable treatment method for septic ankle arthrosis and demonstrated outcomes similar to the more traditional open approach, with fewer complications. Additional, appropriately weighted, randomized controlled studies with long-term follow-up are warranted. Ó 2014 by the American College of Foot and Ankle Surgeons. All rights reserved.
Keywords: arthroscopy bacterial arthritis infection pyarthrosis septic arthritis
Septic arthrosis is a common medical emergency, accounting for 0.2% to 0.7% of all hospital admissions and with a mortality rate of 11% (1–4). Pyarthrosis of the ankle has been reported in 10% to 15% of all pyarthrosis cases in adults and 4% to 13% of cases in children (5,6). Septic arthrosis can affect any joint, with the knee the most common site, followed by the hip, shoulder, and wrist (7,8). The published data are replete with septic arthrosis studies treated with arthroscopic debridement of other major joints, such as the hip and knee, with large patient volumes and long-term follow-up (2,5,6,9–19). Arthroscopic surgical decompression follows the time-honored method of surgical treatment consisting of joint evacuation, decompression, elution, and debridement (16). The arthroscopic approach to septic arthrosis has the advantage of shorter hospital stays, quicker recovery, and avoidance of the complications associated with the more traditional open approaches (14,16–18).
Financial Disclosure: None reported. Conflict of Interest: None reported. Address correspondence to: Thomas S. Roukis, DPM, PhD, FACFAS, Department of Orthopaedics, Podiatry and Sports Medicine, Gundersen Health System, Second Floor Founder’s Building, 1900 South Avenue, La Crosse, WI 54601. E-mail address: [email protected] (T.S. Roukis).
Despite the importance of the clinical presentation, high-quality evidence to guide practitioners in the definitive diagnosis of septic arthrosis is lacking (20). The diagnosis of ankle pyarthrosis can often be difficult and delayed owing to the multiple confounding symptoms, rarity of ankle joint involvement, and comorbidities that are often present (1,7). The common clinical signs and symptoms often reported have included pain without an inciting event, new joint edema, fever, rigors, and diaphoresis (2). Investigations of suspected pyarthrosis have included prompt microscopic analysis and culture of the synovial fluid. Culture has been more sensitive than microscopy alone, because synovial fluid gram staining will be positive in only 50% of the cases (1,7,11). Patients with a synovial white blood cell count of 2
IV
MSSA
90
65 y
1
III
MSSA
2
F
33 y
1
NR
1
F
72 y
1
NR
3
NR
NR
1 (I), 2 (II and I)
2-(I) 1-(II)
Mycobacterium xenopi Mycobacterium tuberculosis 1, MSSA; 2, NR
2-7, 1-14
5
NR
NR
1 (I), >2 (II and III)
NR
NR
1
NR
NR
3 (I), 1 (II), 1 (III) NR
NR
5
1, TAR; 4, no recurrence No recurrence
1
F
68 y
III
Lactococcus
7
No recurrence
Investigator [EBM]
Ankles (n)
Gender
Age
Hagino et al (22), 2011 [IV] Shadrick et al (4), 2011 [IV] Nagaya-Waguri et al (15), 2007 [IV] Salliot et al (12), 2006 [III] Titov et al (14), 2004 [III] Seara et al (18), 2002 [III]
1
M
3 mo
1
M
39 y
1
M
1
Stuz et al (17), 2000 [III] Parisien et al (16), 1992 [III] Boffeli et al (26), 2013 [IV]
Arthroscopic Debridements (n; stage) 1
1 2 (III)
Symptom Duration (d)
60 128
Outcome
Examination Findings
Culture Proven
No recurrence
Edema, calor, WBC count 14,000 Dolor, edema, ESR 60
Yes
No recurrence, converted to open No recurrence No recurrence; arthrodesis No recurrence; TTC arthrodesis No recurrence
Dolor, erythema, edema Dolor, erythema, edema, calor NR
Yes Yes Yes Yes 2, yes 1, no
Febrile, leukocytosis, increased ESR, dolor, erythema NR
NR
NR
NR
Pain, febrile, WBC count, 11.3; ESR, 111; CRP, 60.1
Yes
Abbreviations: CRP, C-reactive protein (mg/dL); EBM, evidence-based medicine (level); ESR, erythrocyte sedimentation rate (mm/h); F, female; M, male; MSSA, methicillin-sensitive Staphylococcus aureus; NR, not reported; TAR, total ankle replacement; TTC, tibiotalocalcaneal; WBC, white blood cell (3.8–10.8 103/mL). * G€ achter criteria for septic staging used (25).
M.R. Mankovecky, T.S. Roukis / The Journal of Foot & Ankle Surgery xxx (2014) 1–5
3
Fig. (A) Initial arthroscopic appearance of ankle joint with copious amounts of acute synovitis viewed from anteromedial portal. (B) Arthroscopic-directed procurement of intra-articular cultures from anteromedial portal. (C) Arthroscopic synovial tissue specimen procurement viewed from anteromedial portal. (D) Intraoperative photograph of synovial tissue sample obtained for culture and histopathologic analysis. (E) Arthroscopic synovectomy and debridement of acute synovitis viewed from anteromedial portal. (F) Arthroscopic image after synovectomy, irrigation, and debridement viewed from anteromedial portal.
agent in the studies that reported it was methicillin-sensitive Staphylococcus aureus (4 of 15). The mean interval before presentation was 32 days. No recurrence was reported, 1 case ended in total ankle arthroplasty and 2 in tibiotalar arthrodesis. Five stage I (33.3%), three stage II (20%), three stage III (20%), and one stage IV (6.7%) were reported in the studies that included staging. Arthroscopic debridement was required only once in 7 (46.7%) of the septic ankles (2 stage I, 1 stage II, and 1 stage III). Additional procedures were required in 6 ankles (40%), with 1 requiring open arthrotomy. Three studies did not include the stage. Retrospective Chart Review Results Of the patients treated at our facility, 8 patients (8 ankles) were included in the present study. The weighted mean age was 54 years (range 17 to 86), including 3 men and 5 women. The most common organism identified was methicillin-sensitive S. aureus (3 of 8). All cases resolved without recurrence after protocol-driven arthroscopic Table 2 Joint sepsis staging Stage*
Intraoperative Findings
I II III IV
Synovitis, cloudy effusion Purulent fluid, hypertrophy of hemorrhagic synovial tissue Villous synovitis, adhesions with multiple pouches Osseous erosions chondrolysis
*
G€ aechter criteria for septic joint staging used (25).
synovectomy and debridement, as described. One treatment culminated in planned arthroscopic tibiotalar arthrodesis. The mean white blood cell count, erythrocyte sedimentation rate, and C-reactive protein at presentation was 12,230 cells/mL, 56.8 mm/hr, and 11.2 mg/ dL, respectively. The interval before presentation was 13 days (range 3 to 30). The source of the infection could not be determined in 4 patients (50%). Iatrogenic infections occurred in 2 patients after elective ankle arthroscopy (25%). One patient developed polyarticular sepsis that included contralateral septic wrist involvement. One patient had sustained a cat scratch puncture wound to her medial ankle joint, leading to infection. All but 1 patient had received intravenous antibiotics through a peripherally inserted central catheter, with the therapy guided by the culture findings and sensitivities, with close follow-up and monitoring by our infectious disease specialists. Discussion Regardless of the approach, the goal of treating septic arthrosis includes the removal of all inflammatory cells, enzymes, debris, and foreign bodies, elimination of destructive pannus, and functional recovery. Surgical evacuation of an ankle joint with extensile incisions, continuous irrigation, and daily needle aspiration have been evaluated and discussed in many published studies (3,5,17,27–30). However, these studies have been inconclusive regarding their recommendations. Advocates for early surgical lavage have noted decreased morbidity with early evacuation. Arthroscopy will allow for a direct magnified view of the intra-articular anatomy, joint lavage,
4
M.R. Mankovecky, T.S. Roukis / The Journal of Foot & Ankle Surgery xxx (2014) 1–5
Table 3 Case series of patients treated at our facility Pt. No.
Gender
Age (y)
Arthroscopic Stage* Infectious Agent Interval to LRS Irrigationy Antibiotic Debridements (n) Presentation (d) (L) Therapy
l
F
34
1
I
MSSA
8
9, plain; 6, Bacitracin
2
M
77
1
IV
MSSA
14
3
M
32
1
I
4
F
66
1
IV
30 Staphylococcus lutetiensis/ Staphylococcus bovis Staphylococcus 12 epidermidis
15, plain; 6, Bacitracin 5, plain; 6, Bacitracin
5
M
56
1
I
MSSA
12
6
F
61
1
II
Pasteurella multocida
16
7
F
17
1
III
MSSA
3
60, plain; 6, Bacitracin
8
F
86
1
II
Beta-hemolytic group G strepoccocus
5
30, plain; 6, Bacitracin
Total Male, 3 54 (17–86) 8 Female, 5
3, 2, 1, 2,
I II III IV
d
13 (3–30)
18, plain; 6, Bacitracin
10, plain; 6, Bacitracin 15, plain; 6, Bacitracin
21, plain (9–60); 6, Bacitracin
WBC/ESR/CRP Outcome
Clindamycin 6 d, 9.1/29/2.1 doxycycline 2 wk Ceftriaxone 9.63/114/23 6 wk Ceftriaxone 21 d, 5.89/5/0.2 Augmentin 20 d
Source
No recurrence
Iatrogenic
No recurrence
Unknown
No recurrence
Unknown
PMMA antibiotic 4.8/29/9.5 beads, IV vancomycin 6 wk Cefazolin 3 wk 17.9/75/12.2
Planned ankle Unknown arthrodesis with external fixation No recurrence Unknown
IV ertapenem 8.76/48/6.8 1 g Q24h, 4 wk
No recurrence. Cat scratch arthrofibrosis puncture wound No recurrence Iatrogenic
9.3/55/7.1 IV ceftriaxone 2 g Q24h, 4 wk, cefadroxil 1000 mg PO, BID, 2 wk 33/99/28.5 IV ceftriaxone 2 g Q24h, 4 wk d
12.23/56.8/ 11.2
No recurrence
Immunocompromised polyarticular sepsis
d
d
Abbreviations: BID, twice daily; CRP, C-reactive protein (mg/dL); ESR, erythrocyte sedimentation rate (mm/h); F, female; IV, intravenous; M, male; MSSA, methicillin-sensitive Staphylococcus aureus; PMMA, poly (methyl methacrylate) with 1.2 g tobramycin and 160 mg gentamycin; PO, orally; Q24h, every 24 hours; TTC, tibiotalocalcaneal; WBC, white blood cell (count) (3.8–10.8 103/mL). * G€ achter criteria for septic staging used (25). y Bacitracin, 50,000 U/3 L.
synovectomy of the septic pannus formation without the morbidity of extensile surgical incisions, quicker recovery, and a lower incidence of iatrogenic injury (9,29,31,32). With the morbidity of septic arthrosis at 31%, we believe that arthroscopy of the ankle can play a significant role in early treatment of septic arthrosis (19). Clear, high-level evidencebased medicine has not yet been reported regarding arthroscopic treatment of septic ankle arthrosis (28). In a systematic review conducted by Glazebrook et al (33), only a grade C recommendation (i.e., lacking direct evidence) was given for the arthroscopic treatment of septic ankles from the data reported by Stuz et al (17) and Seara et al (18). The published data have been only slightly more robust for arthroscopic treatment of septic arthrosis involving other joints, such as the knee and hip. Many of these studies were retrospective observational case series (20). Many results have shown greater morbidity with medical management and joint aspiration alone, although no statistical significance could be reached (3,17,18,20,27,32–36). Additionally, very little information is available regarding the medical and surgical management of pyarthrosis of the ankle. However, considering the success of arthroscopic treatment in other joints, we can infer similar outcomes in relationship to the ankle. When diagnosed appropriately and prompt protocol-driven arthroscopic treatment has been activated, such as we have reported, many of the sequela associated with septic arthrosis can be avoided. Our specific technique included immediate procurement of 3 sets of deep aerobic and anaerobic cultures on entry into the ankle joint, biopsy of the synovial tissues, synovectomy under high-volume antibiotic-impregnated irrigation, and thorough debridement of the
anterior, medial, and lateral aspects of the ankle joint. This approach has provided excellent results, with no recurrence of pyarthrosis, regardless of the stage, after a single debridement in our 8 patients. Intraoperative staging offers guidance on the extent of debridement required to remove all infectious material, eliminating the need for repeat procedures. The most common organism found in our case series was methicillin-sensitive S. aureus, a finding similar to that of other studies (4,7,12,14–18,24,37). Our results are similar to those previously reported, with 100% elimination of the infectious process. Arthroscopic synovectomy, irrigation, and debridement, with concomitant intravenous antibiotics, is an acceptable treatment method for septic ankle arthrosis, demonstrating similar outcomes to the more traditional open approach but with fewer complications. High-volume irrigation has been noted to be consistent with 10 to 15 L of sterile fluid (38). In our series of 8 patients, a mean of 21 L (range 9 to 60) was used, with the stage of infection and extent of debridement required the primary guides for the amount of fluid used. The incidence of arthroscopic complications of the ankle has been reported at 7.7%, most commonly transient cutaneous nerve injury and superficial infection (23); however, we did not encounter these problems. The present study had some weaknesses, including that the search for potential references for inclusion was performed predominantly through electronic databases. Although relevant peer-reviewed journals were also manually searched, it is possible that not all pertinent references were identified. An expanded attempt at searching multiple electronic databases might have yielded additional studies for inclusion. In addition, the included studies were all retrospective,
M.R. Mankovecky, T.S. Roukis / The Journal of Foot & Ankle Surgery xxx (2014) 1–5
with 5 level III and 4 level IV evidence-based, medicine-level studies. Additionally, the operative techniques were not standardized in the included studies, and perhaps this lack of homogenicity could have altered the outcomes. Finally, the included sample size from the systematic review of only 15 ankles and patients was small, with the largest individual study sample size including only 5 ankles. However, larger sample sizes, with long-term follow-up data, have not been reported. Our present description of 8 septic ankles treated with protocoldriven arthroscopic synovectomy, irrigation, and debridement is therefore the largest of its type. However, we acknowledge that the retrospective portion of the present study had weaknesses, including its design. The clinical notes and operative reports were used for information regarding the patients postoperatively, and electronic records are rarely robust. In addition, patients might have been treated at other institutions after completing their care with us and that subsequent treatment information would be unavailable to us. However, this was unlikely because our practice is a closed tertiary referral system, and few providers treat ankle septic arthrosis within the geographic area. Also, any patient seen outside our section, but within the institution, would have had electronic medical records reflecting their care and follow-up that would have been easily accessed by us. The relatively short follow-up at completion of our report was an additional weakness. However, given the organisms cultured, we would anticipate that recurrence of septic ankle arthrosis could be expected to occur in the short term and not years later. We routinely consulted with our infectious disease specialists to assist in determining the duration of antibiotic therapy. The duration varied in the present study according to a number of factors. However, in a systematic review and meta-analysis by Stengel et al (39), little highquality evidence for guiding antibiotic therapy for the treatment of septic arthrosis was found. In conclusion, we have performed a systematic review of published reports of septic ankle arthrosis treated with arthroscopic synovectomy and debridement. Elimination of the infection occurred in all reported cases. Studies reported the need for repeat procedures when a higher stage of infection was encountered. We did not find the need for repeat procedures when adequate arthroscopic synovectomy, irrigation, and debridement had been performed during the index procedure. We have also provided a retrospective analysis of 8 culture-proven septic ankles treated at our facility. The most common infectious organism reported in the systematic review and in our case series was methicillin-sensitive S. aureus. Our experience has demonstrated that with a protocol-driven approach and without regard for stage, a single arthroscopic procedure was required in all 8 cases. Our results are similar to those reported regarding the cure rate, complications, and antibiotic management. To our knowledge, this is the largest case series reporting the treatment of septic ankle joint arthrosis alone. To truly validate our findings multiple, appropriately weighted, well-designed, prospective comparative studies are warranted. The results of our systematic review and case series provide a rationale for a protocol-driven guide for arthroscopic treatment of septic ankle arthrosis and could, in the future, provide information to assist with prospective cohort studies and randomized controlled trials.
References 1. Holtom PD, Lawrence B, Zalavras CG. Hematogenous septic ankle arthritis. Clin Orthop Relat Res 466:1388–1391, 2008. 2. Gupta MN, Sturrock RD, Field M. A prospective 2-year study of 75 patients with adult-onset septic arthritis. Br J Rheumatol 40:24–30, 2001. 3. Manadan AM, Block JA. Daily needle aspiration versus surgical lavage for the treatment of bacterial septic arthritis in adults. Am J Ther 11:412–415, 2004.
5
4. Shadrick D, Mendicinio RW, Catanzariti AR. Ankle joint sepsis with subsequent osteomyelitis in an adult patient. J Foot Ankle Surg 50:354–360, 2011. 5. Kang SN, Sanghera T, Mangwani J, Paterson JM, Ramachandran M. The management of septic arthritis in children. J Bone Joint Surg Br 91-B:1127–1133, 2009. 6. Esterhai JL Jr, Gleb I. Adult septic arthritis. Orthop Clin North Am 22:503–514, 1991. 7. Lee CH, Chen YJ, Ung SWN, Hsu RWH. Septic arthritis of the ankle joint. Chang Gung Med J 23:420–426, 2000. 8. LeDantac L, Maury F, Flipo R, Laskri S, Cortet B, Duquesnoy B, Delcabre B. Peripheral pyogenic arthritis: a study of one hundred seventy-nine cases. Rev Rheum 63:103–110, 1996. €der M, Castro WH. Septic arthritis: arthroscopic 9. Jerosch J, Hoffstetter I, Schro management with local antibiotic treatment. Acta Orthop Belg 61:127–134, 1995. 10. Richard JC, Vilian R. Acute septic arthritis of the fingers: a clinical study of 87 cases. Ann Chir Main 1:214–220, 1982. 11. Kaandorp CJ, Krijnen P, Moens HJ, Habbema JD, van Schaardenburg D. The outcome of bacterial arthritis. A prospective community‐based study. Arthritis Rheum 40:884–892, 1997. 12. Salliot C, Desplaces N, Boisrenoult P, Koeger AC, Beaufils P, Vincent V, Mamoudy P, Ziza JM. Arthritis due to Mycobacterium xenopi: a retrospective study of 7 cases in France. Clin Infect Dis 43:987–993, 2006. 13. Smith MJ. Arthroscopic treatment of the septic knee. Arthroscopy 2:30–34, 1986. 14. Titov AG, Nakonechniy GD, Santavirta S, Serodobintzev MS, Mazurenko SI, Konttinen YT. Arthroscopic operations in joint tuberculosis. Knee 1:57–62, 2004. 15. Nagaya-Waguri Y, Goto H, Kobayashi M, Hisazaki S, Otsuka T, Tsuchiya A. Septic arthritis of the right ankle caused by Staphylococcus aureus infection in a rheumatoid arthritis patient treated with etanercept. Mod Rheumatol 17:338–340, 2007. 16. Parisien JS, Shaffer B. Arthroscopic management of pyarthrosis. Clin Orthop 275:243–247, 1982. € ck F, G€ 17. Stuz G, Kuster MS, Kleinstu achter A. Arthroscopic management of septic arthritis: stages of infection and results. Knee Surg Sports Traumatol Arthosc 8:270–274, 2000. 18. Seara JV, Barthel T, Schmitz H, Eulert J. Arthroscopic treatment of septic joints: prognostic factors. Arch Orthop Trauma Surg 122:204–211, 2002. 19. Weston VC, Jones AC, Bradbury N, Fawthrop F, Dohtery M. Clinical features and outcomes of septic arthritis in a single UK health district 1982–1991. Ann Rheum Dis 58:214–219, 1999. 20. Mathews CJ, Kingsley G, Field M, Jones A, Weston VC, Phillips M, Walker D, Coakley G. Management of septic arthritis: a systematic review. Ann Rheum Dis 66:440–445, 2007. 21. Coutlakis PJ, Roberts WN, Wise CM. Another look at synovial fluid leukocytosis and infection. J Clin Rheumatol 8:67–71, 2002. 22. Hagino T, Wako M, Ochial S. Arthroscopic washout of the ankle for septic arthritis in a three-month-old boy. Sports Med Arthroscop Rehab Ther Tech 3:21, 2011. 23. Deng DF, Hamilton GA, Lee M, Rush S, Ford LA, Patel S. Complications associated with foot and ankle arthroscopy. J Foot Surg 51:281–284, 2012. 24. Valentin A, Winge S, Stark A, Eriksson E. Late follow-up results of operative ankle arthroscopy in patients under local anesthesia. Knee Surg Sports Traumatol Arthosc 2:250–254, 1994. €echter A. Der Gelenkinfekt. Inform Artz 6:35–43, 1986. 25. Ga 26. Boffeli TJ, Thompson JC. Arthroscopic management of septic ankle joint: case report of a stage-guided treatment. J Foot Ankle Surg 52:113–117, 2013. 27. Goldenberg DL, Brandt KD, Cohen AS, Cathcart ES. Treatment of septic arthritis: comparison of needle aspiration and surgery as initial modes of joint drainage. Arthritis Rheum 18:83–89, 1975. 28. Field M. Optimum therapy in septic arthritis: to cut or not to cut? Nat Rev Rheumatol 6:8–10, 2010. 29. Broy SB, Stulber DS, Schmid FR. The role of arthroscopy in the diagnosis and management of the septic joint. Clin Rheum Dis 12:489–500, 1986. 30. Renner JB, Agee MW. Treatment of suppurative arthritis by percutaneous catheter drainage. Am J Radiol 154:135–138, 1990. 31. Ayral X. Arthroscopy and joint lavage. Best Pract Res Clin Rheumatol 19:401–415, 2005. 32. Morgan DS, Fisher D, Merianos A, Currie BJ. An 18 year clinical review of septic arthritis from tropical Australia. Epidemiol Infect 117:423–428, 1996. 33. Glazebrook MA, Ganapathy V, Bridge MA, Stone JW, Allard J. Evidence-based indications for ankle arthroscopy. Arthroscopy 25:1478–1490, 2009. 34. Wirtz DC, Marth M, MIltner O, Schneider U, Zilkens KW. Septic arthritis of the knee in adults: treatment by arthroscopy or arthrotomy. Int Orthop 25:239–241, 2001. 35. Thiery JA. Arthroscopic drainage in septic arthritides of the knee: a multicenter study. Arthroscopy 5:65–69, 1989. 36. Schurman DJ, Smith LR. Surgical approach to the management of septic arthritis. Orthop Rev 4:241–245, 1987. 37. Peters RHJ, Rasker JJ, Jacobs JWG, Prevo RL, Karthaus RP. Bacterial arthritis in a district hospital. Clin Rheumatol 11:351–355, 1992. 38. Kirchhoff C, Braunstein V, Paul J, Imhoff A, Hinterwimmer S. Septic arthritis as a severe complication of elective arthroscopy: clinical management strategies. Patient Safety Surg 3:6, 2009. 39. Stengel D, Bauwens K, Sehouli J, Ekkernkamp A, Porzsolt F. Systematic review and meta-analysis of antibiotic therapy for bone and joint infections. Lancet 1:175– 188, 2001.
