Time to Overhaul the “Rule Out Sepsis” Workup

For neonatologists, decisions about how to manage well-appearing newborns with risk factors for sepsis remain fraught. Maternal chorioamnionitis (inflammation of the chorion and amnion) caused by intrauterine bacterial infection affects 0.1% to 2% of pregnancies^1,2 and increases the newborn’s risk of early-onset sepsis.^1,3 For this reason, clinical guidelines published by the Centers for Disease Control and Prevention in 2010⁴ and the American Academy of Pediatrics in 2012 (with a subsequent clarifying supplement in 2013)^5,6 recommended that all well-appearing term newborns whose mothers were diagnosed with chorioamnionitis undergo laboratory screening for sepsis, including blood culture, and receive at least 48 hours of broad-spectrum antibiotic therapy.

However, consensus has shifted in the years since these recommendations were published. Multiple studies and commentaries have suggested that performing sepsis evaluations on all well-appearing term newborns with a maternal history of chorioamnionitis is unnecessary and may have significant downsides, including: interference with breastfeeding and bonding; alteration of the neonatal microbiome (with uncertain long-term consequences); and risk of medical errors and treatment complications.^7–11 Furthermore, the diagnosis of chorioamnionitis is problematic. A 2012 survey of over 200 obstetricians found that 26% diagnose chorioamnionitis based on maternal fever alone, which is an unreliable indicator of intrauterine inflammation.^12,13

A retrospective study by Jan and colleagues¹⁴ in this issue of Pediatrics examines an alternative management strategy for asymptomatic term and late-preterm (≥35 weeks’ gestation) newborns with maternal history of chorioamnionitis. Patients in their cohort were admitted to a mother-baby unit and underwent initial blood culture and 2 complete blood counts and C-reactive protein measurements. These patients did not receive empirical antibiotic therapy unless they became symptomatic, had a positive culture, or had abnormal laboratory studies, in which case they were transferred to the NICU and treated.

The rationale for this approach is that well-appearing term newborns are extremely unlikely to have sepsis, regardless of their risk factors. Two studies that together included over 26 000 asymptomatic term and late-preterm newborns reported zero cases of culture-positive sepsis.^15,16

It is therefore surprising that Jan’s group¹⁴ reports 10 asymptomatic newborns (4.2% of their cohort) with initial blood cultures positive for pathogenic bacteria. Did these 10 patients benefit from routine blood cultures? The answer is not entirely clear, but our opinion is that they may not have. Because the infants were treated while they were well-appearing, and cultures were not repeated before treatment with antibiotics, it is not possible to know with certainty whether their bacteremia was persistent or transient. However, screening blood cultures without an intention to administer antibiotics has not been shown to improve outcomes.¹⁵ Furthermore, we do not agree with the policy of transferring asymptomatic babies with abnormal complete blood count or C-reactive protein results to the NICU and treating them. The positive predictive value of these laboratory studies is too low to be of use in diagnosing sepsis.^17,18

Jan’s group¹⁴ has taken management of asymptomatic term and near-term newborns with a maternal history of chorioamnionitis in the right direction and has shown that their strategy is safe. But we believe a better approach would be to forego routine laboratory evaluations among this population altogether and manage them using clinical signs alone. Additional studies of this strategy would make a valuable contribution to neonatology and are scientifically and ethically justified.

Two key caveats are important to state. First, in the immediate postpartum period, mild respiratory distress among term or near-term newborns may be attributable to the physiologic transition, which occurs in all newborn infants. It is not necessary to draw laboratories or start antibiotics on these patients as long as their symptoms improve and resolve within the first 6 hours of life.¹⁹ Second, if newborns with a maternal history of chorioamnionitis are to be monitored for signs of sepsis outside the NICU setting, observations must be frequent (at least hourly for the first 6 hours of life and then every 3 hours for the next 18 hours) and performed by adequately trained medical staff. In the absence of frequent, reliable observation, there is a possibility that the early signs of sepsis will be missed and go untreated with potentially severe consequences.

The often-cited Centers for Disease Control and Prevention and American Academy of Pediatrics guidelines for managing newborns with risk factors for sepsis are outdated. Studies like the one performed by Jan and colleagues¹⁴ reflect an interest in proving what most suspect: that well-appearing term and near-term newborns can be safely managed with frequent observation alone. Updates to official guidelines are undoubtedly being developed. In the meantime, clinicians should feel comfortable incorporating evidence-based modifications to published recommendations.

• Copyright © 2017 by the American Academy of Pediatrics