THE USEFULNESS OF C-REACTIVE PROTEIN-LEVELS IN THE IDENTIFICATION OF CONCURRENT SEPTIC ARTHRITIS IN CHILDREN WHO HAVE ACUTE HEMATOGENOUS OSTEOMYELITIS - A COMPARISON WITH THE USEFULNESS OF THE ERYTHROCYTE SEDIMENTATION-RATE AND THE WHITE BLOOD-CELL COUNT

Thirty-six children who had bacteriologically confirmed acute hematoge nous osteomyelitis but did not have concurrent septic arthritis, and t en children who had confirmed acute hematogenous osteomyelitis and con current septic arthritis, were followed for one year to compare the ch anges in the C-reactive protein level in the blood, the erythrocyte se dimentation rate, and the white blood-cell count. In both groups, the mean C-reactive-protein values were high (eighty-four milligrams per l iter in the children who had septic arthritis and osteomyelitis and si xty-five milligrams per liter in those who had osteomyelitis only) at the time of admission to the hospital. However, in the group that had septic arthritis, the increase was significantly higher (p < 0.01) as early as the second day and a normal level (less than twenty milligram s per liter) was reached significantly later (p < 0.001) than in the g roup that had osteomyelitis only (11 +/- 7 days compared with 6 +/- 3 days [mean and standard deviation]). The erythrocyte sedimentation rat e showed the same tendency, but the difference in the rates between th e groups did not become evident until the fifth to fourteenth days aft er admission. A normal erythrocyte sedimentation rate (less than twent y millimeters per hour) was reached in 25 +/- 12 days in the children who had septic arthritis and in 17 +/- 10 days in those who did not (p < 0.05). In the group that had septic arthritis, the white blood-cell count was significantly higher (p < 0.001) on admission and, except f or one measurement period, a significant difference persisted between the groups thereafter; however, the levels were within the normal rang e (<12 x 10(9) per liter, or <12,000 per cubic millimeter(3)), and the overlap between the groups was so great that the white blood-cell cou nt was not a useful value with which to distinguish between the groups . If a child who has signs and symptoms of acute hematogenous osteomye litis initially has a moderate leukocytosis and an elevated C-reactive -protein value that increases almost twofold within twenty-four hours, the possibility of osteomyelitis with associated arthritis should be borne in mind. In children who have osteomyelitis without septic arthr itis, the C-reactive protein level then usually decreases rapidly; if it does not, associated septic arthritis should be suspected. We found that when the level of the C-reactive protein on the third day was mo re than 1.5 times the level at the time of admission to the hospital, the likelihood ratio that septic arthritis was also present was 6.5. C hanges in the erythrocyte sedimentation rate gave the same information , but later.