Delayed Infections After Posterior TSRH Spinal Instrumentation for Idiopathic Scoliosis Revisited

SPINE Volume26,Number18,pp1990–1996©2001,Lippincott Williams&Wilkins,Inc.Delayed Infections After Posterior TSRH Spinal Instrumentation for Idiopathic Scoliosis RevisitedB.Stephens Richards,MD,and Khaled M.Emara,MDObjective.To determine the incidence of delayed in-fections in idiopathic scoliosis treated with TSRH instru-mentation,proper wound management after instrumen-tation removal,and whether the previously identifiedbacterial trend remains consistent.Methods.All patients with idiopathic scoliosisՆ2years after surgery with posterior TSRH instrumentation were in-cluded.Those cases with delayed infections were retrospec-tively reviewed.Time of presentation(infection)from index operation,clinical picture,sedimentation rate,presence of pseudarthrosis,organisms grown on culture,type of woundclosure,and duration of antibiotics were examined.Results.A total of489patients were identifiedՆ2years postoperation;23had delayed infections(4.7%).Time ofpresentation averaged27months after initial surgery.Spon-taneous drainage occurred in15patients,fluctuance in6,and neither in the remaining2(pain and fever).Sedimenta-tion rate averaged48mm/hr.All patients had instrumenta-tion removed.Primary closure(1stage)was performed in14patients,and delayed primary closure(Ն2stages)wasperformed in nine patients.All wounds healed uneventfully.Cultures at the time of instrumentation removal grew Pro-pionibacterium acnes in12patients,Staphylococcus epider-midis(or Staphylococcus coagulase-negative)in4,Micro-coccus varians in1,and Staphylococcus aureus in1.Fivepatients had negative cultures.After removal,patients re-ceived parenteral antibiotics;in21of these patients this was followed by oral antibiotics.Conclusions.Low-virulent skin organisms are primar-ily responsible for delayed infections.Intraoperativeseeding followed by subclinical quiescent periods ap-pears to be the method by which infection occurs.Theincreased bulk and modularity of modern instrumenta-tion systems can lead to inflammation and bursa forma-tion,thus contributing significantly to the activation ofthese infections.Primary wound closure results in suc-cessful wound healing.Delayed closure after48hours isunnecessary.Short-term postoperative parenteral antibi-otics(2–5days)followed by short-term oral antibiotics(7–14days)is recommended.[Key words:delayed infec-tions,spinal instrumentation,Propionibacterium acnes,Staphylococcus epidermidis]Spine2001;26:1990–1996 Multiple hook-screw-rod spinal instrumentation sys-tems,one of which is Texas Scottish Rite Hospital (TSRH)spinal instrumentation,have been used for the treatment of spinal deformities over the past12–15 years.These modern modular systems have provided greater versatility in the correction of scoliosis than did the earlier Harrington instrumentation.Unfortunately, these newer systems increased the amount and bulk of instrumentation being implanted.Over time,several stud-ies emerged reporting delayed spontaneous drainage in sco-liosis patients,ranging in occurrence from1%to 6.7%.2,4,8,11,13Some authors concluded that this delayed drainage represented an aseptic process resulting from an inflammatory soft tissue reaction to fretting corrosion of the metal implants.4,13Others conclude that these episodes of drainage represented delayed bacterial infections.2,8,11 In1995,we reported on what we considered delayed deep wound infections in thefirst149patients at our institution treated with posterior TSRH instrumentation for idiopathic scoliosis.8Ten patients with delayed infec-tions were identified(incidence,6.7%).The infections were diagnosed an average of25months after the initial procedure.Cultures taken from specimens at the time of instrumentation removal identified several low-virulent skinflora,which had been considered nonpathogenic in humans20years ago.These organisms included Propi-onibacterium acnes,Staphylococcus epidermidis,and Micrococcus varians.We found that P.acnes required an extended duration(7–10days)of incubation to be iden-tified.If thefinal culture reading had been completed at 72hours,the causative organism clearly could have been missed.This fact was borne out in a similar study of22 patients reported by Clark and Shufflebarger in1999.2In theirfirst10patients,cultures were performed for only 72hours and were negative in nine.Subsequently,when cultures were continued for7days,11of the remaining 12patients were positive.S.epidermidis was the pre-dominant organism in their report.This study was undertaken to determine1)the inci-dence of delayed infections in a larger idiopathic scoliosis population treated with TSRH instrumentation,2)the proper wound management after instrumentation re-moval,and3)whether the previously identified bacterial trend(low-virulent organisms)has remained consistent over time in a larger population.Materials and MethodsAll patients with idiopathic scoliosis treated at Texas Scottish Rite Hospital who were atϾ2years after their primary surgery with posterior TSRH instrumentation were included.A total of 489patients were identified who had been operated on be-tween May1988and December1997.Surgery was performed by eight different staff orthopedic surgeons.From this popula-tion23patients were identified who had developed delayed infections.Medical records and radiographs of these23pa-tients were retrospectively reviewed to determine age of theFrom the Texas Scottish Rite Hospital for Children,Dallas,Texas. Acknowledgment date:November8,2000.Revision date:February23,2001.Acceptance date:March29,2001.Device status category:11.Conflict of interest category:12.1990patient,duration of the operation,perioperative use of antibi-otics,loss of blood,amount of blood salvaged with the use of cell saver,type of bone graft used,length of fusion,number of hooks and crosslinks used,presence of risk factors(such as urinary tract infections or previous spinal operations),dura-tion of follow-up after the initial operation,clinical presenta-tion for the delayed infection,erythrocyte sedimentation rate, presence of pseudarthrosis,organisms grown on culture,type of wound closure(primary or delayed)performed after the removal of the instrumentation,duration of antibiotic treat-ment,and outcome.Tissue removed at the time of surgery was not histologically examined for the presence of metallic debris.ResultsTwenty-three patients(20female,3male)presented with delayed infections,for an incidence of4.7%.For these 23patients,the average age at the time of the index operation was14.7years(range,12.4–19years),the average duration of the operation was4hours13min-utes(range,2hours50minutes to6hours20minutes), and the average estimated blood loss was877mL(range, 150–2500mL)(Table1).All of the patients received cefazolin intraoperatively and for24–48hours after sur-gery.An average of11levels were included in the arth-rodesis.All of the patients received autologous iliac crest bone graft to the fusion site.Eighteen of the patients received intraoperative salvage of autologous blood with use of the cell saver;the average blood return for these patients was335mL(range,120–900mL).Fourteen patients received before surgery donated autologous blood during the surgical procedure.Risk factors were identified retrospectively for only two of the patients.One patient(case5)had a reopera-tion6weeks after the initial procedure because of dis-lodgement of an inferior hook.Another patient(case7) had abused drugs intravenously;this was not known by the surgeon at the time of the initial procedure.Twelve patients(cases1–4,7,8,10,12,14–16,and 20)were operated on by one staff surgeon who had op-erated on92of the489patients.The remaining11pa-tients had been operated on byfive staff surgeons,as follows.Cases9,13,and17occurred from a group of 115patients;cases5and21occurred from a group of92 patients;cases11and22occurred from a group of84 patients;case6occurred from a group of42patients; and cases18,19,and23occurred from a group of34 patients.Two other staff surgeons had operated on the remaining30patients;none of these patients had a de-layed infection.There were no apparent substantial dif-ferences in operative technique,personnel,or equipment to explain the higher prevalence of infection seen in two of the staff surgeons–patient populations.None of the23patients who presented with delayed infection had an acute postoperative infection or any signs of an impending postoperative infection.The delayed infec-tions were diagnosed at an average of27months(range, 11–79months)after the initial procedure(Table2).Spon-taneous drainage occurred in15patients,fluctuance was present in six,and two patients clinically presented only with pain and fever(Figure1).None of the patients had exposed instrumentation.There was no clinical evidence toTable1.Data on the Initial Operation for ScoliosisPatientNo.SexAge atOperation(yr)Date ofOperationDuration of Operation(hrϩmin)EstimatedBlood Loss(mL)Cell Saver(mL)AutologousPacked Blood(U)No.of Vertebraein ArthrodesisNo.of Hooks/No.of Crosslinks1F132/28/894ϩ1560037521210/2 2F13.53/9/893ϩ258502501108/2 3F14.23/14/892ϩ50600120N/A108/2 4F133/21/893ϩ401300N/A11210/2 5a F12.44/27/895ϩ15150042021410/2 5b6/14/894ϩ30750350N/A13/4 6F15.27/24/894ϩ30500N/A298/2 7M198/15/893ϩ401800900N/A1110/2 8F14.73/8/903ϩ306002401118/2 9F14.64/19/904ϩ201000N/A11010/2 10F14.79/18/905ϩ0160049021312/2 11F15.112/12/904ϩ30120024021311/2 12F12.810/20/923ϩ508004751108/2 13F151/6/944ϩ10900245N/A1413 14F14.96/28/943ϩ50500125N/A87 15F13.99/2/944ϩ30100035011413 16F15.46/19/955ϩ10250050021211/2 17M16.810/25/956ϩ201300675N/A119/2 18M17.44/27/964ϩ50400125N/A910/2 19F14.47/9/963ϩ30500130N/A1011/2 20F16.37/2/963ϩ20422125N/A98/2 21F13.57/18/964ϩ45100020011211/2 22F13.84/23/973ϩ55150N/A N/A910/2 23F156/11/974ϩ0600N/A11211/2 N/Aϭnot applicable.1991Delayed Infections Revisited•Stephens Richards and Emarasuggest erosion of the skin over prominence of instrumen-tation.The average erythrocyte sedimentation rate(ESR) for the18patients for whom this information was available was48mm/hours(range,17–100mm/h).For the two pa-tients whose clinical presentation included only back pain and fever(cases17and23),the ESRs were92and94 mm/h.All of the patients were taken to the operating room for removal of their instrumentation.Granulation tissue was seen around the rods and was frequently dark in color when it was in proximity to loose hook–rod or crosslink–rod interfaces.The tissue was not sent for histologic evaluation.After removal of the instrumentation and debridement of the tissues,pri-mary closure(one stage)was performed in14patients. Delayed primary closure was performed in nine patients. In eight of these nine patients the wound was closed during the second operative intervention(2to4days later).In the remaining patient,the wound was irrigated and debrided twice before delayed closure on the6th day.All of the wounds healed uneventfully.Cultures of specimens taken from deep within the wound were performed on all patients.P.acnes was identified in12patients,S.epidermidis(or Staphylococ-cus coagulase-negative)in four patients,M.varians in one patient,and rare S.aureus in one patient.Of these18 positive cultures,five(cases1,2,3,11,and21)grew in liquid media only.Five patients had negative cultures despite prolonged incubation(cases6,12,18,20,and 23).Three of thesefive patients(cases12,18,and23) received cephalexin orally for1–2weeks shortly before instrumentation removal,which may have significantly impacted the culture results.Preoperative superficial wound cultures were obtained in several patients.Three were positive for organisms different from those ob-Table2.Data on the Delayed InfectionsPatient No.TimeSinceOperation(mos)Sign of InfectionESR(mm/hr)Pseudarthrosis Type of ClosureOrganism Grown onCultureParenteralAntibiotics(days)Oral Antibiotics(days)141Drainage,low back191stage S.epidermidis Vancomycin11Cloxacillin28Rifampin28 232Drainage,crosslink282stages,3days apartP.acnes Cefazolin14Clindamycin28311Drainage,crosslink261stage P.acnes Cefazolin2Gentamicin2Cephalexin14 Tetracycline60415Fluctuance,low back67T8–T91stage P.acnes Cefazolin5Cefaclor14 5a45Drainage,crosslink201stage S.epidermidis Cefazolin10Gentamicin2Cephalexin14 5b618Drainage,mid back—2stages,3days apartNo growth Cefazolin5—724Drainage,crosslink811stage P.acnes Cefuroxime14Cephalexin? 830Drainage,crosslink321stage P.acnes Cefazolin4Cephalexin12 979Pain,fluctuance,low back—2stages,4days apartP.acnes Cefazolin7Cephalexin141015Fluctuance,crosslink52L1–L21stage Coagulase-negativeStaphylococcus Cefazolin5Cephalexin3Tetracycline141120Drainage,crosslink252stages,3days apartMicrococcus varians†Cefazolin5Cephalexin16 1228Pain,fluctuance,low back371stage No growth*Cefazolin7Cephalexin141342Pain,fluctuance,fever,midthoracic 1002stages,3days apartP.acnes Cefuroxime6Ciprofloxacin71420Drainage,mid thoracic181stage P.acnes Cefazolin4—1524Pain,drainage,mid thoracic—1stage P.acnes Oxacillin3Gentamicin3Cephalexin10 1617Drainage,mid thoracic—L2–L31stage withreinstru-mentationP.acnes Cefazolin5Cephalexin211711Pain,fever922stages,3irrigated anddebrided in6daysRare S.aureus Cefazolin13Cloxacillin71843Pain,drainage,mid thoracic172stages,1dayapartNo growth*Cefazolin5Cephalexin42 1927Pain,drainage,low back182stages,2days apartS.epidermidis Cefazolin6Cephalexin142017Drainage,mid thoracic—1stage No growth Cefazolin4Cephalexin10 2126Pain,drainage,low back961stage P.acnes Cefazolin6Cephalexin282217Fluctuance,low back(crosslink)512stages,2days apartP.acnes Cefazolin8Cephalexin302318Pain,fever941stage No growth*Cefazolin4Cephalexin14 ESRϭerythrocyte sedimentation rate.*Cephalexin for1–2weeks shortly before instrumentation removal.1992Spine•Volume26•Number18•2001tained on deep cultures:Clostridium beijerinckii(case5), S.aureus(case7),and S.epidermidis(case9).After instrumentation removal,all patients received parenteral antibiotics for2–14days.Twenty-one of the patients then received oral antibiotics for approximately 1–9weeks(Table2).Pseudarthrosis was noted in three patients(cases4, 10,and16).In two of these patients(cases4and10),the pseudarthrosis was identified at a level in which two hooks,oriented in opposite direction,had been placed in one intervertebral space.One of the three patients(case 16)had exchange of instrumentation at the initial reop-eration.She then received antibiotics for4weeks.Nearly 2years after reinstrumentation her wound has remained healed.Another of the three patients(case4)was rein-strumented8months after her instrumentation removal because of curve progression.She developed a recurrent infection with P.acnes4months and8months after the reinstrumentation.The wound was debrided and then closed primarily over intact instrumentation,and the pa-tient took clindamycin orally for5months.The incision in the back healed without redness or swelling and re-mained so2years after the reinstrumentation.Thefinal patient presenting with pseudarthrosis was not reinstrumented.DiscussionSeveral reports of delayed,or late-developing,deep infec-tions after treatment for scoliosis with multiple hook-screw-rod instrumentation systems have been published re-cently.2,4,6,8,10,11In a previous study conducted at this institution,Richards found10cases(6.7%)of delayed in-fection in149patients who had been treated with TSRH instrumentation.8Those10patients are incorporated into this current report,which has expanded the number of pa-tients treated with TSRH instrumentation.A lower inci-dence(4.7%)of delayed infections was found in this cur-rent report.Another retrospective review of idiopathic scoliosis patients treated by a single surgeon with either Cotrel–Dubousset or Moss–Miami instrumentation re-ported a1.7%incidence of delayed infections(22of1247 patients).2For thefirst10patients in that study,cultures were performed for only72hours and were negative in nine patients.After a report by Richards,7they extended the duration of the cultures and had a positive report in11of the remaining12,the majority being S.epidermidis or P. acnes.In still another report by Viola et al,11delayed infec-tions developed in eight patients(unknown incidence).11S. epidermidis or P.acnes was responsible for seven of the eight cases.In that study the ESR averaged57mm/h in their patients,an abnormal elevation similar to that found in this current study(48mm/h).Delayed infection may result from one of two possible pathways.Thefirst pathway results from intraoperative seeding of the low-virulent skinflora,followed by a qui-period.2,8,11At some point in the future,the infec-becomes clinically evident with the presence of backfever,localized swelling along the incision,and/orspontaneous drainage.This time interval between primarysurgery and the presentation of infection averaged2.3years(range,0.9–6.7years)in this study and3.1years(range,1.2–8.5years)in the Clark and Shufflebarger study.2Westrongly believe,as did Clark and Shufflebarger,2that thispathway is responsible for the majority of the infections.Unfortunately,this mechanism of intraoperative seed-ing remains conjectural and is difficult,if not impossible,to prove.Support for this mechanism,however,was pro-vided in a study by Dietz et al in1991.3They describedpositive cultures of low-virulent organisms(P.acnes andcoagulase-negative Staphylococcus)in a series of40pa-tients who underwent routine culture of the operativesite during elective clean orthopedic(nonscoliosis)oper-ations.The patients had no evidence of infection at thetime.The bacteria,which are normal skinflora,werethought to have been carried into the wound during theoperation.The authors concluded that it remained clin-ically difficult to know whether these organisms could beresponsible for a true delayed infection(such as thoughtto be the case in this present study)or whether the presenceof these organisms is only a coincidentalfinding.In partic-Figure1.This16.7-year-old girl(case14)presented20monthsafter her primary posterior spinal instrumentation and fusion witha draining sinus in the midthoracic region of her incision.Her deepcultures grew P.acnes.1993Delayed Infections Revisited•Stephens Richards and Emaraduring the primary operation and2)to minimize the amount and bulk of instrumentation being implanted. Intraoperative antibiotics should be routinely used.A wide surgical preparation should be performed,and after draping,the operativefield should be sealed with an io-dine-impregnated adherent plastic barrier.Incisions should never extend to the edge of the drapes because of the likely violation of the barrier.Irrigation should be used frequently during the procedure.Although topical irrigation(bacitracin or polymyxin)has been shown to decrease bacterial growth on cultures obtained intraop-eratively during clean spine surgery,its efficacy is un-proven in the prevention of wound infection.9Use of subcutaneous drains should be considered to decrease hematoma formation.Doing so diminishes the possibil-ity of persistent hematomas or drainage through the in-cision,both of which create favorable bacterial culture media.However,no studies have shown conclusively that the use of drains will prevent infection.Perhaps the most significant problem is the nature, bulk,and modularity of modern implant systems.With these more extensive implants,bursal formation can de-velop over prominent areas of the rod system,particu-larly the crosslinks.In some patients this soft tissue irri-tation likely establishes an environment conducive to growth of these surgically introduced low-virulent or-ganisms,which then develop into clinically apparent de-layed infections.Decreasing the size of the implant and minimizing the modular components(to lessen the po-tential for loosening),all the while maintaining sufficient strength of the construct,should be goals addressed by surgeons and spinal instrumentation companies alike. With this belief that less instrumentation might decrease the delayed infection rate without altering the surgical results,the concept of single-rod instrumentation evolved at our institution.12Initial mechanical testing found the single-rod construct withfixation at every level to be as resistant to torsional stress as a double-rod con-struct over10vertebrae.Unfortunately,an unacceptable rate of instrumentation failure and pseudarthrosis oc-curred with the use of a single rod,and this is no longer used.Continued efforts to decrease the amount and size of instrumentation are being made.A second potential pathway leading to the delayed infections includes hematogenous seeding and has been described previously by Heggeness et al.6The organisms responsible for the delayed infections in their study were also P.acnes and S.epidermidis.Although we believe the evidence strongly suggests that these patients with late-developing drainage(even those with negative cultures)represent delayed bacterial infec-tions,this opinion is not universally shared.Initially,there were reports that the late drainage represented an aseptic result many of the patients were culture negative at72 hours,and the process was thought to be noninfectious. Wimmer and Cluch13continue to believe that late drain-age,without pyogenic appearance,is an inflammatory re-sponse to metal corrosion secondary to micromotion be-tween components.They reported delayed drainage in6of 94patients who had undergone Cotrel–Dubousset instru-mentation for idiopathic scoliosis.None of their patients had positive cultures despite incubation periods of10days. Based on theirfindings,their conclusion that these episodes of delayed drainage represent an aseptic process is under-standable.However,with the high yield of positive cultures found in our study,we remainfirm in our conclusion that the delayed drainage is most likely representative of bacte-rial infection.Once late-developing drainage orfluctuance has de-veloped,the most important step in its treatment is re-moval of all components of the instrumentation.This is the only way in which the infection can be eliminated because the organisms remain protected by a biofilm (glycocalyx)that is adherent to the rods.5Antibiotics and the body’s immune system cannot effectively penetrate this barrier.When the soft tissue and bone surrounding the implants are inspected,it is clear that these deep infections represent soft tissue infections and not osteo-myelitis.The fusion mass of bone around the rods is viable and never is sequestrum seen.Once the instrumen-tation has been removed,primary closure of the wound at the same operative setting is effective and well toler-ated by the patient.In the14patients who had primary closure in this report,there were no complications re-lated to the wound.Delayed primary closure appears to be equally effective,but the patient is subjected to addi-tional procedures2–4days after thefirst operation.It appears that delayed primary closure is unnecessary in achieving successful wound healing.This conclusion was further substantiated by Clark and Shufflebarger.2 The two organisms responsible for the majority of the deep infections in this and in other series are P.acnes and S.epidermidis.These two skinflora have been clearly recognized in the recent literature as being causative of implant-related infections.1,2,8,10,11,14P.acnes requires an extended period of incubation before identification;final culture results in this clinical setting should not be reported out at72hours.After5days of incubation there may be several colonies in the anaerobic or thioglycolate culture.These colonies will then need to be replated,iso-lated,and interpreted.Its identification may require up to 10days.P.acnes should not be dismissed as a contaminant.The postoperative role of antibiotics is becoming bet-ter defined.The duration of parenteral antibiotics in this report ranged from2to14days.The duration of oral antibiotics thereafter ranged from1to9weeks.Thosea clinical outcome as those receiving longer periods of antibiotic treatment.It appears reasonable that paren-teral antibiotics be given for 2–5days followed by oral antibiotics for a period of 7–14days.This conclusion is also further substantiated by the Clark and Shufflebarger study.2They used parenteral antibiotics for 2–3days fol-lowed by oral antibiotics for 10days and achieved excellent wound healing.First-generation cephalosporins (cefazolin followed by cephalexin)are quite effective in the treatment for P.acnes and S.epidermidis (if not methicillin-resistant).Because identification of these organisms may take up to 10days,antibiotics should be used after surgery even in the absence of early positive cultures.In summary,the incidence of delayed deep wound infection was found to be 4.7%in a large population with idiopathic scoliosis who were treated with TSRH instrumentation.Low-virulent skin flora (particularly P.acnes and S.epidermidis )continue to be responsible for the majority of these delayed deep wound infections.Af-ter instrumentation removal,primary wound closure at the same time is satisfactory in achieving excellent heal-ing of the wound.First-generation cephalosporins are effective in eliminating the offending organism once the instrumentation has been removed.Short-term postop-erative parenteral antibiotics followed by short-term oral antibiotics are recommended.1.Brook I,Frazier EH.Infections caused by Propionibacterium species.Rev Infect Dis 1991;13:819–22.2.Clark CE,Shufflebarger te-developing infection in instrumented idio-pathic scoliosis.Spine 1999;24:1909–12.3.Dietz FR,Koontz FP,Found EM,et al.The importance of positive bacterial cultures of specimens obtained during clean orthopaedic operations.J Bone Joint Surg Am 1991;73:1200–7.4.Dubousset J,Shufflebarger H,Wenger te “infection”with CD instru-mentation.Orthop Trans 1994;18:121.5.Gristina AG,Costerton JW.Bacterial adherence to biomaterials and tissue:the significance of its role in clinical sepsis.J Bone Joint Surg Am 1985;67:264–73.6.Heggeness MH,Esses SI,Errico T,et te infection of spinal instrumen-tation by hematogenous seeding.Spine 1993;18:492–6.7.Richards BS,Herring JA,Johnston CE,et al.Treatment of adolescent idio-pathic scoliosis using Texas Scottish Rite Hospital (TSRH)instrumentation.Spine 1994;19:1598–605.8.Richards BS.Delayed infections following posterior spinal instrumentation for the treatment of idiopathic scoliosis.J Bone Joint Surg Am 1995;77:524–9.9.Savitz SI,Savitz MH,Goldstein HB,et al.Topical irrigation with polymixin and bacitracin for spinal surgery.Surg Neurol 1998;50:208–12.10.Schofferman L,Zucherman J,Schofferman J,et al.Diphtheroids and associ-ated infections as a cause of failed instrument stabilization procedures in the lumbar spine.Spine 1991;16:356–8.11.Viola RW,King HA,Adler SM,et al.Delayed infection after elective spinalinstrumentation and fusion.Spine 1997;22:2444–51.12.Wattenbarger JM,Richards BS,Herring JA.A comparison of single-rodinstrumentation with double-rod instrumentation in adolescent idiopathic scoliosis.Spine 2000;25:1680–8.13.Wimmer C,Cluch H.Aseptic loosening after CD instrumentation in thePoint of ViewFrederick R.Dietz,MDI was asked to write this Point of View commentary because of a small study I performed concerning the sig-nificance of positive cultures in clean orthopedic opera-tions.The study I performed consisted of swabbing and biopsying the base of the wound after the approach in 40orthopedic operations of various types.We found posi-tive cultures in 58%of the cases.The vast majority of these positive cultures were coagulase-negative Staphy-lococcus.The next most common group was Propi-onibacterium acnes .All of the positive coagulase-negative Staphylococcus cultures had fewer than five colonies or grew in broth only.P.acnes was the only。