BackgroundThe primary objective was to measure the proportion of episodes where care delivery was inconsistent with selected recommendations of a clinical practice guideline (CPG) on fever and neutropenia (FN) management. The influence of site size on CPG‐inconsistent care delivery, and association between patient outcomes and CPG‐inconsistent care were described.MethodsThis retrospective, multicenter study included patients less than 21 years old with cancer who were at high risk of poor FN outcomes and were previously enrolled to a Children's Oncology Group (COG) study at participating National Cancer Institute Community Oncology Research Program (NCORP) institutions from January 2014 through December 2015. Patients were randomly selected for chart review by participating sites from a COG‐generated list. Care delivered in each episode was adjudicated (CPG‐consistent or CPG‐inconsistent) against each of five selected recommendations.ResultsA total of 107 patients from 22 sites, representing 157 FN episodes, were included. The most common CPG‐inconsistent care delivered was omission of pulmonary computerized tomography in patients with persistent FN (60.3%). Of 74 episodes where assessment of four (episodes without persistent FN) or five (episodes with persistent FN) recommendations was possible, CPG‐inconsistent care was delivered with respect to at least one recommendation in 63 (85%) episodes. Site size was not associated with CPG‐inconsistent care delivery. No statistically significant association between CPG‐inconsistent care and fever recurrence was observed.ConclusionsIn this cohort of pediatric patients at high risk of poor FN outcomes, CPG‐inconsistent care was common. Opportunities to optimize resource stewardship by boosting supportive care CPG implementation are highlighted.