To the Editor:

We appreciate the comments of Dr. Yagupsky. One of the main reasons we performed out our study was to assess the validity of the clinical prediction algorithm previously published by Kocher et al. In their study model, peripheral white blood cell count greater than 12,000 cells per cubic millimeter, non-weightbearing status, history or fever (greater than 38.5 degrees Celsius), and erythrocyte sedimentation rate of at least 40 millimeters per hour were demonstrated to be statistically significant predictors between the two diagnoses. We were impressed with the adjusted odds ratios between 14.4 and 38.6 for the 4 predictors and their ability to identify septic arthritis in 99.8% of cases if all 4 were present. If we could validate these 4 variables in our patient population we could improve our ability to accurately identify septic arthritis with the minimal amount of painful interventions. To maximize the ability to compare our results to that of Kocher et al, we used their study definitions, specifically the definitions of “true” septic arthritis, “presumed” septic arthritis, and transient synovitis.

The definition of septic arthritis is imprecise and challenging, and can only be made definitively with positive culture results from the joint fluid. We agree with Dr. Yagupsky that the inclusion of several pathogens is controversial, specifically the coagulase-negative staphylococci and alpha-hemolytic streptococci. The reason for their inclusion into the septic arthritis category was based not only on the culture results, but also the associated findings of a white-blood cell count in the joint fluid of at least 50,000 cells per cubic millimeter. As defined by Kocher et al, based on cell count only, these patients would be classified as “true” or “presumed” septic arthritis.

If we dismissed the culture results as a contaminant, these patients would be re-classified from the “true” to the “presumed” septic arthritis groups, a change which would not have altered the statistical analysis or the ultimate findings of our study. Thus, based on our findings, institution- specific algorithms for septic arthritis of the hip do not appear to be generalized to other medical centers.

Of interest there were two cases, one each of coagulase-negative staphylococci and alpha-hemolytic streptococci, that had positive cultures obtained by percutaneous arthrocentesis in the radiology suite and in the operating room suite. Cultures obtained in the operating room suite were done after formal open anterior approach to the hip with joint fluid obtained by syringe immediately prior to arthrotomy (1 case) or after arthrotomy (1 case).

References:

Kocher MS, Zurakowski D, Kasser JR. Differentiating Between Septic Arthritis and Transient Synovitis of the Hip in Children: An Evidence- Based Clinical Prediction Algorithm. J Bone Joint Surg Am. 1999:81:1662- 70.

To the Editor:

I read with interest the article by Luhmann et al. in which the distribution of a variety of demographic, clinical and laboratory parameters in children with septic arthritis or transient synovitis of the hip was compared (1). The authors found that the two populations differed in terms of history of fever, erythrocyte sedimentation rate values, white blood cell (WBC) count in the synovial fluid and cellular composition of the peripheral blood and joint fluid leukocytes. Of 47 children included in the composite septic arthritis group, 20 had either a positive synovial fluid culture or a positive blood culture and a synovial fluid WBC count > = 50 x 109/L (”true septic arthritis”), and 27 had negative blood and synovial fluid cultures and >=50 x 109 WBC/L of synovial fluid (presumed septic arthritis). Although no explicit criteria for defining a positive bacteriological culture were stated in the Materials and Methods section of the article, examination of the list of organisms isolated from patients in the true septic arthritis population group raises some concerns.

Overall, bacteria of dubious clinical significance, which are normal components of the skin and mucosal flora, were detected in 11 of 20 patients [coagulase-negative staphylococci in seven children and alpha-hemolytic streptococci (“Streptococcus viridans”) in four]. Isolation of these organisms from blood or synovial fluid cultures of immunocompetent individuals without intravenous devices or prosthetic joints is usually indicative of contamination of the specimen and related to the technical difficulties in obtaining blood and synovial fluid aspirates from young and uncooperative patients (2). Obviously, inclusion of children with contaminated (false-positive) cultures in the septic arthritis group, could have only attenuated real differences between patients with truly infected joints and those with transient synovitis. I believe that a allocation of patients in whom coagulase-negative staphylococci or alpha-hemolytic streptococci were isolated in the septic arthritis population should be reconsidered.

References

1. Luhmann SJ, Jones A, Schootman M, Gordon JE, Schoenekker PL, Luhmann JD. Differentiation between septic arthritis and transient synovitis of the hip in children with clinical prediction algorithms. J Bone Joint Surg Am 2004;86A:956-962.

2. Trujillo M, Nelson JD. Suppurative and reactive arthritis in children. Sem Pediatr Infect Dis 1997;8:242- 249.