ABSTRACT
BACKGROUND: The prognosis for late-onset sepsis depends largely on a timely diagnosis. We assess central-peripheral temperature difference monitoring as a marker for late-onset neonatal sepsis diagnosis. METHODS: We performed a prospective, observational study focusing on a cohort of 129 very low-birth-weight infants. Thermal gradient alteration was defined as a difference of > 2°C maintained during 4 hours. We then determined its association with the late-onset sepsis variable through logistic regression. RESULTS: We enrolled 129 preterm babies in 52 months. Thermal gradient alterations showed an adjusted odds ratio for late-onset sepsis of 23.60 (95% confidence interval [CI], 6.80-81.88), with a sensitivity of 83% and negative predictive value of 94%. In 71% of cases, thermal gradient alteration was the first clinical sign of sepsis, while C-reactive protein was < 1.5 mg/dL in 64% of cases and procalcitonin < 2 ng/mL in 36%. These figures indicate potential for early diagnosis. CONCLUSIONS: Sustained increases of central-peripheral temperature differences are an early sign of evolving late-onset sepsis.
