Antibiotic use in neonates with hypoxic-ischemic encephalopathy undergoing therapeutic hypothermia: time to rethink universal empirical treatment.

Neonatal hypoxic-ischemic encephalopathy (HIE) treated with therapeutic hypothermia (TH) can alter the immune system and may present symptoms similar to sepsis. This has led to widespread use of empirical antibiotics, despite limited evidence that support the need for universal empirical treatment. We aimed to evaluate the incidence of culture-proven sepsis and current antibiotic use patterns in neonates undergoing TH. We conducted a multicenter retrospective study including term and near-term neonates with HIE treated with TH in two European NICUs with different antibiotic strategies: one with universal empirical antibiotics (Italy, n = 228) and one with selective use (Belgium, n = 74). We collected data on blood culture results, antibiotic exposure, inflammatory biomarkers (CRP, PCT, leukocytes), and clinical outcomes. Of 302 newborns (276 term and 26 late preterm) who underwent TH, 278 (92%) received antibiotics at birth, with varying regimens for a median of 6 days. Blood cultures were performed in 228 Italian babies and in 72/74 Belgian infants (300 in total, equal to 99.3%), revealing only 2 cases of culture-confirmed early-onset sepsis (0.6%). In Belgium, all cultures from eight infants with clinical suspicion of sepsis were negative, resulting in a Number Needed to Treat (NNT) of 111 to prevent one confirmed infection. Late-onset sepsis occurred in four of 302 patients (1.3%). Biomarkers (CRP, PCT, leukocytes) fluctuated during cooling and rewarming but were not predictive of EOS.

conclusionThe incidence of confirmed early-onset sepsis in neonates with HIE undergoing TH was very low, despite extensive empirical antibiotic use. These findings question the necessity for routine universal empiric antibiotics and advocate for a selective risk-based approach based on clinical assessment. Careful monitoring and prudent antibiotic use are crucial to reduce unnecessary exposure and its potential harms in this vulnerable population. WHAT IS KNOWN: • Distinguishing the symptoms of HIE from those of EOS at birth or shortly thereafter is very complex, as they are often similar and can mimic one another. • Most newborns receiving hypothermic treatment are treated with empirical antibiotics in the suspicion of sepsis, even though the incidence of early-onset sepsis in the European healthcare context is very low. WHAT IS NEW: • Our multicentric study found a remarkably low incidence of confirmed EOS in newborns undergoing TH, irrespective of the antibiotic strategy used in each single center. • We provide data that can support a restriction on the use of empirical antibiotic therapy in neonates undergoing therapeutic hypothermia, also considering the bacterial resistance that the imprudent use of antibiotics is turning into a global emergency.