A strategy for reduction of antibiotic use in new patients admitted to a neonatal intensive care unit

Background: In order to reduce the unnecessary use of and provide appropriate guidance for the administration of antibiotics, the neonatal bacterial infections screening score (NBISS) was developed to assess each new patient that is admitted to the neonatal intensive care unit (NICU). Methods: NBISS was designed based on maternal risk factors, clinical presentations, and laboratory data. The total score of each new patient is calculated at the time of admission. The first period of this study was an observational survey. Receiver operating characteristic (ROC) curves were used to determine the best cut-off NBISS for the diagnosis of bacterial infection (BI) and guide the use of antibiotics during the second period of this study. Results: Of 250 neonates who were admitted to the NICU, 237 (94.8%) received antibiotics during the first period of study. The initial total scores were not statistically different between the BI and non-BI groups (p = 0.155). We weighted C-reaction protein (CRP) (by 8×), the presence of a bulging fontanelle, pus from the ear canal, redness around the umbilicus, reduced movement, and being unable to feed (each by 5×) as significantly different between the BI and non-BI groups (p = 0.015). Weighted scores >8 points demonstrated the best diagnostic accuracy for indicating BI. After introducing NBISS for predicting BI in new patients admitted to NICU, the rate of antibiotic use significantly decreased from 94.8% to 60.3% between the two periods. Conclusion: Using this simple screening strategy, we were able to clinically reduce the use of unnecessary antibiotics.