Title: SERUM C-REACTIVE PROTEIN, ERYTHROCYTE SEDIMENTATION RATE AND WHITE BLOOD CELL COUNT IN SEPTIC ARTHRITIS OF CHILDREN
Abstract: One hundred twenty years ago more than 50% of children with acute septic arthritis died and many were left with persistent sequelae.1 Today the case fatality is generally <1%, but sequelae are still not uncommon.2,3 The traditional laboratory indices for monitoring septic arthritis are the erythrocyte sedimentation rate (ESR) and the white blood cell (WBC) count. ESR is a nonspecific index of inflammation primarily reflecting concentration of fibrinogen and immunoglobulins in the plasma. It usually rises 24 h or more after the onset of symptoms and signs of infection and then slowly returns to normal within about 4 weeks.4 C-reactive protein (CRP) is an acute phase protein; it tends to increase several hundred fold after a triggering stimulus, and the rise may occur within 6 h.5 With a doubling time of about 8 h, peak values are usually reached within 50 h,5 after which CRP concentrations decrease rapidly, if the chosen therapy is correct. Sequential determination of the concentration of CRP has also proved to be of clinical value in differential diagnosis and in monitoring the course of the infections.5,6 CRP has been useful in diagnosis and follow-up in acute osteomyelitis.7,8 To compare the clinical values of CRP, ESR and WBC count in bacteriologically proved acute septic arthritis, we devised the present prospective study. Patients and methods. This investigation is a part of a prospective multicenter follow-up study on pediatric orthopedic infections,7,8 which was launched in 8 large hospitals of Finland in 1982. The study protocol was approved by the ethical committees of the hospitals involved. The series comprised 100 children age 6 months to 14 years (median, 5.6 years; mean, 5.9 years) with bacteriologically confirmed septic arthritis. Septic arthritis was defined by the characteristic signs and symptoms of a joint infection and a positive joint or blood culture or oth. Patients with an underlying chronic illness and those with subacute arthritis (a history of >2 weeks on admission), penetrating wounds or spine infections were excluded. ESR was measured by the conventional Westergren tube method on admission, and on Days 3, 5, 7, 10, 14, 19 and 29 of treatment. CRP was measured quantitatively by the turbidimetric or nephelometric method on the same days as ESR and in addition on Days 2, 9, 12, 17 and 23. The value of 19 mg/l was taken as the upper normal value for CRP and 19 mm/h was chosen for ESR. WBC count was examined on admission and on Days 5, 10, 19 and 29. Values <12 × 109/l were regarded as normal. A body temperature exceeding 37.5°C (99.5°F) was considered fever. Blood cultures were taken before antimicrobial treatment was instituted. A first generation cephalosporin or clindamycin were the antimicrobics used, and the treatment was initiated intravenously but switched to oral within a few days. The mean duration of medication was 23 days. According to the length of history, the patients were divided into two groups: the "short history group" (45% of the series) comprised those in whom symptoms and signs had persisted for 48 h or less before diagnosis; the remaining patients comprised the "long history group" (55%). Results.Staphylococcus aureus was responsible for 62 of 100 cases, Haemophilus influenzae type b for 25 cases, group A streptococci for 6 cases, pneumococcus for 5 cases and group B meningococcus for 2 cases. The culture obtained locally from the joint was positive in 80% of cases, blood culture was positive in 63% cases and both cultures were positive in 43% cases. The hip joint was affected in 36 cases, the ankle in 27, the knee in 22, the humeroscapular joint in 8 (8%) and the elbow in 5. Polyarthritis was found in 2 cases. ESR was increased in 90% of cases, the mean value being 56 mm/h (Fig. 1). ESR continued to increase so that the peak was reached on Day 5 (mean 79 mm/H). The ESR values (mean ± SD) on admission and on Days 3, 5, 7, 10, 14, 19 and 29 were 55 ± 28, 78 ± 27, 79 ± 33, 68 ± 32, 63 ± 37, 54 ± 35, 37 ± 25 and 21 ± 16, respectively.Fig. 1: CRP and ESR values (means ± SEM) in 100 cases of childhood septic arthritis, during a 1-month follow-up. Stippled area, normal range of values.On admission, CRP was elevated in 95% of cases: the mean value was 85 mg/l (Fig. 1). The peak values were reached on Day 2 (mean 112 mg/l). The CRP values (mean ± SD) on admission and on Days 2, 3, 5, 7, 9, 10, 12, 14, 17, 19, 23 and 29 were 85 ± 49, 112 ± 57, 98 ± 63, 58 ± 42, 36 ± 37, 22 ± 23, respectively, and normal (<20) on Day 10 and thereafter, respectively. Only 60% of the children had leukocytosis on admission. The WBC values (mean ± SD) on admission and on Days 5, 10, 19 and 29 were 15 ± 6.7, 9.3 ± 3.5, 8.7 ± 2.6, 8.1 ± 3.6 and 8.3 ± 3.4, respectively. After reaching the highest levels on the second day of treatment, CRP commenced a rapid normalization, and values at 19 mg/l or less were reached in 9 days (mean, 8.8 days). This was in contrast to ESR which required ∼5 weeks to normalize. The mean time to reach a level at 19 mm/h or less was 34 days for ESR and in 26 patients normalization took >29 days. The difference between normalization time of CRP and ESR was significant (P < 0.0001). Defervescence occurred, on average, in 4.2 days (range, 0 to 20 days). The mean duration of illness before the start of antimicrobial treatment was 3 days (range, 1 to 7; median, 4 days). In the short history group the initial mean values for CRP, ESR and WBC were 76 mg/l, 43 mm/h and 16.1 × 109/l, respectively. The peak value for ESR was reached on Day 5 and for CRP on Day 2 (P < 0.01). In the long history group ESR and CRP were more elevated on admission, the mean values being 93 mg/l for CRP, and 64 mm/H for ESR. In contrast, the mean WBC count was only 14.1 × 109/l. In this group also, highest values were reached faster for CRP than for ESR (Day 2 vs. Day 5, P < 0.001). Whereas the ESR values were significantly higher in the long history group on admission (P < 0.01), the groups did not differ in CRP values. The WBC counts were slightly higher in the short history group (P not significant). In the short history group 30% had normal WBC counts (<12.0 × 109/l), and in the long history group 53% had normal WBC counts on admission. Fever lasted longer in the short history group but no other differences were found in the indices investigated. Discussion. The concentration of serum CRP, an acute phase protein, initially rose and thereafter decreased rapidly during the course of bacterial arthritis in children. Behavior of CRP in septic arthritis was the same as in other bacterial diseases such as acute osteomyelitis and bacterial meningitis.6-8 With a doubling time of 8 h peak values of CRP are reached within 2 days5, after which concentration decreases rapidly. On the contrary ESR rises slowly during several days and then normalizes slowly during several weeks. The question is, however, what benefit does physician achieve when using CRP determinations in addition to ESR (and, to a lesser extent, WBC) in diagnosis and especially follow-up of septic arthritis in children. There are three important reasons for using CRP in addition to ESR determinations. First, CRP is a simple test carried out with readily available laboratory equipment, nephelometer or turbidimeter, and it requires only a finger prick sample of blood. Ease of sampling and determination of CRP means that it can be done frequently during the acute phase of illness. The CRP values should be measured quantitatively, and the test result must be obtained on the same day, because the information reflects the actual situation of infection more reliably than any other simple laboratory test. Measuring CRP from a finger prick is even easier now that an apparatus is available in which whole blood, instead of serum, is used without the need for centrifugation.9 Secondly, CRP increases and decreases much more quickly than ESR (Fig. 1). CRP was clearly elevated on admission, peaked within 48 h and normalized in about 1 week. In contrast ESR continued increasing, despite active treatment, until Day 5 and required more than 1 month to normalize. The rapidity of CRP is of a major clinical relevance because CRP also has the potential for detecting complications sooner than other laboratory indices we are aware of; CRP reacts sensitively to commencing complications in bacterial diseases.6,10 If CRP is decreasing inflammation is likely to be subsiding, even though the clinical symptoms and signs of the patient have not yet started to diminish. In case of doubt CRP could be retested after 12 h, and if the concentration of CRP is decreasing, resolution of infection is even more likely. Thirdly, because CRP is an acute phase protein, it is already markedly elevated 24 h after the beginning of clinical symptoms. On admission to the hospital there was no significant difference in CRP levels in children whose symptoms and signs had lasted either longer or shorter than 48 h. On the contrary the ESR levels were clearly lower in the short history group. Because of the dissimilar kinetics of CRP and ESR, we recommend determination of both indices initially on admission to hospital. Then CRP could be measured quantitatively every 24 to 48 h until normal values (<20 mg/l) are reached. If CRP values remain high or especially if they increase again during therapy, it may be a warning sign of a new activation of the inflammatory process.10 Should this happen thorough examination of the patient and careful evaluation of the antibiotic therapy is very important. Markku J. T. Kallio, M.D. Leila Unkila-Kallio, M.D. Kari Aalto, M.D. Heikki Peltola, M.D. Children's Hospital (MJTK, KA, HP) Departments I and II of Obstetrics and Gynecology (LUK) University of Helsinki; Helsinki, Finland