Outcome of orthopedic implant infections due to different staphylococci

International Journal of Infectious Diseases j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / i j i d Outcome of orthopedic implant infections due to different staphylococci§ Dorota Teterycz Tristan Ferry Daniel Lew , Richard Stern Mathieu Assal Pierre Hoffmeyer Louis Bernard Ilker Uc¸kay a Orthopedic Surgery Service, Geneva University Hospitals and Faculty of Medicine, University of Geneva, 24, Rue Micheli-du-Crest, 1211 Geneva 14, Switzerlandb Service of Infectious Diseases, Geneva University Hospitals and Faculty of Medicine, University of Geneva, Switzerlandc Service of Infectious Diseases, Raymond Poincare´ University Hospital, Garches, Franced Assistance Publique-Hoˆpitaux de Paris, Garches, France Background: Comparisons of different staphylococci in orthopedic implant infections have rarely been reported. In this study we assessed total joint arthroplasty infections and other orthopedic implant infections due to methicillin-sensitive Staphylococcus aureus (MSSA), methicillin-resistant Staphylococcus aureus (MRSA), and coagulase-negative staphylococci (CoNS).
Methods: This was a retrospective study performed at the Geneva University Hospitals for the periodJanuary 1996 to June 2008.
Results: There were 44 infections due to MRSA, 58 due to MSSA, and 61 due to CoNS. Overall cure was achieved in 57% (25/44) of MRSA infections, 72% (42/58) of MSSA infections, and 82% (50/61) of CoNS infections, after a minimum follow-up of 1 year. In the subgroup of arthroplasty infections only, cure was achieved in 39% (7/18) of MRSA, 60% (15/25) of MSSA, and 77% (30/39) of CoNS episodes. In multivariate analysis, arthroplasty (odds ratio (OR) 0.2, 95% confidence interval (95% CI) 0.1–0.6) and MRSA infections (OR 0.3, 95% CI 0.1–0.9) were inversely associated with overall cure for all implants. CoNS infection (OR3.0, 95% CI 1.2–8.0) and the insertion of a new implant (OR 4.5, 95% CI 1.6–13.1) were associated withhigher cure results. Methicillin resistance, immunosuppression, sex, age, duration of antibiotic therapy,one-stage revision, rifampin use, and total number of surgical interventions did not influence cure.
MRSA-infected patients had more post-infection sequelae than patients with MSSA or CoNS (Chi-squaretest 13/44 vs. 93/119, OR 3.4, 95% CI 1.3–8.9, p = 0.004).
Conclusions: In orthopedic implant infections, S. aureus is more virulent than CoNS. MRSA has the worstoutcome and CoNS the best.
ß 2010 International Society for Infectious Diseases. Published by Elsevier Ltd. All rights reserved.
experience suggests that methicillin-resistant infections mighthave more recurrences and more sequelae than methicillin- Staphylococci comprise up to two-thirds of all pathogens in orthopedic implant these infections are difficult to The objective of this study was to assess the clinical features treat because of the ability of the organisms to form small-colony and outcome in patients with orthopedic implant infections (total variantsand to grow into Additionally, foreign material joint arthroplasties and fracture fixation devices) due to the three itself inhibits neutrophil antibacterial activity.
main groups of staphylococci: methicillin-resistant Staphylococcus Methicillin-resistant staphylococcal species may adversely aureus (MRSA), methicillin-sensitive Staphylococcus aureus (MSSA), influence treatment outcome, as has previously been shown for and coagulase-negative staphylococci (CoNS). In a second step, we assessed risk factors for recurrent disease for implant infections Comparative studies regarding the epidemiology and outcomes overall, and stratified by total joint arthroplasties and fracture of localized orthopedic implant-related infections stratified by staphylococci or type of orthopedic implant are rare. Clinical Presented in part as a poster at the Annual Meeting of the Swiss Society for Infectious Diseases, June 2009, Geneva and the Annual Congress of the European The Geneva University Hospitals form a 2200-bed tertiary Society for Surgical Research, June 2010, Geneva.
* Corresponding author. Tel.: +41 22 372 3311; fax: +41 22 372 3987.
hospital with a high MRSA endemicity (30% of all clinical S. aureus sequence type 228 is the predominant MRSA strain.
1201-9712/$36.00 – see front matter ß 2010 International Society for Infectious Diseases. Published by Elsevier Ltd. All rights reserved.
doi: D. Teterycz et al. / International Journal of Infectious Diseases 14 (2010) e913–e918 The Orthopedic Surgical Service has 119 acute care beds and a follow-up by analysis of patient medical records during the last dedicated infectious diseases specialist; 5374 surgical procedures visit to the Geneva University Hospitals, independent of orthopedic were performed here in 2007. This service has run a cohort for total reasons. This passive follow-up had no upper time limit, unless there was general censoring on 30 September 2009, the date ofclosure for data sampling. Recurrence of infection meant new clinical signs of infection with the same microorganism at least 2weeks after the end of treatment for the first episode. A Databases from the Laboratory of Bacteriology, the Geneva pseudarthrosis without proof of the formerly infecting Staphylo- Arthroplasty Registry,the Septic Orthopedic Cohort, and the coccus spp according to study definitions was interpreted as hospital’s administrative coding were retrospectively searched for sequelae, but not as recurrent infection. The duration and staphylococcal infections related to orthopedic implants for the modalities of antibiotic treatment concomitant to surgery were period January 1996 to June 2008. Sixty variables for each episode undertaken according to expert opinion.
were assessed with information pertaining to demographiccharacteristics, microbiology, treatment modalities, and outcomes.
A surgeon and a physician independently recorded each variableon a spreadsheet for analysis. In the case of discordance, a Comparisons of the groups of staphylococcal infections were consensus was obtained by involving a third co-author. Patients performed using the Pearson Chi-square test, Fisher’s exact test, or were followed-up to 30 September 2009. A minimum follow-up the Wilcoxon rank sum test, as appropriate. Logistic regression time of 1 year after the end of treatment was required for study analyses determined associations with cure. Independent variables with a p-value of 0.2 in univariate analysis were added stepwisein the multivariate analysis. The following variables were introduced into the final model independently of their associationin univariate analysis: sex, age, duration of antibiotic treatment, The microbiological procedures were unchanged during the number of surgical interventions, and methicillin resistance. All study period and based on the Clinical and Laboratory Standards variables were checked for confounding, collinearity, and interac- Institute (CLSI) guidelines.In order to enhance specificity, only tion; the latter by Mantel–Haenszel estimates. p-Values of 0.05 cultures that were grown on plates were considered. Staphylococci (two-tailed) were considered significant. STATA software (v. 9.0; were characterized to the species level by slidex agglutination (Pastorex1, Bio-Rad), DNAse tests (homemade), the ID32 Staphy-lococcus Gallery (bioMe´rieux, Marcy l’Etoile, France), and/or the Vitek ID system. The staphylococci were interpreted as the same ifsurrogate markers such as species, staphylococcal chromosomal cassettes (SCC), presence of exfoliatins A and B, Panton–Valentineleukocidin, toxic shock syndrome toxins and agr gene regulator A total of 205 episodes of staphylococcal orthopedic implant genes, and antibiotic susceptibility patterns were identical. No infection were retrieved. Of these, 42 were excluded due to: typing was performed. Since clinical specimens had not been follow-up shorter than 3 months or lost to follow-up (n = 32); stored, no retrospective analyses of minimal inhibitory concentra- substantial co-infection with Pseudomonas aeruginosa (n = 2), tions, e.g., for vancomycin against MRSA, could be performed.
Enterobacter cloacae (n = 2), Escherichia coli (n = 1), Propionibacter-ium acnes and Streptococcus constellatus (n = 1), and Enterococcus faecalis (n = 1); infection of spondylodesis material (n = 2). Oneschizophrenic patient with MSSA infection was excluded because Inclusion criteria were: the presence of an implant; local signs of infection such as heat, erythema, pus, or functional impairment; In the final evaluation, a total of 163 primary surgical site a medical report; a targeted antibiotic treatment; and the presence infections in 157 patients (73 females; median age 69 years, of the same Staphylococcus sp in more than one intraoperative interquartile range (IQR) 50–80 years) underwent further analysis.
The median follow-up time was 2.3 years (IQR 1.1–4.3 years).
Exclusion criteria were: antibiotic medication in the preceding four weeks (to avoid a potential bias by ‘selection’ of methicillin- resistant strains in microbiological samplings that might havebeen pre-treated with beta-lactam antibiotics, thus hiding other The infected implants in the arthroplasty group included total susceptible pathogens); an active follow-up shorter than 3 months hip (n = 52), total knee (n = 29), and total ankle (n = 1). In the after the end of treatment; and infections occurring after spinal fracture fixation devices group, the infected implants included surgery. Co-pathogens were accepted only if the Staphylococcus spp plates/screws (n = 40), intramedullary nails (n = 16), external outnumbered them by at least three-fold in intraoperative fixation (n = 13), hip screws (n = 4), other screws (n = 4), patellar cerclage wire (n = 3), and pins (n = 1).
Since infections of total joint arthroplasties might be different from those following other orthopedic implant procedures, all analyses were repeated for arthroplasty infections and fracturefixation devices separately. Arthroplasties were defined as total There were 44 episodes due to MRSA, 58 due to MSSA, and 61 hip, total knee, and total ankle prostheses. The fracture fixation due to CoNS. Community-acquired MRSA was not encountered.
device group encompassed intramedullary nails, plates/screws, The species of CoNS were Staphylococcus epidermidis (n = 36), screws alone, external fixation, wires, and pins.
Staphylococcus lugdunensis (n = 3), Staphylococcus capitis (n = 2), Cure was defined as complete clinical and microbiological Staphylococcus hominis (n = 2), and one episode each of Staphylo- resolution of the former infection after a minimum follow-up time coccus intermedius, Staphylococcus simulans, Staphylococcus xylosus of 1 year following the end of treatment. This follow-up was active, and Staphylococcus schleiferi. The CoNS were not further identified e.g., regular postoperative controls. There was also a passive to the species level for 14 episodes.
D. Teterycz et al. / International Journal of Infectious Diseases 14 (2010) e913–e918 Table 1Characteristics and comparisons between three groups of staphylococcal orthopedic implant-associated infections All types of implant infections (N = 163) Median time delay between previous implantation and infection onset MRSA, methicillin-resistant Staphylococcus aureus; MSSA, methicillin-sensitive Staphylococcus aureus; CoNS, coagulase-negative staphylococci.
Group comparisons were performed with the Wilcoxon rank sum test, Fisher’s exact test, or the Pearson Chi-square test, as appropriate.
a Only statistically significant p-values of 0.05 (two-tailed) are displayed.
b Diabetes mellitus, transplantation, chronic alcoholism, neoplasia, Child’s class C cirrhosis, AIDS, steroid medication. Polytrauma did not count as chronic immunosuppression.
Eighty-seven staphylococci (53%; MRSA and CoNS) were 3.6. Overall outcomes for all implant infections methicillin-resistant. Three staphylococci (all CoNS) were resistantto rifampin. Five MSSA patients, four MRSA patients, and two Two patients died of septic shock due to MSSA and MRSA, patients with CoNS were co-infected with other pathogens, with respectively. They were not included in the final analysis as they the Staphylococcus spp outnumbering these by at least three-fold in were among the 32 patients excluded because of insufficient microbiological specimens; pathogens included Klebsiella pneu- moniae (n = 2), Proteus mirabilis (n = 2), CoNS (n = 2), and one case Among the remaining 163 infections, cure was achieved in 57% each of Enterococcus faecalis, Streptococcus pyogenes, MRSA, (25/44) of all episodes of MRSA, in 72% (42/58) of MSSA, and in 82% Streptococcus constellatus, and Enterobacter cloacae.
(50/61) of CoNS (These differences were not statisticallysignificant. Forty-three patients (27%, 43/157) had sequelae of 3.4. Clinical presentation of all implant infections former infection, including: Girdlestone hip (n = 8), arthrodesis(n = 8), amputation (n = 4), and other functional handicaps and/or summarizes the differences in clinical presentation incapacitating pain (n = 23). MRSA-infected patients had signifi- among the three groups of staphylococci for all types of implant cantly more sequelae than MSSA or CoNS patients (13/44 vs. 93/ infections. Statistically significant differences included the follow- 119, odds ratio (OR) 3.4, 95% confidence interval (95% CI) 1.3–8.9, ing: MRSA infections had shorter incubation times; patients with MSSA infections were younger than patients with methicillin- Recurrence of infection always occurred locally and was seen in resistant staphylococci; and bacteremia was witnessed only in S.
20 episodes (12%, 20/163) with a median delay of 94 days after the aureus (MRSA and MSSA), but not in CoNS infections.
end of treatment. The number of recurrences were not statisticallydifferent between arthroplasty vs. fracture fixation device infec- tions (12 vs. 8 recurrences, p = 0.36).
The median length of hospital stay for all staphylococcal All patients received systemic antibiotic therapy directed towards infections was 36 days (IQR 16–82 days). Patients with methicillin- the causative pathogen for a median duration of 7 weeks (IQR 6–12 resistant infections (MRSA and resistant CoNS) stayed significantly weeks). There were no significant differences in duration of treatment between the staphylococcal groups or betweenarthroplasty and fracture fixation device infections (8 vs. 7 weeks, p = 0.21). There were no clear preferences for the choice of antibioticagents. MRSA and CoNS were treated with vancomycin, doxycycline,and combinations of ciprofloxacin–rifampin or fusidic acid–rifam- Upon stratification of the results into the staphylococcal groups, pin. For MSSA, clindamycin, vancomycin, floxacillin, rifampin, and for arthroplasty infection, cure was achieved in 39% (7/18) of MRSA ciprofloxacin were used in the majority of cases. In 91 infections episodes, in 60% (15/25) of MSSA episodes, and in 77% (30/39) of (56%, 91/163), rifampin was used in combination therapy.
CoNS episodes. These differences were statistically significant All but two patients underwent surgery, and the median (Pearson Chi-square test between MRSA and CoNS, p = 0.008).
number of interventions to cure was two (IQR 1–2). There were no Patients with arthroplasty infections (n = 82, 50%) were signifi- significant differences in terms of number of surgical interventions cantly older than those with fracture fixation device infections between the staphylococcal groups or between the groups of (n = 81, 50%; median age 73 vs. 55 years, p < 0.001), were more arthroplasty vs. fracture fixation device infections (median immunosuppressed (28/82 vs. 14/81, p = 0.014), had a longer number two vs. two interventions, p = 0.65). In contrast, a revision incubation time (median delay 176 vs. 50 days, p < 0.034), had arthroplasty was more frequently performed in patients with CoNS significantly lower cure rates (52/82 vs. 65/81, p = 0.017), a infections than those with S. aureus infections significantly shorter recurrence time to infection (median delay 71 D. Teterycz et al. / International Journal of Infectious Diseases 14 (2010) e913–e918 Table 2Characteristics and comparisons between three groups of staphylococcal arthroplasty infections Median time delay between previous implantation and infection onset MRSA, methicillin-resistant Staphylococcus aureus; MSSA, methicillin-sensitive Staphylococcus aureus; CoNS, coagulase-negative staphylococci.
Group comparisons were performed with the Wilcoxon rank sum test, Fisher’s exact test, or the Pearson Chi-square test, as appropriate.
a Only statistically significant p-values of 0.05 (two-tailed) are displayed.
b Diabetes mellitus, transplantation, chronic alcoholism, neoplasia, Child’s class C cirrhosis, AIDS, steroid medication.
vs. 129 days, p = 0.011), and a longer hospitalization (median disease (for all types of infection). In contrast, infection due to duration 52 vs. 24 days, p < 0.001), and were more likely to receive MRSA was inversely associated with as was the case in the a new implant (34/82 vs. 15/81, p = 0.001). In contrast, the group of arthroplasty infections compared to the group with proportions of implant removal were similar (62 vs. 59, p = 0.69).
fracture fixation devices, which were mostly removed when summarizes the characteristics of arthroplasty patients infected. As previously reported by others,patient demo- stratified by staphylococcal infection. Of note, MRSA infections graphics (immunosuppression, sex, age), disease intensity (bac- occurred at a significantly higher patient age, had shorter teremia), and treatment modalities (rifampin duration of incubation times, and benefited less from the use of rifampin.
antibiotic therapy, number of surgical interventions, proportion ofone-stage revisions) did not influence cure.
One explanation for the lower cure rates of MRSA infections might lie in the lower proportion of new implants inserted in MRSA summarizes the univariate and the multivariate results infections as compared to those patients with CoNS disease. This is of logistic regression. In multivariate analysis including all implant highlighted by significantly more sequelae for patients with MRSA infections, arthroplasty (OR 0.2, 95% CI 0.1–0.6) and MRSA infections. Theoretically, a higher sequelae risk might be a sign that infections (OR 0.3, 95% CI 0.1–0.9) were significantly associated surgeons did not perform revision surgery and did not put in a new with lower cure outcomes, whereas CoNS infection (OR 3.0, 95% CI implant. We cannot completely exclude this decision bias.
1.2–8.0) and the insertion of a new implant (OR 4.5, 95% CI 1.6– It is clear that infection with S. aureus demonstrates an 13.1) were significantly associated with higher cure results.
enhanced virulence. For example bacteremic disease, a hallmark In the separate analysis for the group of arthroplasty infections of S. aureus infection, was not seen in CoNS disease. Two patients in only, a new implant (OR 12.8, 95% CI 2.7–61.9) showed a our study died secondary to S. aureus septicemia. Contrary to the statistically significant association with cure, while in the group evidence for staphylococcal bloodstream infections,non-pros- of fracture fixation device infections, no parameter reached thetic surgical site infections,and community-acquired MRSA, it remains unclear whether staphylococcal methicillin resistanceamong S. aureus or CoNS results in failure of treatment in localized tissue infections. While in vitro point to this concept, invivo studies show conflicting results. Al-Nammari et al. reported Our study shows that clinical features and outcomes differ the same duration of antimicrobial therapy and the same number considerably in orthopedic implant infections due to MRSA, MSSA, of surgical interventions for the treatment of septic arthritis or CoNS. Overall, cure increased from 57% for MRSA, to 72% for whether due to MRSA or Volin et al. demonstrated that MSSA, and to 82% for CoNS. This tendency was also similar when methicillin resistance did not influence the probability of cure in considering arthroplasty infections separately, with corresponding patients with two-stage re-implantation after total joint infec- cure rates of 39%, 60%, and 77%, respectively. In our study, the tion.In contrast, Kilgus et al. showed that infection following hip overall infection recurrence rate was only 12%, which is less than arthroplasty secondary to MRSA was treated successfully in only reported rates of 26%and 38%in the literature.
48% of cases, as compared to 81% with MSSA infection.Salgado When adjusted in multivariate analysis, revision arthroplasty et al. attributed a nine-fold higher hazard ratio to treatment failure with insertion of a new implant (for arthroplasty infections) was in prosthetic joint infections due to MRSA than due to MSSA.In the most significant protective factor for cure, followed by CoNS our analysis, methicillin resistance per se was not a risk factor for D. Teterycz et al. / International Journal of Infectious Diseases 14 (2010) e913–e918 Table 3Predictors of cure in staphylococcal orthopedic implant-associated infections MRSA, methicillin-resistant Staphylococcus aureus; MSSA, methicillin-sensitive Staphylococcus aureus; CoNS, coagulase-negative staphylococci.
a Only statistically significant p-values of 0.05 (two-tailed) are displayed.
b MRSA and methicillin-resistant CoNS.
failure, but was related to a higher age and a longer length of hospital stay.This reflects nosocomial aspects rather than damageby the pathogen itself, because methicillin-resistant infections We are indebted to Christophe Barea, Medical Informatics, for more likely occur in the elderly population with more co- his help in retrieving data. We thank the teams of the Orthopedic Service and the Laboratory of Bacteriology for their clinical Our study has limitations: (1) It was retrospective, from a single institution, and with a small sample size, thus limiting the Ethical approval: The study was approved by the hospital ethics generalizability of the findings. (2) Patients with an infection committee (No. 05-017, 05-041, 08-017R, 08-029R). No informed treated in another hospital may have been undetected. However, given that the Geneva University Hospitals comprise the largest Conflict of interest: No conflict of interest to declare.
and only public hospital in the area, and given the active post-discharge follow-up of our patients, we consider this selection bias to be minimal. (3) We used databases with microbiologicaldocumentation, with another possible selection bias for infections 1. Steckelberg JM, Osmon D, Prosthetic joint infections. 3rd ed. Bisno AL, Waldvo- where the microbiological cultures did not grow any staphylococci.
gel FA, editors. Infections associated with indwelling medical devices. Washing-ton DC: American Society for Microbiology Press; 2000, p. 173–209.
Half of the infections were due to methicillin-resistant strains, 2. Moran E, Masters S, Berendt AR, McLardy-Smith P, Byren I, Atkins BL. Guiding which are unlikely to be masked by unreported pre-hospitalization empirical antibiotic therapy in orthopaedics: the microbiology of prosthetic antibiotic use. In the literature culture-negative prosthetic joint infection managed by debridement, irrigation and prosthesis retention. JInfect 2007;55:1–7.
infections account for only 7% of cases, and the outcome appears 3. Neut D, van der Mei HC, Bulstra SK, Busscher HJ. The role of small-colony to be indistinguishable from that due to cultured bacteria.(4) The variants in failure to diagnose and treat biofilm infections in orthopedics. Acta potential influence of small-colony variants was not investigated since it was a retrospective study and many specimens had not 4. Uc¸kay I, Pittet D, Vaudaux P, Sax H, Lew D, Waldvogel F. Foreign body infections due to Staphylococcus epidermidis. Ann Med 2008;41:109–19.
been stored. Small-colony variants are resistant to aminoglyco- 5. Zimmerli W, Lew PD, Waldvogel FA. Pathogenesis of foreign body infection.
sides, considered difficult to treat, and responsible for recur- Evidence for a local granulocyte defect. J Clin Invest 1984;73:1191–200.
However, we consider their influence to be minimal 6. Bernard L, Vaudaux P, Huggler E, Stern R, Frehel C, Franc¸ois P, et al.
Inactivation of a subpopulation of human neutrophils by exposure to because small-colony variants are inherent to all three groups of ultrahigh-molecular-weight polyethylene wear debris. FEMS Immunol Med staphylococci independent of their methicillin In addition, our episodes were not recurrent infections and patients 7. Melzer M, Eykyn SJ, Gransden WR, Chinn S. Is methicillin-resistant Staphylo- coccus aureus more virulent than methicillin-susceptible S. aureus? A compar- were not subjected to long-term anti-staphylococcal therapy ative cohort study of British patients with nosocomial infection and before the onset of their infection.
bacteraemia. Clin Infect Dis 2003;37:1453–60.
D. Teterycz et al. / International Journal of Infectious Diseases 14 (2010) e913–e918 8. Shurland S, Zhan M, Bradham DD, Roghmann MC. Comparison of mortality risk 17. Stengel D, Bauwens K, Sehouli J, Ekkernkamp A, Porzsolt F. Systematic review associated with bacteraemia due to methicillin-resistant and methicillin-sus- and meta-analysis of antibiotic therapy for bone and joint infections. Lancet ceptible Staphylococcus aureus. Infect Control Hosp Epidemiol 2007;28:273–9.
9. Salgado CD, Dash S, Cantey JR, Marculescu CE. Higher risk of failure of methi- 18. Perlroth J, Kuo M, Tan J, Bayer AS, Miller LG. Adjunctive use of rifampin for the cillin-resistant Staphylococcus aureus prosthetic joint infections. Clin Orthop treatment of Staphylococcus aureus infections: a systematic review of the literature. Arch Intern Med 2008;168:805–19.
10. Kilgus DJ, Howe DJ, Strang A. Results of periprosthetic hip and knee infections 19. Engemann JJ, Carmeli Y, Cosgrove SE, Fowler VG, Bronstein MZ, Trivette SL, et al.
caused by resistant bacteria. Clin Orthop Relat Res 2002;404:116–24.
Adverse clinical and economic outcomes attributable to methicillin-resistance 11. Blanc DS, Pittet D, Ruef C, Widmer AF, Mu¨hlemann K, Petignat C, et al. Molecular among patients with Staphylococcus aureus surgical site infection. Clin Infect Dis epidemiology of predominant clones and sporadic strains of methicillin-resis- tant Staphylococcus aureus in Switzerland and comparison with European 20. Duckworth GJ, Jordens JZ. Adherence and survival properties of an epidemic epidemic clones. Clin Microbiol Infect 2002;8:419–26.
methicillin-resistant strain of Staphylococcus aureus compared with those of 12. Lu¨bbeke A, Stern R, Garavaglia G, Zu¨rcher L, Hoffmeyer P. Differences in out- methicillin-sensitive strains. J Med Microbiol 1990;32:195–200.
comes of obese women and men undergoing primary total hip arthroplasty.
21. Al-Nammari SS, Bobak P, Venkatesh R. Methicillin-resistant Staphylococcus aureus versus methicillin-sensitive Staphylococcus aureus adult haematogenous 13. Performance standards for antimicrobial susceptibility testing. Seventeenth septic arthritis. Arch Orthop Trauma Surg 2007;127:537–42.
informational supplement. Standard M100-S17. Wayne, PA: Clinical and Lab- 22. Volin SJ, Hinrichs SH, Garvin KL. Two-stage reimplantation of total joint infections: a comparison of resistant and non-resistant organisms. Clin Orthop 14. Harbarth S, Franc¸ois P, Schrenzel J, Fankhauser-Rodriguez C, Hugonnet S, Koessler T, et al. Community-associated methicillin-resistant Staphylococcus 23. Berbari EF, Marculescu C, Sia I, Lahr BD, Hanssen AD, Steckelberg JM, et al.
aureus, Switzerland. Emerg Infect Dis 2005;11:962–5.
Culture-negative prosthetic joint infection. Clin Infect Dis 2007;45:1113–9.
15. Yun HC, Branstetter JG, Murray CK. Osteomyelitis in military personnel wound- 24. Vaudaux P, Kelley WL, Lew DP. Staphylococcus aureus small colony variants: ed in Iraq and Afghanistan. J Trauma 2008;64:163–8.
difficult to diagnose and difficult to treat. Clin Infect Dis 2006;43:968–70.
16. Ure KJ, Amstutz HC, Nasser S, Schmalzried TP. Direct-exchange arthroplasty for 25. Adler H, Widmer A, Frei R. Emergence of a teicoplanin-resistant small colony the treatment of infection after total hip replacement. An average ten-year variant of Staphylococcus epidermidis during vancomycin therapy. Eur J Clin follow-up. J Bone Joint Surg Am 1998;80:961–8.
Microbiol Infect Dis 2003;22:746–8.

Source: http://www.swissfootcenter.ch/pdf/2010-Outcome-of-orthopedic-implant-infections-due-to-different-staphylococci.pdf