Hans de Graaf, Priya Sukhtankar, Barbara Arch, Nusreen Ahmad, Amanda Lees, Abigail Bennett, Catherine Spowart, Helen Hickey, Annmarie Jeanes, Kate Armon, Andrew Riordan, Jethro Herberg, Scott Hackett, Carrol Gamble, Delane Shingadia, Ann Pallett, Stuart C Clarke, Philip Henman, Marieke Emonts, Mike Sharland, Adam Finn, Andrew J Pollard, Colin Powell, Peter Marsh, Claire Ballinger, Paula R Williamson, Nicholas MP Clarke, and Saul N Faust
Background: There is little current consensus regarding the route or duration of antibiotic treatment for acute osteomyelitis (OM) and septic arthritis (SA) in children. Objective: To assess the overall feasibility and inform the design of a future randomised controlled trial (RCT) to reduce the duration of intravenous (i.v.) antibiotic use in paediatric OM and SA. Design: (1) A prospective service evaluation (cohort study) to determine the current disease spectrum and UK clinical practice in paediatric OM/SA; (2) a prospective cohort substudy to assess the use of targeted polymerase chain reaction (PCR) in diagnosing paediatric OM/SA; (3) a qualitative study to explore families’ views and experiences of OM/SA; and (4) the development of a core outcome set via a systematic review of literature, Delphi clinician survey and stakeholder consensus meeting. Setting: Forty-four UK secondary and tertiary UK centres (service evaluation). Participants: Children with OM/SA. Interventions: PCR diagnostics were compared with culture as standard of care. Semistructured interviews were used in the qualitative study. Results: Data were obtained on 313 cases of OM/SA, of which 218 (61.2%) were defined as simple disease and 95 (26.7%) were defined as complex disease. The epidemiology of paediatric OM/SA in this study was consistent with existing European data. Children who met oral switch criteria less than 7 days from starting i.v. antibiotics were less likely to experience treatment failure (9.6%) than children who met oral switch criteria after 7 days of i.v. therapy (16.1% when switch was between 1 and 2 weeks; 18.2% when switch was > 2 weeks). In 24 out of 32 simple cases (75%) and 8 out of 12 complex cases (67%) in which the targeted PCR was used, a pathogen was detected. The qualitative study demonstrated the importance to parents and children of consideration of short- and long-term outcomes meaningful to families themselves. The consensus meeting agreed on the following outcomes: rehospitalisation or recurrence of symptoms while on oral antibiotics, recurrence of infection, disability at follow-up, symptom free at 1 year, limb shortening or deformity, chronic OM or arthritis, amputation or fasciotomy, death, need for paediatric intensive care, and line infection. Oral switch criteria were identified, including resolution of fever for ≥ 48 hours, tolerating oral food and medicines, and pain improvement. Limitations: Data were collected in a 6-month period, which might not have been representative, and follow-up data for long-term complications are limited. Conclusions: A future RCT would need to recruit from all tertiary and most secondary UK hospitals. Clinicians have implemented early oral switch for selected patients with simple disease without formal clinical trial evidence of safety. However, the current criteria by which decisions to make the oral switch are made are not clearly established or evidence based. Future work: A RCT in simple OM and SA comparing shorter- or longer-course i.v. therapy is feasible in children randomised after oral switch criteria are met after 7 days of i.v. therapy, excluding children meeting oral switch criteria in the first week of i.v. therapy. This study design meets clinician preferences and addresses parental concerns not to randomise prior to oral switch criteria being met. Funding: The National Institute for Health Research Health Technology Assessment programme.