Antibiotic prophylaxis for wound infections in total joint arthroplasty

Transcription

1 Antibiotic prophylaxis for wound infections in total joint arthroplasty A SYSTEMATIC REVIEW B. AlBuhairan, D. Hind, A. Hutchinson From the University of Sheffield, Sheffield, England We reviewed systematically the published evidence on the effectiveness of antibiotic prophylaxis for the reduction of wound infection in patients undergoing total hip and total knee replacement. Publications were identified using the Cochrane Library, MEDLINE, EMBASE and CINAHL databases. We also contacted authors to identify unpublished trials. We included randomised controlled trials which compared any prophylaxis with none, the administration of systemic antibiotics with that of those in cement, cephalosporins with glycopeptides, cephalosporins with penicillin-derivatives, and second-generation with firstgeneration cephalosporins. A total of 26 studies ( participants) met the inclusion criteria. Methodological quality was variable. In a meta-analysis of seven studies (3065 participants) antibiotic prophylaxis reduced the absolute risk of wound infection by 8% and the relative risk by 81% compared with no prophylaxis (p < ). No other comparison showed a significant difference in clinical effect. Antibiotic prophylaxis should be routine in joint replacement but the choice of agent should be made on the basis of cost and local availability. B. AlBuhairan, BSc, MA, MPH, PhD Research Student A. Hutchinson, FRCGP, FFPH, Professor in Public Health Section of Public Health, D. Hind, BA, MA, PhD, Research Fellow Clinical Trials Research Unit School of Health and Related Research, University of Sheffield, Regent Court, 30 Regent Street, Sheffield S1 4DA, UK. Correspondence should be sent to Dr B. AlBuhairan; B.AlBuhairan@sheffield.ac.uk 2008 British Editorial Society of Bone and Joint Surgery doi: / x.90b $2.00 J Bone Joint Surg [Br] 2008;90-B: Received 16 November 2007; Accepted 15 February 2008 Although total hip replacement (THR) and total knee replacement (TKR) are considered to be clean operations, approximately 1% to 5% of wounds develop a superficial or deep infection. 1-4 The prevalence of peri-prosthetic infection is 1.3% after primary THR, 3.2% after revision THR, 2% after primary TKR and 5.6% after revision TKR. 5 The most prevalent organisms in prosthetic-related infections are the Grampositive bacteria, Staphylococcus aureus and Staphylococcus epidermidis. They are normally present as skin flora and can adhere to implants and multiply in polymers. 1 Increasing infection secondary to methicillinresistant strains of Staph. aureus and Staph. epidermidis has also emerged. The classes of antibiotic used in the treatment of wound infections include the β-lactams such as cephalosporins, penicillin and its derivatives, glycopeptides such as teicoplanin, and aminoglycosides such as gentamicin. In this article, we bring a previous systematic review 6 up to date by identifying and reviewing the evidence for the influence of antibiotic prophylaxis on the reduction of wound infections in patients undergoing total joint replacement. Materials and Methods The following electronic bibliographic databases were interrogated: the Cochrane Library, Issue 2, 2006; 7 OVID MEDLINE, 1966 to July 2007; 8 OVID EMBASE, 1980 to 2006 week 23; 9 and the Cumulative Index to Nursing and Allied Health Literature (CINAHL), to July In order to prevent time-lag bias and to identify unpublished studies, a search for grey literature was undertaken using the Google Scholar engine 11 and AMEDEO Medical literature guide. 12 We screened the reference lists of included articles to identify any further studies and contacted primary authors thus identified by for additional unpublished clinical trials. All the titles and abstracts were examined to assess their relevance to the review. Only studies meeting the following eligibility criteria were included: 1) types of participant, patients undergoing a primary or revision THR or TKR, irrespective of the type of prosthesis; 2) types of antibiotic administered at any time pre-operatively, irrespective of dose and route of administration and including β-lactams, glycopeptides, aminoglycosides and any others; 3) outcome, wound infection being defined as visible purulent exudate at the surgical site VOL. 90-B, No. 7, JULY

2 916 B. ALBUHAIRAN, D. HIND, A. HUTCHINSON Potentially relevant citations identified by searches (n = 354) Relevant citations retrieved for detailed evaluation (n = 38) (representing 32 RCTs) RCTs included in the meta-analyses (n = 26) (from 33 citations) Citations excluded on the basis of clearly irrelevant title or abstract (n = 316) Citations excluded after screening full paper (n = 7) Dosage/administration (n = 6) (5 RCTs) Wound infections not an outcome (n = 1) Fig. 1 Flow diagram of the selection of studies included in the review. (deep or superficial) reported at the maximum follow-up time; and 4) types of study (randomised controlled trial (RCT)). The following comparisons were analysed: 1) antibiotics versus no antibiotics; 2) systemic antibiotics versus antibiotic-impregnated cement; 3) teicoplanin versus firstand second-generation cephalosporins; 4) first- versus second-generation cephalosporins; and 5) any generation cephalosporin versus any penicillin derivative. Only publications in English were included because of constraints on time and resources. Studies were excluded if wound infection was not an outcome or if they only compared different doses of the same drug. The methodological quality of the included studies was independently assessed by one reviewer (BA) using the predefined criteria described by the Delphi list. 13 The following parameters were assessed: 1) the method of randomisation, classified as adequate (computer-generated randomnumber tables or similar methods), inadequate (the use of alternate days, coin tossing or similar methods) or unclear (process not reported); 2) treatment allocation concealment, classified as adequate (centralised or pharmacycontrolled or similar methods), inadequate (use of open random-number lists, sealed envelopes or similar methods) or unclear (not reported); 3) blinding, classified as triple (patient, health-care-provider and outcome-assessor blinded), double (patient and health-care-provider blinded), single (patient blinded) or unclear (no description); 4) baseline comparability, classified as adequate (similar groups) inadequate (groups not similar characteristics) or not reported baseline comparability was not assessed using a statistical test, but was determined by casually reviewing the studies; 5) description of eligibility criteria as defined (inclusion criteria specified), undefined (inclusion criteria not specified), or not reported; 6) sample size, classified as reported (sample-size calculation undertaken) or not reported; 7) intention-to-treat analysis, classified as adequate (analysis reported on all of the randomised joint arthroplasties), inadequate (analysis not reported on all the randomised joint arthroplasties) or unclear (no reporting on how the analysis was undertaken); and 8) exclusion, classified as reporting the number of cases excluded (joint replacements lost to follow-up, withdrawals) or unclear (no report of exclusions). Clinical outcome data were abstracted. When papers distinguished between the number of joints and patients randomised, we used the number of joints as our unit of analysis. We analysed the results of wound infection in all randomised joints. When this was not possible because of poor reporting, we used an available case analysis. When reported, we recorded the number of exclusions from the study, either without a justification or as a result of one or more of the following protocol violations: missing data, lack of resources, patient-related causes or requirement of additional surgery. We considered death without wound infection to be a competing risk, not an exclusion. Statistical analysis. When appropriate we undertook a meta-analysis. A random-effects Mantel-Haenszel model 14 was used for dichotomous data and the relative risk and 95% confidence interval (CI) were calculated. Inconsistency between trial results was assessed using the I 2 statistic, 15 which is the proportion of variation that is due to statistical heterogeneity rather than chance. The I 2 values of 25%, 50%, and 75% were interpreted as representing low, moderate and high heterogeneity. 16 The level of significance was set at 95%. Results The searches identified 354 citations including 11 from hand-searching, of which 316 were rejected on the basis that the title and abstract were irrelevant. There were 38 full papers retrieved involving 32 studies. Seven citations relating to six individual studies were excluded after reading the full paper, either because they were dose-finding studies or because wound infection was not a reported outcome. 23 The remaining 26 RCTs were included in the systematic review. A flow chart detailing the study selection was compliant with the Quality of Reporting of Meta- Analysis (QUOROM) statement (Fig. 1). 24 Seven studies were from North America, two from East Asia 32,33 and 17 from Europe, 28, were singlecentre 26,29,32,33,35,41-43,46,48,48-50 and 12 multicentre trials 27,28,30,31,36,37,39,40,44,45,47,51 and in two data were not reported. 25,38 There were two studies which reported only TKR patients, 32,33 11 THR patients, 34-43,50 11 both THR THE JOURNAL OF BONE AND JOINT SURGERY

3 ANTIBIOTIC PROPHYLAXIS FOR WOUND INFECTIONS IN TOTAL JOINT ARTHROPLASTY 917 and TKR patients, 26-29,44-50 one THR, TKR and other undefined replacements 30 and one undefined total joint replacement and other surgery such as gynaecological and gastrointestinal operations. 31 Some studies failed to specify whether the operations were primary or revision, the numbers of primary or revision procedures and which types of prosthesis were used. The methodological quality of the studies was variable. The method of randomisation was adequate in five studies 27,37,42,50,52 and unclear in the remaining 21. Allocation concealment was adequate in four studies 27,29,37,50 and unclear in the remaining 22. Blinding was well-described in one triple-blind study, 50 two double-blind studies 25,30 and one single-blind study. 31 In five double-blind 26,27,34,37,52 and five singleblind 33,42,45,46,49 studies there was insufficient information as to who was blinded. A total of 12 did not describe blinding. Baseline characteristics were adequately reported and well-balanced in 17 studies, 26-33,37,41-45,47,48,50 but unclear in nine. 25,34,35,38-40,46,49,52 Eligibility criteria were undefined in one study 25 and unclear in another. 38 Four presented an intention-to-treat analysis, 27,29,36,37 eight did not but gave sufficient data to allow us to do so. 26,30,34,39,41,42,47,50 Four studies presented only enough information to reconstruct an available case analysis, 28,35,40,46 and in ten there was insufficient information to determine whether the analyses were by intentionto-treat. 25,31-33,38,43-45,48,49 In five, 28,35,40,45,47 fewer than 5% of patients were excluded from the analysis, in eight, 29,30,36,37,39,41,42,45 between 5% and 20%, in two, 27,46 more than 20% and in 11 25,26,31-34,38,43,44,48,49 the number of exclusions was unclear. Eight studies defined wound infection as superficial or deep 27,32,33,39-41,44,45 and five considered only deep infections. 34,37,42,47,51 Nine did not provide a definition of wound infection, 26,29,30,35,43,48-50,52 although one of these distinguished wounds which contained purulent exudate. 49 Follow-up periods ranged from ten days to ten years, with most not specifying the minimum and maximum follow-up times (we report only the latter). The follow-up time was unclear in two studies. 35,43 A total of 14 studies reported the route of administration of antibiotics for all doses 25,26,28,32-34,40,41,44,45,47-50,51 and six reported the route for the first, but not the subsequent doses. 27,35,38,39,42,46 The remaining six studies did not report the route of administration for all doses ,36,37,43 The recorded route of administration in most studies was intravenous and only five used antibiotic-impregnated cement. 32,33,39,44,45 Antibiotics were either administered pre-operatively, at the induction of anaesthesia, up to 24 hours before surgery, just before incision, intra-operatively or post-operatively. In seven studies, 28,42,43,46,47,49,53 one treatment arm received a single dose compared with multiple doses in the other treatment arm. The duration of antibiotics also varied from one to 14 days. In a pooled analysis of seven studies 32-34,36,38,41,43 (n = 3065) the administration of antibiotics reduced the relative risk (RR) of wound infection by 81% (RR 0.19; 95% CI 0.12 to 0.31; chi-squared test, p < ). Because such events are rare, this translates to an absolute risk reduction of 8%, meaning that one wound infection would be prevented for every 13 people treated compared with no administration of antibiotics (risk difference -0.08; 95% CI to -0.12). There was no statistical heterogeneity (I 2 = 0%). The pooled analysis of three studies 39,44,45 (n = 2388) comparing the administration of systemic antibiotics with that of antibiotic-impregnated cement found no significant difference in clinical effect (RR 0.88; 95% CI 0.59 to 1.31; chi-squared test, p = 0.52). There was a low level of statistical heterogeneity (I 2 = 36.9%). The pooled analysis of five studies 28,42,46,47,49 (n = 2625) comparing the use of cephalosporins with that of teicoplanin found no significant difference in clinical effect (RR 1.22; 95% CI 0.64 to 2.34; chi-squared test, p = 0.54). There was no statistical heterogeneity (I 2 = 0%). The pooled analysis of three studies 40,48,53 comparing the use of cephalosporins with that of penicillin-derivatives (n = 386) found no significant difference in clinical effect (RR 1.17; 95% CI 0.31 to 4.41; chi-squared test, p = 0.81). There was a small amount of statistical heterogeneity (I 2 = 7.4%). In the pooled analysis of eight studies 25-27,29,30,35,37,50 (n = 2879) comparing the use of second-generation with that of first-generation cephalosporins, there was no significant difference in clinical effect (RR 1.08; 95% CI 0.63 to 1.84; chi-squared test, p = 0.79). There was no statistical heterogeneity (I 2 = 0%). Discussion Antibiotic prophylaxis is effective in reducing wound infection in patients following total joint replacement. There was insufficient evidence to suggest that there was a significant difference in the efficacy of cephalosporins compared with that of teicoplanin or penicillin-derivatives, or that a particular generation of cephalosporins was more effective than another. There was no difference in clinical effect whether antibiotics were administered systemically or through impregnated cement. Two comparisons showed low levels of inconsistency between the results of primary studies. When comparing the use of systemic antibiotics with antibiotic-impregnated cement, one study 45 appeared to be an outlier, despite being of similar methodological quality and incorporating comparable populations, interventions and outcome measures. Statistical heterogeneity in the comparison of cephalosporins with penicillin-derivatives was probably due to chance in the study of Jones and Wojeski, 53 in which a single event resulted in a disproportionately large clinical effect because of the small sample size. The larger trial of Pollard et al 40 was the most influential in the meta-analysis and represented a more reasonable estimate of effect. VOL. 90-B, No. 7, JULY 2008

4 918 B. ALBUHAIRAN, D. HIND, A. HUTCHINSON There are discrepancies between our findings and those of a previous systematic review. 6 The latter included studies in all languages, whereas time and resource constraints meant that we could not. Therefore, relevant non-englishlanguage papers published since 1998 were not identified in our review. However, when it has any effect, the exclusion of non-english-language papers from systematic reviews usually exaggerates the estimate of clinical effect, because non-significant results may be more likely to be published in non-english-language journals. 54 We avoided repeating comparisons made in the original review which we considered to be clinically uninformative, for example, those between cephamandole or teicoplanin and any other antibiotic. There are discrepancies between the number of wound infections reported for a study in the original review and in our review. The reasons were twofold. First, in three studies 36,39,50 the original review did not use outcome data from the longest available follow-up. Secondly, our analysis was, as far as possible, based on all randomised joints, while the original was not. Failure to analyse all randomised joints introduced bias into statistical analyses and increased the likelihood of false-positive results. The identification of optimal dosages and the duration of systemic therapy was beyond this review. The original review found that single-dose or short-term administration was as effective as long-term administration, but with lower overall costs and reduced risk of toxicity or bacterial resistance. 6 Therefore, teicoplanin was administered as a single dose and cephalosporins in fewer doses and over a shorter period than penicillin derivatives. We identified a paucity of RCTs in this important area of clinical research, which could be because the rates of wound infection after total joint replacement are low. The reported rates for wound infection, however, may have been underestimated because most trials in this review varied in their definition of wound infection and followup period. Nevertheless, the demand for total joint replacement will increase in the coming years, predominantly because of an ageing population and there is consequently a greater need for further RCTs. We recommend that future research should focus on improving the rigour of RCTs in accordance with the Consort Standards. 55 We suggest that any RCT should consider carefully its measure of outcome and adoption of a common internationally accepted definition to allow comparisons between trials. Our results add to the evidence base on the prevention of post-operative infection in total joint replacement. 56,57 For every 13 patients treated, rather than untreated, one wound infection would be prevented. Since the effect of postoperative infection in such patients is serious they should be offered antibiotic prophylaxis. 58 Since our results show that there is no evidence that any type of antibiotic prophylaxis has better results than any others, selection should be on the basis of cost and local availability. Supplementary Material A further opinion by Professor S. Hughes is available with the electronic version of this article. In addition, supplementary tables and figures showing details of the studies and meta-analysis are available on our website at No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. References 1. Nehrer S, Thalhammer F, Schwameis E, Breyer S, Kotz R. Teicoplanin in the prevention of infection in total hip replacement. Arch Orthop Trauma Surg 1998;118: Zimmerli W, Trampuz A, Ochsner P. Prosthetic-joint infections. N Engl J Med 2004;351: Garner J. Guideline for prevention of surgical wound infections. wonder/prevguid/p /p asp (date last accessed 27 October 2005). 4. Nathens AB, Delinger EP. Surgical site infections. Curr Treat Options Infect Dis 2000;2: König DP, Schierholz JM, Münnich U, Rütt J. Treatment of staphylococcal implant infection with rifampicin-ciprofloxacin in stable implants. Arch Orthop Trauma Surg 2001;121: Glenny A, Song F. Antimicrobial prophylaxis in total hip replacement: a systematic review. Health Technol Assess 1999;3: No authors listed. Cochrane Library, Issue view/item/review_one.jsp?j=603 (date last accessed 27 May 2008). 8. No authors listed. OVID MEDLINE, 1966 to July ovidweb.cgi?new+database=single/13&s=bfhbpdiecohfooiefnhlahihhppaa00 (date last accessed 25 April 2008). 9. No authors listed. OVID EMBASE, 1980 to 2006 week spb/ovidweb.cgi?new+database=single/12&s=jgpmfpge (date last accessed 25 April 2008). 10. No authors listed. Cumulative Index to Nursing and Allied Health Literature (CINAHL). BFHBPDIEC (date last accessed 25 April 2008). 11. No authors listed. Google Scholar. (date last accessed 27 May 2008). 12. No authors listed. AMEDEO. (date last accessed 27 May 2008). 13. Verhagen AP, De Vet HC, de Bie RA, et al. The Delphi list: a criteria list for quality assessment of randomized clinical trials for conducting systematic reviews developed by Delphi consensus. J Clin Epidemiol 1998;51: Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst 1959;22: Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002;21: Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ 2003;327: delalla F, Viola R, Pellizer G, et al. Regional prophylaxis with teicoplanin in monolateral or bilateral total knee replacement: an open study. Antimicrob Agents Chemother 2000;44: Nelson CL, Green TG, Porter RA, Warren RD. One day versus seven days of preventive antibiotic therapy in orthopedic surgery. Clin Orthop 1983;176: Heydemann JS, Nelson CL. Short-term preventive antibiotics. Clin Orthop 1986;205: Wymenga B, Hekster YA, Theeuwes A, et al. Antibiotic use after cefuroxime prophylaxis in hip and knee joint replacement. Clin Pharmacol Ther 1991;50: Wymenga A, Van Horn J, Theeuwes A, Muytjens H, Slooff T. Cefuroxime for prevention of postoperative coxitis: one versus three doses tested in a randomized multicenter study of 2,651 arthroplasties. Acta Orthop Scand 1992;63: Ritter MA, Campbell E, Keating EM, Faris PM. Comparison of intraoperative versus 24 hour antibiotic prophylaxis in total joint replacement: a controlled prospective study. Orthop Rev 1989;18: Johnson DP. Antibiotic prophylaxis with cefuroxime in arthroplasty of the knee. J Bone Joint Surg [Br] 1987;69-B: Moher D, Cook DJ, Eastwood S, et al. Improving the quality of reports of metaanalyses of randomised controlled trials: the QUORUM statement: quality of reporting of meta-analyses. Lancet 1999;354: THE JOURNAL OF BONE AND JOINT SURGERY

5 ANTIBIOTIC PROPHYLAXIS FOR WOUND INFECTIONS IN TOTAL JOINT ARTHROPLASTY Ritter M, Conway M, Stringer E, Williams J. Cephalosporin for total hip replacement. Orthopedics 1983;6: DeBenedictis KJ, Rowan NM, Boyer BL. A double-blind study comparing cefonicid with cefazolin as prophylaxis in patients undergoing total hip or knee replacement. Rev Infect Dis 1984;6: Mauerhan DR, Nelson CL, Smith DL, et al. Prophylaxis against infection in total joint arthroplasty: one day of cefuroxime compared with three days of cefazolin. J Bone Joint Surg [Am] 1994;76-A: Periti P, Stringa G, Donati L, et al. Teicoplanin: its role as systemic therapy of burn infections and as prophylaxis for orthopaedic surgery: Italian Study Groups for Antimicrobial prophylaxis in orthopaedic surgery and burns. Eur J Surg 1992;567(Suppl): Soave R, Hirsch JC, Salvati EA, Brause BD, Roberts RB. Comparison of ceforanide and cephalothin prophylaxis in patients undergoing total joint arthroplasty. Orthopedics 1986;9: Davis WA, Kane JG. Antimicrobial prophylaxis for arthroplasty: a comparative study of cefonicid and cefazolin. Orthopedics 1987;10: Jones RN, Slepack JM, Wojeski WV. Cefotaxime single-dose surgical prophylaxis in a prepaid group practice: comparisons with other cephalosporins and ticarcillin/clavulanic acid. Drugs 1988;35: Chiu FY, Chen CM, Lin CF, Lo WH. Cefuroxime-impregnated cement in primary total knee arthroplasty: a prospective, randomized study of three hundred and forty knees. J Bone Joint Surg [Am] 2002;84-A: Chiu FY, Lin CF, Chen CM, Lo WH, Chaung T-Y. Cefuroxime-impregnated cement at primary total knee arthroplasty in diabetes mellitus: a prospective, randomised study. J Bone Joint Surg [Br] 2001;83-B: Carlsson AS, Lidgren L, Lindberg L. Prophylactic antibiotics against early and late deep infections after total hip replacements. Acta Orthop Scand 1977;48: Davies AJ, Lockley RM, Jones A, el-safty M, Clotheir JC. Comparative pharmocokinetics of cefamandole, cefuroxime and cephradine during total hip replacement. J Antimicrob Chemother 1986;17: Doyon F, Evrard J, Mazas F, Hill C. Long-term results of prophylactic cefaxolin versus placebo in total hip replacement. Lancet 1987;1: Evrard J, Doyon F, Acar JF, et al. Two-day cefamandole versus five-day cephazolin prophylaxis in 965 total hip replacements: report of a multicentre double blind randomised trial. Int Orthop 1988;12: Gunst JP, Deletang S, Rogez MJ, et al. Prophylactic antibiotic therapy by cefamandole in total hip surgery replacement using Charnley s tent: a randomized study. Pathol Biol (Paris) 1984;32:567-9 (in French). 39. Josefsson G, Kolmert L. Prophylaxis with systematic antibiotic versus gentamicin bone cement in total hip arthroplasty. Clin Orthop 1993;292: Pollard JP, Hughes SP, Scott JE, Evans MJ, Benson MKD. Antibiotic prophylaxis in total hip replacement. BMJ 1979;1: Schulitz KP, Winkelman W, Schoening B. The prophylactic use of antibiotics on alloarthroplasty of the hip joint for coxarthrosis. Arch Orthop Trauma Surg 1980;96: Suter F, Avai A, Fusco U, et al. Teicoplanin versus cefamandole in the prevention of infection in total hip replacement. Eur J Clin Microbiol Infect Dis 1994;13: Wollinsky KH, Oethinger M, Büchele M, et al. Autotransfusion: bacterial contamination during hip arthroplasty and efficacy of cefuroxime prophylaxis: a randomized controlled study of 40 patients. Acta Orthop Scand 1997;68: McQueen M, Littlejohn A, Hughes SP. A comparison of systemic cefuroxime and cefuroxime loaded bone cement in the prevention of early infection after total joint replacement. Int Orthop 1987;11: McQueen MM, Hughes SP, May P, Verity L. Cefuroxime in total joint arthroplasty: intravenous or in bone cement. J Arthroplasty 1990;5: Mollan RA, Haddock M, Webb CH. Teicoplanin vs cephamandole for antimicrobial prophylaxis in prosthetic joint implant surgery: preliminary results. Eur J Surg 1992;567(Suppl): Periti P, Stringa G, Mini E. Comparative multicenter trial of teicoplanin versus cefazolin for antimicrobial prophylaxis in prosthetic joint implant surgery: Italian Study Group for Antimicrobial prophylaxis in orthopaedic surgery. Eur J Clin Microbial Infect Dis 1999;18: Vainionpää S, Wilppula E, Lalla M, Renkonen OV, Rokkanen P. Cefamandole and isoxazolyl penicillins in antibiotic prophylaxis of patients undergoing total hip or knee-joint arthroplasty. Arch Orthop Trauma Surg 1988;107: Wall R, Klenerman L, McCullough C, Fyfe I. A comparison of teicoplanin and cefuroxime as prophylaxis for orthopaedic implant surgery: a preliminary report. J Antimicrob Chemother 1988;21: Bryan CS, Morgan SL, Caton RJ, Lunceford EM Jr. Cefazolin versus cefamandole for prophylaxis during total joint arthroplasty. Clin Orthop 1988;228: Ericson C, Lidgren L, Lindberg L. Cloxacillin in the prophylaxis of postoperative infections of the hip. J Bone Joint Surg [Am] 1973;55-A: Hill C, Flamant R, Mazas F, Evrard J. Prophylactic cefazolin versus placebo in total hip replacement: report of a multicentre double-blind randomised trial. Lancet 1981;1: Jones RN, Wojeski WV. Single-dose surgical prophylaxis using ticarcillin/clauvulanic acid (Timentin): a prospective, randomised comparison with fecotaxime. Diagn Microbiol Infect Dis 1987;7: Egger M, Zellwager-Zähner T, Schneider M, et al. Language bias in randomised controlled trials published in English and German. Lancet 1997;350: No authors listed. Consort Group. (date last accessed 5 October 2006). 56. Quenon JL, Eveillard M, Vivien A, et al. Evaluation of current practices in surgical antimicrobial prophylaxis in primary total hip prosthesis: a multicentre survey in private and public French hospitals. J Hosp Infect 2004;56: Bratzler DW, Houck PM, Richards C, et al. Use of antimicrobial prophylaxis for major surgery: baseline results from the National Surgical Infection Prevention Project. Arch Surg 2005;140: Graves N, Halton K, Lairson D. Economics and preventing hospital-acquired infection: broadening the perspective. Infect Control Hosp Epidemiol 2007;28: VOL. 90-B, No. 7, JULY 2008