A new clinical practice guideline from the Pediatric Infectious Diseases Society and the Infectious Diseases Society of America provides recommendations for the diagnosis and management of acute hematogenous osteomyelitis (AHO) in otherwise healthy North American children ages 1 month to 18 years.
The guideline, available at https://academic.oup.com/jpids/article/10/8/801/6338658, was developed by a committee comprised of physicians with expertise in pediatric infectious diseases, pediatric hospital medicine, general pediatrics, pediatric emergency medicine, pediatric orthopedic surgery and epidemiology. AAP sections and committees also provided input.
Recommendations are labeled as “strong” or “conditional” with further classification as supported by very low, low, moderate or high quality evidence based on the GRADE (Grading of Recommendations Assessment, Development and Evaluation) approach.
Blood culture and C-reactive protein (CRP) are recommended as part of the initial evaluation of children with suspected AHO. Blood cultures yield the causative bacterium in approximately one-third of cases (median positivity 34.2%; range 11% to 48.5%) and typically are positive within 12-14 hours. While serum CRP has low accuracy in establishing the diagnosis of AHO, the initial serum CRP can serve as a baseline value for sequential monitoring. Data were insufficient to recommend procalcitonin.
Complete blood counts (CBCs) often are performed as part of the initial evaluation and can provide adjunctive information for decision-making in children with suspected or confirmed AHO. However, the committee recognized that CBCs generally do not help discriminate AHO from other potential causes of a child’s symptoms or signs.
Plain radiography can help identify or exclude other important diagnoses (strong recommendation to perform; moderate certainty of evidence). MRI is suggested over bone scan to confirm the diagnosis of AHO (conditional recommendation; very low certainty of evidence).
Empiric antibiotic therapy should be started immediately in children with presumed AHO who appear ill (strong recommendation; moderate certainty of evidence) but can be deferred in those who are not clinically ill if bone aspiration or biopsy is planned (conditional recommendation; very low certainty of evidence).
Staphylococcus aureus is the most common bacterial pathogen overall, though Kingella kingae predominates in younger children. Those with underlying hemoglobinopathies also are at risk for Salmonella spp. infection.
No controlled trials have assessed the efficacy of empiric regimens that provide coverage for methicillin-resistant S. aureus (MRSA) vs. activity against only methicillin-susceptible S. aureus. Given the wide geographic variation in the prevalence of methicillin resistance among S. aureus isolates, the committee felt that selection of an empiric antibiotic agent should be informed by local antibiotic susceptibility patterns.
In areas where rates of community-acquired MRSA are low (i.e., less than 10% of all community S. aureus isolates), empiric cefazolin, oxacillin or nafcillin is reasonable. In area with higher MRSA prevalence, empiric therapy with clindamycin or vancomycin may be more appropriate. The duration of recommended therapy is three to four weeks.
For the majority of children who will respond well to initial intravenous therapy, conversion to oral therapy at or prior to discharge is recommended. This recommendation places high value on avoidance of harms and costs associated with prolonged intravenous therapy and the proliferation of well-tolerated oral antibiotic options. The switch to oral antibiotics can occur even if 1) an organism has not been identified so long as the child has demonstrated improvement in clinical and laboratory parameters and 2) in the presence of S. aureus bacteremia (without venous thrombosis or septic emboli) since children do not carry the same risk of occult endovascular infection as adults.
In addition to assessing clinical response (e.g., resolution of fever, reduction in pain and swelling, improvement in musculoskeletal function), decrease in serially monitored CRP values (by 50% or to values of 2-3 mg/dL) also can inform decisions regarding switch to oral therapy and hospital discharge. In uncomplicated AHO, CRP typically peaks two to four days after initiation of appropriate treatment and returns to normal within nine to 12 days.
End-of-therapy imaging is not recommended routinely except in cases with involvement of the physis or complicated AHO (conditional recommendation; very low certainty of evidence).
Dr. Shah is a member of the AAP Committee on Infectious Diseases.