Copyright © 2006 by The Journal of Bone and Joint Surgery, Inc.
Commentary & Perspective
Commentary & Perspective by
James R. Kasser, MD
Department of Orthopaedic Surgery, Children's Hospital, Boston, MA
This article is an important contribution because it tests the validity of treatment algorithms for the differentiation of septic arthritis from transient synovitis. The initial work by Kocher et al. in this area, at Children's Hospital in Boston, defined four factors which, when combined, were instrumental in decision-making, forming a treatment algorithm1. The four factors in Kocher's study were (1) refusal to bear weight, (2) fever >38.5°C, (3) serum white blood-cell count of >12,000/mm3 (>12.0 × 109/L), and (4) erythrocyte sedimentation rate of >40 mm/hr. Subsequently, Luhmann et al., in a retrospective study, found the validity of this treatment algorithm not sufficient to be of great benefit in clinical use2.
However, a subsequent prospective study at Children's Hospital in Boston again demonstrated the validity of this treatment algorithm3. During the period of use of the algorithm, C-reactive protein has become a routine index of inflammation and has been found to be much more accurate than sedimentation rate. Therefore, the use of C-reactive protein in the treatment algorithm presented by Caird et al. would be expected to improve the accuracy of this algorithm, as is the case.
In this prospective study of forty-eight patients, fourteen patients had transient synovitis and thirty-four had either septic arthritis or presumed septic arthritis. The organisms involved were predominantly Staphylococcus aureus, (ten patients), Streptococcus pneumoniae (two patients), or Staphylococcus epidermidis (two patients). The diagnosis of confirmed septic arthritis was based on positive blood culture or positive culture from synovial fluid.
Lyme arthritis was differentiated from septic arthritis based on serologic profile. Univariate and multivariate analyses were performed. Patients with septic arthritis differed significantly different from those with transient synovitis with regard to fever, erythrocyte sedimentation rate, and C-reactive protein level (p < 0.05). No patient with transient synovitis had a fever that was >38.5°C at any time during the hospital stay. The C-reactive protein level was believed to have the highest predictive value and was the only risk factor strongly associated with outcome at a 5% significance level.
In the study by Kocher et al., children with no predictors for septic arthritis had a 0.2% chance of having this diagnosis, whereas those with all four factors positive had a 99.6% chance of having septic arthritis3. In this study, the predicted probability of septic arthritis with all predictors positive was 98%.
In summary, when clinical algorithms are developed, they must be tested in multiple settings to ensure that they are valid. It is clear from this study from Children's Hospital of Philadelphia that the addition of a C-reactive protein level >2.0 mg/dL is a substantial addition to the previously developed algorithm by Kocher et al.
I believe that clinical algorithms must be developed to supplement decision-making and to aid in accuracy of diagnosis. In addition to the laboratory tests and the findings stated, a comprehensive physical examination is of great value. The simple observation of "refusal to walk" is not a satisfactory substitute for an actual physical examination showing limitation of motion and spasm. The difficulty in developing a clinical prediction guideline is in quantifying a physical finding so that it is reproducible from one examiner to another. Therefore, the algorithms do not include physical examination, and this may be one of the greatest criticisms of such a diagnostic protocol. Clearly, observation of the patient and careful physical examination are part of the required evaluation for the purpose of making appropriate management decisions for patients. The diagnostic algorithm should supplement clinical decision-making in all cases, and, based on this study as well as that by Kocher et al.3, the validity of this rule is confirmed.
*The author did not receive grants or outside funding in support of his research for or preparation of this manuscript. He did not receive payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution, or other charitable or nonprofit organization with which the author is affiliated or associated.
References
1. 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.
2. Luhmann SJ, Jones A, Schootman M, Gordon JE, Schoenecker 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;86:956-62.
3. Kocher MS, Mandiga R, Zurakowski D, Barnewolt C, Kasser JR. Validation of a clinical prediction rule for the differentiation between septic arthritis and transient synovitis of the hip in children. J Bone Joint Surg Am. 2004;86:1629-35.
|
