Bone and joint infections are challenging therapeutic situations requiring rapid antibiotic therapy as soon as bacteriology specimens have been obtained. Laboratory tests (C reactive protein, erythrocyte sedimentation rate, white cell count) and clinical findings are used to assess therapeutic efficacy. Most of the clinical signs however, particularly in children or after a surgical procedure, are not explicit enough to allow proper assessment of the clinical course under antibiotic therapy. Body temperature is the only parameter currently used in routine practice. But the measurement of body temperature is not always reliable and variations observed during treatment should not always be attributed to treatment failure. The purpose of this work was to assess the significance of changes in body temperature observed in children given effective intravenous antibiotic treatment for bone and joint infections.

We reviewed retrospectively the files of 60 children treated in our unit for acute bone and joint infections. The patients had acute hematogeneous osteomyelitis (n = 27), septic arthritis (n = 25), and infectious osteoarthritis (n = 8). A bacterial strain was identified on cultures of blood, joint fluid, or metaphysis puncture samples in all cases. Blood test results (C-reactive protein, erythrocyte sedimentation rate, white cell count) were recorded during treatment. Body temperature was recorded three times a day until normalization then daily until discharge. We searched for correlations between variations in the temperature curve observed during treatment and blood test results.

Ninety percent of the children had fever at admission (mean 39.1 °C). Among the six children without fever, the temperature rose in 5 during the first 48 hours of hospitalization. Even when the treatment was effective, apyrexia was achieved slowly, on the average after 8 days of antibiotic treatment. We also observed that the peak temperature occurred during the first 5 days of antibiotic treatment considered effective. C-reactive protein level normalized within a satisfactory time (10.5 days), reflecting the efficacy of the antibiotics.

The efficacy of antibiotic treatment must always be verified in patients with acute bone and joint infections. Generally, biological parameters are used to monitor efficacy. Currently, C-reactive protein appears to be the most reliable parameter to assess efficacy, its rapid decline reflecting clinical cure. Erythrocyte sedimentation rate and white cell counts are poor surveillance parameters. Finally, body temperature is not a specific surveillance parameter and persistent fever during treatment does not necessarily signify ineffective antibiotic treatment. In light of this fact, body temperature should always be compared with C-reactive protein level to draw any conclusion concerning therapeutic failure.