RESUMO
INTRODUCTION: It is unclear if serum procalcitonin (PCT) estimated at sepsis suspicion can help detect culture-positive sepsis in neonates. We evaluated the diagnostic performance of PCT in culture-positive sepsis in neonates. METHODS: This was a prospective study (February 2016 to September 2020) conducted in four level-3 units in India. We enrolled neonates suspected of sepsis in the first 28 days of life. Neonates with birth weight <750 g, asphyxia, shock, and major malformations were excluded. Blood for PCT assay was drawn along with the blood culture at the time of suspicion of sepsis and before antibiotic initiation. The investigators labeled the neonates as having culture-positive sepsis or "no sepsis" based on the culture reports and clinical course. PCT assay was performed by electrochemiluminescence immunoassay, and the clinicians were masked to the PCT levels while assigning the label of sepsis. Primary outcomes were the sensitivity, specificity, and likelihood ratios to identify culture-positive sepsis. RESULTS: The mean birth weight (SD) and median gestation (IQR) were 2,113 (727) g and 36 (32-38) weeks, respectively. Of the 1,204 neonates with eligible cultures, 155 (12.9%) had culture-positive sepsis. Most (79.4%) were culture-positive within 72 h of birth. The sensitivity, specificity, and positive and negative likelihood ratios at 2 ng/mL PCT threshold were 52.3% (95% confidence interval: 44.1-60.3), 64.5% (60.7-68.1), 1.47 (1.23-1.76), and 0.74 (0.62-0.88), respectively. Adding PCT to assessing neonates with 12.9% pretest probability of sepsis generated posttest probabilities of 18% and 10% for positive and negative test results, respectively. CONCLUSION: Serum PCT did not reliably identify culture-positive sepsis in neonates.