Predictive Factors for Differentiating Between Septic Arthritis and Lyme Disease of the Knee in Children

Transcription

1 721 COPYRIGHT Ó 2016 BY THE JOURNAL OF BONE AND JOINT SURGERY, INCORPORATED A commentary by Elan J. Golan, MD, an Jeffrey D. Thomson, MD, is linke to the online version of this article at jbjs.org. Preictive Factors for Differentiating Between Septic Arthritis an Lyme Disease of the Knee in Chilren Keith D. Balwin, MD, MSPT, MPH, Christopher M. Brusalis, BA, Afamefuna M. Nuaguba, MD, an Wubhav N. Sankar, MD Investigation performe at The Chilren s Hospital of Philaelphia, Philaelphia, Pennsylvania Backgroun: Differentiating between septic arthritis an Lyme isease of the knee in enemic areas can be challenging an has major implications for patient management. The purpose of this stuy was to ientify a preiction rule to ifferentiate septic arthritis from Lyme isease in chilren presenting with knee pain an effusion. Methos: We retrospectively reviewe the recors of patients younger than 18 years of age with knee effusions who unerwent arthrocentesis at our institution from 2005 to Patients with either septic arthritis (positive joint flui culture or synovial white bloo-cell count of >60,000 white bloo cells/mm 3 with negative Lyme titer) or Lyme isease (positive Lyme immunoglobulin G on Western blot analysis) were inclue. To avoi misclassification bias, uniagnose knee effusions an joints with both a positive culture an positive Lyme titers were exclue. Historical, clinical, an laboratory ata were compare between groups to ientify variables for comparison. Binary logistic regression analysis was use to ientify inepenent preictive variables. Results: One hunre an eighty-nine patients were stuie: 23 with culture-positive septic arthritis, 26 with culturenegative septic arthritis, an 140 with Lyme isease. Multivariate binary logistic regression ientifie pain with short arc motion, history of fever reporte by the patient or a family member, C-reactive protein of >4 mg/l, an age younger than 2 years as inepenent preictive factors for septic arthritis. A simpler moel was evelope that showe that the risk of septic arthritis with none of these factors was 2%, with 1 of these factors was 18%, with 2 of these factors was 45%, with 3 of these factors was 84%, or with all 4 of these factors was 100%. Conclusions: Although septic arthritis of the knee an Lyme monoarthritis share common features that can make them ifficult to istinguish clinically, the presence of pain with short arc motion, C-reactive protein of >4.0 mg/l, patientreporte history of fever, an age younger than 2 years were inepenent preictive factors of septic arthritis in peiatric patients. The more factors that are present, the higher the risk of having septic arthritis. Level of Evience: Diagnostic Level III. See Instructions for Authors for a complete escription of levels of evience. Peer Review: This article was reviewe by the Eitor-in-Chief an one Deputy Eitor, an it unerwent bline review by two or more outsie experts. The Deputy Eitor reviewe each revision of the article, an it unerwent a final review by the Eitor-in-Chief prior to publication. Final corrections an clarifications occurreuring one or more exchanges between the author(s) an copyeitors. Acute bacterial septic arthritis is most often a surgical emergency. However, ifferentiating septic arthritis of the knee from Lyme isease in chilren with symptoms of knee pain an swelling can be very challenging. Clinical presentations of both conitions can be quite similar. Often patients present with fever, joint effusion, irritability, an inability to bear weight on the affecte extremity. In aition, laboratory measurements, incluing erythrocyte seimentation rate (ESR), C-reactive protein (CRP), an peripheral an synovial white bloo-cell (WBC) counts, may be elevate in both conitions 1-4. Disclosure: One author of this stuy (K.D.B.) receive funs from JBJS for manuscript preparation. On the Disclosure of Potential Conflicts of Interest forms, which are provie with the online version of the article, one or more of the authors checke yes to inicate that the author ha a relevant financial relationship in the biomeical arena outsie the submitte work. J Bone Joint Surg Am. 2016;98:

2 722 TABLE I Patient Data Groups Lyme Disease Culture-Negative Septic Arthritis Culture-Positive Septic Arthritis No. of cases* Mean age (yr) 8.03 ± ± ± 1.6 Sex* Male Female *The values are given as the number of patients. The values are given as the mean an the stanareviation. In spite of the overlap in both clinical presentation an laboratory testing, it is imperative that the treating physician makes an accurate iagnosis because each conition has very ifferent implications for treatment. Because pyogenic septic arthritis can result in rapi articular cartilage estruction, emergent irrigation anrainage in the operating room are warrante. In contrast, Borrelia burgorferi, the causative organism in Lyme isease, stimulates an immune complex-meiate inflammatory reaction that oes not enanger the articular cartilage to the same egree, but oes cause systemic immuneriven meical complications 5. Stanar treatment generally involves an appropriate course of antibiotics an observation for other sequelae of systemic Lyme isease. In most peiatric meical centers, confirmatory testing for Lyme isease involves serum Western blot analysis for Lyme immunoglobulin G (IgG), but the results of this testing may take several ays to be finalize. As a result, frontline clinicians are often face with a challenging scenario: a chil with knee pain an effusion an ineterminate serum an synovial laboratory values. In these circumstances, one must weigh the risk of elaye or untreate septic arthritis against the potential for an unnecessary surgical proceure or serial aspirations, which are less than ieal in a chilren s hospital where anxiety at multiple neele sticks is unesirable. In the absence of a quick an accurate test to efinitively iagnose Lyme isease, preiction algorithms that combine multiple clinical an laboratory factors have been evelope to ientify patients at high risk for pyogenic arthritis an to help to guie treatment ecisions 6-9. Although these moels have proven helpful in ifferentiating septic arthritis from transient synovitis of the hip, their utility in ifferentiating septic arthritis from Lyme isease of the knee TABLE II Univariate Analysis of Culture-Positive Septic Arthritis Compare with Lyme Arthritis for Clinical an Laboratory Finings Culture-Positive Septic Arthritis* (N = 23) Lyme Disease* (N = 140) P Value Duration of symptoms () 3.2 ± ± History of antibiotic use 21.7% 6.4% History of fever 73.9% 40.7% Raiographic effusion 78.3% 95% Micromotion tenerness 86.4% 6.3% <0.001 No weight-bearing 61.1% 15% Joint pain 95.7% 100% Recent illness 8.7% 12.9% Triage temperature ( F) ± ± Joint warmth 100% 43.6% <0.001 Serum WBC count ( 10 9 cells/l) 14.5 ± ± 3.2 <0.001 Absolute neutrophil count (cells/mm 3 ) 10,009 ± 6,345 6,126 ± 2,638 <0.001 ESR (mm/hr) 47 ± ± CRP (mg/l) 11.8 ± ± Synovial WBC count (cells/mm 3 ) 128,080 ± 109,714 53,640 ± 38, Synovial neutrophil (%) 89.3 ± ± *The values are given as either the mean an the stanareviation or as the percentage of patients.

3 723 TABLE III Univariate Analysis of Culture-Negative Septic Arthritis Compare with Lyme Arthritis for Clinical an Laboratory Finings Culture-Negative Septic Arthritis* (N = 26) Lyme Disease* (N = 140) P Value Duration of symptoms () 4.5 ± ± History of antibiotic use 23.1% 6.4% History of fever 76.9% 40.7% Raiographic effusion 96.2% 95% Micromotion tenerness 58.8% 6.3% <0.001 No weight-bearing 46.2% 25.0% Joint pain 100% 100% Recent illness 38.5% 12.9% Triage temperature ( F) ± ± Joint warmth 66.7% 43.6% Serum WBC count ( 10 9 cells/l) 11.6 ± ± Absolute neutrophil count (cells/mm 3 ) 5,989 ± 1,905 6,126 ± 2, ESR (mm/hr) 50 ± ± CRP (mg/l) 5.7 ± ± Synovial WBC count (cells/mm 3 ) 137,204 ± 65,415 53,640 ± 38,652 <0.001 Synovial neutrophil (%) 88.6 ± ± *The values are given as either the mean an the stanareviation or as the percentage of patients. in Lyme isease-enemic areas has not been as thoroughly evaluate 10,11. Therefore, the purpose of this stuy was to evelop a preictive moel to help ifferentiate septic arthritis of the knee from Lyme isease of the knee in chilren with knee pain an effusion an to inform the ecision to operate with a higher level of certainty. TABLE IV Univariate Analysis of All Septic Arthritis Compare with Lyme Arthritis for Clinical an Laboratory Finings All Septic Arthritis* (N = 49) Lyme Disease* (N = 140) P Value Duration of symptoms () 3.9 ± ± History of antibiotics use 22.4% 6.4% History of fever 78.7% 40.7% <0.001 Raiographic effusion 89.6% 95% Micromotion tenerness 74.4% 6.3% <0.001 No weight-bearing 52.3% 25.0% Joint pain 98% 100% Recent illness 24.4% 12.9% Triage temperature ( F) ± ± Joint warmth 81.8% 43.6% <0.001 Serum WBC count ( 10 9 cells/l) 13.0 ± ± 3.2 <0.001 Absolute neutrophil count (cells/mm 3 ) 7,915 ± 4,982 6,126 ± 2, ESR (mm/hr) 49.0 ± ± CRP (mg/l) 8.6 ± ± Synovial WBC count (cells/mm 3 ) 133,352 ± 85,886 53,640 ± 38,652 <0.001 Synovial neutrophil (%) 88.9 ± ± *The values are given as either the mean an the stanareviation or as the percentage of patients.

4 724 Materials an Methos We conucte a retrospective stuy of iniviuals younger than 18 years of age who presente to the emergency epartment of a large tertiary care chilren s hospital between 2005 an 2013 an in whom an arthrocentesis was performe for knee joint effusion. The stuy was approve by our institutional boar review prior to ata collection. Stuy patients were ientifie through a query of the laboratory atabase of our electronic meical recor system for all patients in the emergency epartment for whom joint or synovial flui analysis was orere. Ientifie patients were then cross-reference with the laboratory results of Gram-staining, microbial culture, an synovial WBC count to ientify patients with septic arthritis an were cross-reference with the results of the Lyme Western blot analysis to ientify patients iagnose with Lyme isease. The emergency epartment recors of each ientifie patient were reviewe to confirm the iagnosis of either pyogenic septic arthritis or Lyme arthritis an to extract clinical an laboratory ata. To create the cleanest cohort to minimize misclassification bias, we efine septic arthritis as either the presence of a positive joint flui culture irrespective of synovial cell count (culture-positive septic arthritis) or synovial WBC count of >60,000 white bloo cells/mm 3 in a joint with a negative Lyme titer (culture-negative septic arthritis). Lyme isease was efine as the presence of a positive serum Lyme IgG on Western blot base on our reference laboratory efinitions. Patients who i not meet the criteria for a iagnosis of septic arthritis an who i not have a Lyme titer sent within the same emergency epartment evaluation were exclue from the stuy. To avoi misclassification bias, uniagnose knee effusions an joints with both a positive culture an positive Lyme titers were exclue. In patients with multiple peripheral bloo analyses an joint aspirations uring the same hospitalization, we limite our ata analysis to the first peripheral laboratory values an to the initial aspirate sample. For each patient, we recoreata on emographic characteristics, meical comorbiities, historical elements, physical examination finings, an results of laboratory testing, incluing complete bloo-cell count, Lyme serological testing, bloo an joint flui cultures, Gram stains of the joint flui, ESR, an CRP. Statistical Analysis Data were analyze using SPSS version 22.0 (IBM). Descriptive elements of the ata are presente as means (an ranges) an categorical variables are presente as counts (an percentages). Cutoff values for continuous variables, CRP of >4 mg/l an age younger than 2 years, were etermine with receiver operating characteristic curve analysis. Mann-Whitney U tests were use to evaluate ifferences between groups in continuous or orinal variables in cases where the ata were not normally istribute an to ientify variables for multivariate analysis. A chi-square test was use to etermine significance in the case of binary or categorical ata. Level of significance was establishe at a twosie alpha level of p < Variables that significantly influence the iagnosis of septic arthritis were inclue in a binary logistic regression with a stepwise backwar elimination metho (with a cutoff for elimination of 0.1) to etermine factors preictive of septic arthritis. Simpler ichotomize moels were generate using the risk factors an were confirme using 22 log likelihoo methoology. Moels were generate for Lyme isease compare with culture-positive septic arthritis an for Lyme isease compare with all cases of septic arthritis. Patient Demographic Characteristics During the stuy perio, a total of 841 chilren were ientifie as having ha an arthrocentesis performe in the emergency epartment, of which 498 were knee aspirations. After cross-referencing ientifie patients with the results of the microbial culture an Lyme isease immunoblot assays an applying further exclusion criteria (Fig. 1), we ientifie189patientsasourstuy cohort (49 with septic arthritis an 140 with Lyme isease) (Table I). Of the 23 patients with culture-positive septic arthritis, organisms inclue methicillin-susceptible Staphylococcus aureus (MSSA) (11 patients), methicillin-resistant Staphylococcus aureus (MRSA) (4 patients), Propionibacterium acnes (2 patients), Clostriium bifermentans (1 patient), Streptococcus pneumoniae Fig. 1 Consoliate Stanars of Reporting Trials (CONSORT) iagram of the stuy population. SA = septic arthritis, CNSA = culture-negative septic arthritis, Dx = iagnosis, WB1 = positive white bloo-cell count, an Hx = meical history.

5 725 TABLE V Univariate Analysis of Operatively Treate Lyme Arthritis Compare with Nonoperatively Treate Lyme Arthritis Operative Group* (N = 46) Nonoperative Group* (N = 94) P-Value Temperature ( F) 99.3 ± ± Serum WBC count ( 10 9 cells/l) 11.2 ± ± Absolute neutrophil count (cells/mm 3 ) 5,727 ± 2,122 5,726 ± 2, ESR (mm/hr) 52 ± ± 21 <0.001 CRP (mg/l) 4.3 ± ± Synovial WBC count (cells/mm 3 ) 89,304 ± 38,829 36,396 ± 24,117 <0.001 *The values are given as the mean an the stanareviation. TABLE VI Multivariate Analysis for 4 Inepenent Preictors Multivariate Preictor Ajuste Os Ratio* P Value History of fever 6.1 (1.2 to 31.7) Short arc pain 67.3 (11.7 to 389.1) <0.001 CRP 1.2 (1.0 to 1.3) Age 0.6 (0.5 to 0.8) *The values are given as the ajuste os ratio, with the 95% CI in parentheses. TABLE VII Algorithm for the (Exact) Probability of Septic Arthritis Factors Present Preictive Probability of Septic Arthritis 0 2% 1 18% 2 45% 3 84% 4 100% (1 patient), Corynebacterium species (1 patient), Group A Streptococcus (1 patient), Moraxella catarrhalis (1 patient), an Serratia marcescens (1 patient). Results Tables II, III, an IV illustrate the ifferences an similarities between the cohorts of septic arthritis (culturepositive, culture-negative, an all cases) an immunologically confirme Lyme isease in clinical presentation an laboratory values. All patients with confirme septic arthritis unerwent emergent incision anrainage of the knee in the operating room. Of the patients who were eventually iagnose with Lyme isease, 46 (33%) of 140 patients unerwent surgical incision anrainage base on a presumeiagnosis of septic arthritis. Patients with Lyme arthritis who unerwent incision anrainage iffere significantly from those who i not unergo a surgical proceure in terms of mean synovial WBC count (89,304 compare with 36,396 cells/mm 3, p < 0.001), serum WBC count ( cells/l compare with cells/l, p = 0.010), ESR (52 compare with 37 mm/hr, p < 0.001), an CRP (4.3 compare with 3.4 mg/l, p = 0.01) (Table V). An elevate synovial WBC count was the variable most consistent with receiving surgical incision anrainage in the setting of Lyme arthritis; only 5 of the 46 patients with Lyme isease who unerwent a surgical proceure ha cell counts below 50,000 cells/mm 3. On multivariate analysis, we ientifie 4 inepenent factors for ifferentiating between septic arthritis an Lyme isease: subjective history of fever reporteuring the emergency epartment encounter by the patient or a family member, severe pain with <30 of motion (pain with short arc motion), CRP of 4 mg/l, an age uner 2 years; ajuste os ratios, p values, an 95% confience intervals (95% CIs) were calculate (Table VI). The probability of septic arthritis with any one factor present was 18% compare with 100% with all 4 factors present (Table VII). Similar risk factors were foun for the culture-positive septic arthritis subgroup. Of note, short arc motion pain was even more strongly associate with culturepositive septic arthritis than in the full moel (os ratio, [95% CI, 14.6 to 2,028]). Discussion In a geographic region with a high prevalence of Lyme isease, istinguishing septic arthritis of the knee from Lyme isease is challenging but essential as the two conitions have profounly ifferent implications for treatment an possible complications in the skeletally immature patient. Although Lyme isease can be treate with a course of oral antibiotics, septic arthritis requires prolonge antibiotic therapy an urgent incision anrainage to prevent long-term complications, incluing articular cartilage amage, osteonecrosis, concomitant osteomyelitis, an systemic sepsis The similarities in clinical an laboratory features of pyogenic septic arthritis an Lyme arthritis 1-4,11,15 create iagnostic challenges. Arthrocentesis for synovial cell count an culture can facilitate ecision-making, but are noniagnostic in many cases 16. Aitionally, our ata show that reliance on cell

6 726 count alone may lea to a high rate of unnecessary surgical proceures. Furthermore, the results of Lyme-specific serological testing can take several ays an therefore are not helpful for making an urgent surgical ecision. As a result, Lyme isease is often mistaken for culture-negative septic arthritis an is treate as such with incision anrainage, unnecessarily exposing these patients to the risks of a surgical proceure 4. In the absence of an easily aministere test with high sensitivity an specificity for efinitively iagnosing septic arthritis, previous stuies have use multiple physical examination an laboratory factors to create preiction moels that help to ientify patients at high risk for pyogenic arthritis. Kocher et al. 8 evelope a wiely use set of criteria to ifferentiate septic arthritis from transient synovitis of the hip using 4 inepenent clinical variables: history of fever, non-weightbearing status, ESR of 40 mm/hr, an serum WBC count of cells/l. This moel has since been valiate internally an externally, an upate to reflect the aitional iagnostic benefit of CRP 6. Although the Kocher moel is often applie to multiple joints even in the setting of Lyme isease, this preictor was not originally intene to istinguish between septic arthritis an Lyme isease an was not valiate for the knee. The septic arthritis cohort stuie by Kocher et al. consiste of all patients with synovial leukocyte counts in excess of 50,000 cells/mm 3 regarless of bacterial culture results an therefore may have inavertently inclue patients with Lyme isease an other misclassification biases. Inee, a recent stuy by Deanehan et al. emonstrate synovial WBC counts in excess of 50,000 cells/mm 3 to be quite common in Lyme arthritis 16, an these finings were reiterate by this current stuy. Three previous stuies have specifically sought to ientify preictive factors for ifferentiating septic arthritis from Lyme isease. Thompson et al. 11 evaluate 179 patients with monoarthritis of ifferent joints (not just the knee) an ientifie history of fever an elevate CRP to be negative preictors of Lyme arthritis an knee involvement to be a positive preictor (moel sensitivity of 88% an specificity of 82%). In spite of this, the authors conclue that there was too much overlap in the ata to create a clinically useful preictive moel. Milewski et al. 15 compare 123 patients with Lyme isease with 51 patients with culture-positive septic arthritis, although several ifferent joints were inclue in their analysis. Base on multivariate analysis, the authors foun that refusal to bear weight was the most preictive factor, with weaker preictors being presence of a fever, WBC count of >12 1,000/mL, an nucleate cell count of >100,000 cells/mm 3. Interestingly, CRP an ESR were not foun to be useful preictors, although the stuy was limite by a very low CRP collection rate uring the stuy perio. A recent stuy by Deanehan et al. is the only previous one, to our knowlege, that compare preictive factors for Lyme isease with those for septic arthritis specifically of the knee 10. Unlike the aforementione stuies (an ours), the authors foun ESR to be an inepenent preictor of septic arthritis in Lyme isease-enemic areas; the other factor that they foun to be preictive was an absolute neutrophil count of 10,000 cells/mm 3, which we also i not etect in our stuy. However, their ata were limite by a low CRP collection rate (77%) an a small number of patients with septic arthritis (19). In aition, the Lyme isease cohort inclue a significant proportion of patients (58%) who i not unergo arthrocentesis. Fig. 2 Map of reporte cases of Lyme isease in the Unite States, (Reprouce from: CDC, National Center for Health Statistics. Lyme Disease Oct 2. Accesse 2016 Jan 26.)

7 727 These latter types of patients can typically be triage to observation by the treating physician an are rarely, by clinical examination, suspecte of having septic arthritis. These patients may therefore not represent the types of clinical presentations that are easily confuse with those having true pyogenic arthritis. In this present stuy aime at ifferentiating septic arthritis of the knee from Lyme monoarthritis, we foun several inepenent preictive factors for pyogenic arthritis: patient or family-reporte history of fever, pain with short arc motion, CRP of 4 mg/l, an age younger than 2 years. The probability of septic arthritis with any one factor present was 18% compare with 100% with all 4 factors present. Logistic regression of preictor variables showe that 87% of cases were correctly preicte with our moel. In our series, a patient with no risk factors ha a 2% chance of being iagnose with septic arthritis, an a patient with 4 factors ha a 100% chance of septic arthritis. Notably, of 46 patients with Lyme isease who unerwent a surgical proceure, 11 patients ha no clinical preictors. Thus, applying this clinical preiction rule may have avoie a surgical proceure in 23.9% of patients in this cohort. Of the remaining 35 patients with Lyme isease who unerwent a surgical proceure, 19 ha one clinical preictor, 14 ha 2 preictors, 2 ha 3 preictors, an no patient ha all 4 preictors. It is important to note that the results of this stuy may not be applicable in a non-enemic area with a lower prevalence of Lyme isease. Interestingly, we foun pain with short arc motion to be the most clinically useful test for septic arthritis, with a sensitivity of 0.78 an a specificity of Multivariate analysis showe that the ajuste os of having septic arthritis with the presence of short arc pain was 67.3, which was many magnitues higher than the os from CRP of 4mg/L,patientreporte history of fever, or age younger than 2 years. Although weight-bearing status is often ocumente in stuies of septic arthritis, to the best of our knowlege, our stuy was the first to show the high accuracy of pain with short arc motion in etermining patients at high risk for septic arthritis. Our institution alreay inclues this examination component as part of our septic arthritis evaluation algorithm. As with other physical examination finings, ientifying short arc pain in chilren younger than 2 years of age can be challenging. At our institution, the use of chil life avocates an working closely with patients guarians have proven to be helpful in lowering patient anxiety. In aition, although analgesic meication frequently lowers chilren s anxiety,it oes not typically affect short arc pain. We highly recommen incorporating this test in other settings where septic arthritis of the knee is suspecte. To provie a clinically useful preiction moel, we efine septic arthritis as either the presence of a positive joint culture or as a synovial WBC count of >60,000 white bloo cells/mm 3 with negative Lyme serological results. Although there is a small risk of misclassification bias using this seconefinition, ignoring the large number of patients who present to emergency epartments with culture-negative septic arthritis woul limit the clinical applicability of any preiction rule. This stuy ha several limitations. First, as a chart review, it ha the potential loss of ata fielity associate with retrospective analysis. Secon, as a single-center stuy in a Lyme isease-enemic area, the results of this stuy may not be applicable to all geographical regions, particularly those with a low prevalence of Lyme isease. The Centers for Disease Control an Prevention (CDC) has reporte that, in 2014, 96% of confirme cases of Lyme isease occurre in 14 U.S. states, concentrate in the Northeast an upper Miwest (Fig. 2) 17. Therefore, the finings of this stuy may be most applicable to these geographic regions. The strengths of the stuy inclue rigorous efinitions of the cohorts anata acquisition an statistical analysis methos that limit misclassification an observer bias. Aitionally, to our knowlege, our stuy is the first to ientify preictive risk factors that, in our cohort, ensure the iagnosis of septic arthritis when all were present (i.e., the iagnosis was septic arthritis in 100% of cases) an resulte in only one case of septic arthritis when all 4 risk factors were absent (2%). In conclusion, our stuy offers a useful preiction algorithm for septic arthritis compare with Lyme isease of the knee. We believe that a patient oler than 2 years of age without history of fever, no short arc pain, an CRP of <4 mg/l may be safely observe in anticipation of serologic stuies. In contrast, patients with all 4 risk factors shoul be consiere for urgent surgical intervention, as they are nearly certain to have septic arthritis. Areas with a lower prevalence of Lyme isease may nee to interpret our results with caution, as other etiologies of knee effusion may be more likely in those environments. n Keith D. Balwin, MD, MSPT, MPH 1 Christopher M. Brusalis, BA 1 Afamefuna M. Nuaguba, MD 1 Wubhav N. Sankar, MD 1 1 The Chilren s Hospital of Philaelphia, Philaelphia, Pennsylvania aress for W.N. Sankar: sankarw@ .chop.eu References 1. Bachman DT, Srivastava G. Emergency epartment presentations of Lyme isease in chilren. Peiatr Emerg Care Oct;14(5): Cristofaro RL, Appel MH, Gelb RI, Williams CL. Musculoskeletal manifestations of Lyme isease in chilren. J Peiatr Orthop Sep-Oct;7 (5): Eichenfiel AH, Golsmith DP, Benach JL, Ross AH, Loeb FX, Doughty RA, Athreya BH. Chilhoo Lyme arthritis: experience in an enemic area. J Peiatr Nov;109(5): Willis AA, Wimann RF, Flynn JM, Green DW, Onel KB. Lyme arthritis presenting as acute septic arthritis in chilren. J Peiatr Orthop Jan-Feb;23 (1): Puius YA, Kalish RA. Lyme arthritis: pathogenesis, clinical presentation, an management. Infect Dis Clin North Am Jun;22(2): , vi-vii. 6. Cair MS, Flynn JM, Leung YL, Millman JE, D Italia JG, Dormans JP. Factors istinguishing septic arthritis from transient synovitis of the hip in chilren. A prospective stuy. J Bone Joint Surg Am Jun;88(6):

8 Jung ST, Rowe SM, Moon ES, Song EK, Yoon TR, Seo HY. Significance of laboratory an raiologic finings forifferentiating between septic arthritis an transient synovitis of the hip. J Peiatr Orthop May-Jun;23(3): Kocher MS, Zurakowski D, Kasser JR. Differentiating between septic arthritis an transient synovitis of the hip in chilren: an evience-base clinical preiction algorithm. J Bone Joint Surg Am Dec;81(12): Molteni RA. The ifferential iagnosis of benign an septic joint isease in chilren. Clinical, raiologic, laboratory, an joint flui analysis, base on 37 chilren with septic arthritis an 97 with benign aseptic arthritis. Clin Peiatr (Phila) Jan;17(1): Deanehan JK, Kimia AA, Tan Tanny SP, Milewski MD, Talusan PG, Smith BG, Nigrovic LE. Distinguishing Lyme from septic knee monoarthritis in Lyme iseaseenemic areas. Peiatrics Mar;131(3):e Epub 2013 Feb Thompson A, Mannix R, Bachur R. Acute peiatric monoarticular arthritis: istinguishing Lyme arthritis from other etiologies. Peiatrics Mar;123(3): Bennett OM, Namnyak SS. Acute septic arthritis of the hip joint in infancy an chilhoo. Clin Orthop Relat Res Aug;(281): Forlin E, Milani C. Sequelae of septic arthritis of the hip in chilren: a new classification an a review of 41 hips. J Peiatr Orthop Jul-Aug;28(5): Gillespie R. Septic arthritis of chilhoo. Clin Orthop Relat Res Oct; (96): Milewski MD, Cruz AI Jr, Miller CP, Peterson AT, Smith BG. Lyme arthritis in chilren presenting with joint effusions. J Bone Joint Surg Am Feb 2; 93(3): Deanehan JK, Nigrovic PA, Milewski MD, Tan Tanny SP, Kimia AA, Smith BG, Nigrovic LE. Synovial flui finings in chilren with knee monoarthritis in Lyme isease enemic areas. Peiatr Emerg Care Jan;30(1): CDC, National Center for Health Statistics. Lyme Disease Oct Accesse 2016 Jan 26.