We recently completed the largest to date, prospective and randomized treatment trial on non-neonatal OM (n = 107), SA (n = 130) and OM+SA (n = 24) in Finland.[9–11] All cases were culture-positive. Our general aim was to simplify the treatment of these usually so intricately handled diseases in children who were primarily healthy and representative of those usually encountered in an industrialized country. Patients considered high risk to treatment failure – anemic, immunodeficient or with severe underlying illnesses – were excluded as among these groups antibiotic therapy may have to be tailored individually. The treatment scheme is summarized in Figure 1.
Simplified treatment scheme for childhood bone and joint infections. AB: Antibiotic; CLI: Clindamycin; CPH: First generation cephalosporin; CRP: C-reactive protein; D: Day; IV: Intravenous; OM: Osteomyelitis; OM+SA: Acute combined osteomyelitis and septic arthritis; PO: Per oral; SA: Septic arthritis. Data from [9–11,44].
Once an acute osteoarticular infection was suspected, only a diagnostic sample was taken by (percutaneous) bone or joint aspiration. Further surgery was done only on specific, rather rare conditions. A synovial or bone aspiration was done under general anesthesia to obtain a representative specimen for bacteriology. Since bone or joint culture may have sensitivity as low as 30–70%,[9,10,15] blood culture was taken in all cases.
Our first-line antibiotics were clindamycin or a first-generation cephalosporin (the choice was randomized), using large doses (40 or 150 mg/kg per day, respectively) and a 6-h (q.i.d.) regimen.[9–11] These agents could be used, because MRSA was and still is rare in Finland. We did not prefer staphylococcal penicillins, because they tend to cause diarrhea when given (orally) in large doses. After 2–5 days intravenously, the administration was switched to the oral route, using the same high doses. No serum assays were ever performed, but good compliance was required. The total course of antibiotic lasted 20 days in OM and OM+SA, and 10–14 days in SA with few exceptions to this general rule. Not a single relapse and only two late reinfections in one child occurred during the follow-up of 12 months post-hospitalization.[9–11]
Clindamycin may cause diarrhea in adult patients, but its good tolerability in children has been previously reported.[47,48] We confirm this finding. Instead, rash developed in two out of 99 patients (Figure 2). A causal association could not be documented, but was likely. We intentionally used exceptionally large doses and administered the q.i.d. regimen believing that both of these principles are pivotal when treating osteoarticular infections with a time-dependent antibiotic (such as clindamycin and a cephalosporin). As administering a dose in the middle of the night would have raised problems, we allowed the four daily doses to be given during the child's wake-up time.
A 15-year-old boy with staphylococcal sacroilitis received clindamycin for 12 days (3 days intravenously) and was discharged on day 13. Soon afterwards a rash developed mainly on the trunk; a dermatologist diagnosed drug allergy. After clindamycin was changed to cefadroxil, rash quickly subsided.
Figure published with permission from Heini Pohjankoski.
Although this simplified treatment strategy worked well in our setting, a few caveats are justified. No data exist for newborns, immunocompromised patients, or cases caused by Salmonella spp., but we assume that longer courses are required in those conditions. Also, a 'regular' case may sometimes warrant attention that deviates from the general rules. Warning signs are persisting fever, but even more importantly persistently high CRP ascending on the fourth day of treatment. In this case, a complication should be suspected and antibiotic may have to be changed or prolonged. Also a surgeon may have to be consulted.
Expert Rev Anti Infect Ther. 2011;9(12):1125-1131. © 2011 Expert Reviews Ltd.