Adam King, Laura Croucher, Charlotte Weeks, Mike Hall, Sarah Davidson
Institution(s)
Princess Anne Hospital, University Hospital Southampton NHS Foundation Trust
Introduction (include hypothesis)
Early Onset Sepsis (EOS) describes infection in the newborn developing before 72 hours. NICE Guideline 149 advises criteria for starting treatment for EOS, but subsequently some infants have no evidence for EOS on investigation. An EOS probability calculator has been shown to reduce the proportion of babies in the USA exposed to antibiotics for EOS. We hypothesise that applying the calculator to each baby at 35 weeks’ gestation, and over in a UK maternity unit, the proportion receiving antibiotics for EOS could be reduced.
Methods (include source of funding and ethical approval if required)
Electronic clinical and laboratory records were used to identify inborn babies who received antibiotics for EOS in a 3-month period from June-August 2018 using the (then current) guidelines based on NICE CG149. All babies born at 35+0/40 and above were included. From September 2018, the Kaiser Permanente EOS calculator replaced the local CG149-based decision-making tool regarding starting antibiotic treatment. A comparison of frequency of antibiotic treatment and positive blood cultures in the first five months (September 2018-January 2019) following the introduction of the calculator was made with the three months preceding introduction.
Results
In the 3 months June – August 2018, 151/1323 (11.4%) received parenteral antibiotics for treatment of suspected EOS. During the period September 2018 – January 2019, following implementation of the EOS calculator, we saw a reduction to 169/2208 (7.6%) of infants received parenteral antibiotics (p=0.0002; Chi squared test). In the pre implementation period 2 infants had Group B Streptococcus (GBS) positive blood cultures. Neither met criteria for starting antibiotics using either CG149 or the EOS calculator, but both met criteria for observation and monitoring. Following implementation 3 infants had positive blood cultures for GBS. None of these were identified as at risk by either screening system. All 3 presented clinically unwell within the first 72 hours, and received empirical antibiotics based on symptoms and abnormal observations.
Conclusions
Implementing an EOS calculator to a UK hospital is feasible, and results in a significant reduction in the proportion of babies born at 35/40 and above receiving antibiotics for EOS. Although neither system detects all cases of EOS there have to date been no cases of sepsis missed by the calculator compared to CG149.
References (include acknowledgement here if appropriate)
Kaiser Permanente Research [2019], Neonatal Early-Onset Sepsis Calculator [Online]. Available from https://neonatalsepsiscalculator.kaiserpermanente.org/ [Accessed April 2019]