Boston Febrile Infant Algorithm 2.0: Improving Care of the Febrile Infant 1–2 Months of Age

Abstract

Introduction: Significant variation exists in the management of febrile infants, particularly those between 1 and 2 months of age. An established algorithm for well-appearing febrile infants 1–2 months of age guided clinical care for three decades in our emergency department. With mounting evidence for procalcitonin (PCT) to detect invasive bacterial infection (IBI), we revised our algorithm intending to decrease lumbar punctures (LPs) and antibiotic administration without increasing hospitalizations, revisits, or missed IBI. Methods: The algorithm’s risk stratification was revised based on the expert review of evidence regarding test performance of PCT for IBI in febrile infants. With the revision, routine LP and empiric antibiotics were not recommended for low-risk infants. We used quality improvement strategies to disseminate the revised algorithm and reinforce uptake. The primary outcomes were the proportion of infants undergoing lumbar punctures or receiving antibiotics. Admission rates, 72-hour revisits requiring admission, and missed IBI were monitored as balancing measures. Results: We studied 616 infants including 326 (52.9%), after the implementation of the revised algorithm. LP was performed in 66.2% prerevision and 31.9% postrevision (34.3% absolute reduction, P < 0.001). Antibiotic administration decreased by 26.2% (pre 62.4% to post 36.2%, P < 0.001) and hospitalization rates decreased by 8.1% (P = 0.03). There have been no missed IBIs. Adherence to the pathway led to a sustained reduction in LPs and antibiotic administration for 24 months. Conclusion: A revised pathway with the addition of PCT resulted in a safe, sustained reduction in LPs and reduced antibiotic administration in febrile infants 1–2 months of age.

A simple clinical decision rule may help to rule out severe bacterial infection and spare antibiotics in febrile young children

“…Moreover, the published regression models that provide a post‐test probability of infection do not fit with the usual clinician process of treatment decisions 14,15 and seem to have little impact on ATB prescription rate and discharge from PEDs 20,21 . Dorney et al conducted an impact study of a the revised CPR proposed previously by Bachur et al 13 and showed reduction in lumbar punctures and antibiotic prescription but only for very young children from 1 to 2 months of age 19 . Thus, the appropriate care of children with fever without source is still an area of clinical debate.…”

“…Clinical prediction rules (CPRs) can improve diagnostic decision‐making for rare but serious pathologies. However, previously published CPRs 11–18 focused on only clinical parameters 16,17 or laboratory parameters 12,13,18 when other approaches combining clinical and laboratory data are difficult to use in current practice because they require systematic blood sample investigation 11,14,15,19 . Moreover, the published regression models that provide a post‐test probability of infection do not fit with the usual clinician process of treatment decisions 14,15 and seem to have little impact on ATB prescription rate and discharge from PEDs 20,21 .…”

“…However, previously published CPRs [11][12][13][14][15][16][17][18] focused on only clinical parameters 16,17 or laboratory parameters 12,13,18 when other approaches combining clinical and laboratory data are difficult to use in current practice because they require systematic blood sample investigation. 11,14,15,19 Moreover, the published regression models that provide a post-test probability of infection do not fit with the usual clinician process of treatment decisions 14,15 and seem to have little impact on ATB prescription rate and discharge from PEDs. 20,21 Dorney et al conducted an impact study of a the revised CPR proposed previously by Bachur et al 13 and showed reduction in lumbar punctures and antibiotic prescription but only for very young children from 1 to 2 months of age.…”

“…20,21 Dorney et al conducted an impact study of a the revised CPR proposed previously by Bachur et al 13 and showed reduction in lumbar punctures and antibiotic prescription but only for very young children from 1 to 2 months of age. 19 Thus, the appropriate care of children with fever without source is still an area of clinical debate.…”