Procalcitonin Use After Clinical Practice Guideline and QI Intervention for Febrile Infants

The 2021 American Academy of Pediatrics Clinical Practice Guideline (CPG) for evaluation of well-appearing febrile infants 8 to 60 days of age strongly recommends the use of inflammatory markers (IMs), including procalcitonin (PCT), to inform decision-making.1,2 Given PCT has better diagnostic accuracy for detecting a bacterial infection than other IMs,3 the recommended practice change can impact clinician decision-making regarding obtaining cerebral spinal fluid, administration of antibiotics, and emergency department discharge. We aimed to describe changes in the trend of PCT use at the patient- and hospital-level with the publication of the CPG paired with a multisite quality improvement (QI) project.We conducted an interrupted time series analysis of infants 8 to 60 days of age presenting with fever to 107 hospitals participating in the REducing Variability in Infant Sepsis Evaluation II (REVISE II) collaborative from November 2020 to October 2022. REVISE II was an international, American Academy of Pediatrics-sponsored QI collaborative.4 Enrollment was concurrent with CPG release with the intent of improving adherence to specific guideline recommendations. Infant inclusion and exclusion criteria mirrored those in the CPG.1We elected to measure the change in PCT as evidence supporting its use in febrile infants is more recent, thus there may be more variability in availability and use.3We estimated changes in the rate of PCT testing using linear regression with mixed-effect modeling for the total sample and by age group (8–21 days, 22–28 days, 29–60 days) before CPG release (November 2020–July 2021) and after REVISE II launch (November 2021–October 2022). We included a 3-month washout period from the release of the CPG and enrollment for REVISE II until the collaborative launch. We compared the hospital-level pre–post proportion of infants with a PCT obtained. Analyses were performed using R statistical software version 4.2.2.This study was deemed nonhuman subjects research by Seattle Children’s institutional review board.We included 15 657 infants from 107 hospitals; 18% (n = 2780) were 8 to 21 days, 13% (n = 2004) were 22 to 28 days, and 69% (n = 10 873) were 29 to 60 days old. The proportion of infants who had a PCT obtained pre-CPG compared with post-CPG had a relative increase of 58.3% (95% confidence interval [CI] 43.7%–71.0%) for infants aged 8 to 21 day, 75.4% (95% CI 57.2%–92.0%) for infants aged 22 to 28 days, and 52.5% (95% CI 45.8%–58.4%) for infants aged 29 to 60 days. The immediate intercept change for obtaining a PCT after REVISE II was 6.5% (95% CI 2.3%–10.7%) with an increasing slope of 2% per month (95% CI 1.7%–2.2%). Using ages 8 to 21 days as a reference where obtaining of IMs was not recommended by the guideline, the rate of obtaining a PCT after REVISE II increased by 9.2% (95% CI 3.3%–15.1%) for infants aged 22 to 28 days and 0.2% (95% CI −4% to 4.5%) for infants aged 29 to 60 days (Fig 1). Although 86% (n = 92) of hospitals demonstrated increased proportions of infants with PCT obtained, there was substantial variation across hospitals (Fig 2).After CPG release and the launch of REVISE II, the proportion of febrile infants with a PCT obtained increased significantly, particularly for infants 22 to 28 days of age. Guideline release followed by participation in a QI collaborative led to an immediate uptake with sustained improvement in this guideline-concordant recommendation. Given the typical lag between evidence dissemination and clinical adoption,5 the rapid and sustained improvement in PCT use in this study may represent a meaningful change in clinician practice. The pairing of a QI collaborative with the release of guidelines may be an effective strategy to improve timely uptake of evidence-based practice. However, our study was limited by hospitals that participated in REVISE II were likely more motivated for change and multisite QI studies require substantial resources; thus, our findings may not be generalizable outside this setting. The variation between hospitals in the pre–post proportion of infants with a PCT obtained may represent variability in PCT availability; however, we hypothesize the overall increase in PCT use at the institutional level is related to the CPG release and QI collaborative participation.Because of data collection limitations, we do not know how clinicians used PCT results to inform decision-making or why a small number of hospitals had a decrease in PCTs obtained after REVISE II implementation. Future studies should investigate how an increase in PCT use impacts clinician decision-making and patient outcomes for well-appearing febrile infants.