1641 Evaluation of the kaiser permanente risk calculator to manage newborn babies at risk of early onset neonatal sepsis: a prospective cohort study

Résumé fourni par la source

Background Early onset neonatal sepsis (EONS) is rare but associated with significant mortality and morbidity. Most postnatal units assess risk and treat for suspected EONS using the NICE guideline (CG149) based on pregnancy and peri-partum risk factors. This sees many clinically well babies receive empirical antibiotics, before investigations rule out any evidence of infection. This process usually involves separating baby and mother and exposure to the risks associated with intravenous antibiotic therapy. As such, there is a clear mandate to validate a tool that can safely risk stratify and identify which babies should receive empirical antibiotics. The Kaiser Permanente Risk Calculator (KPRC) combines peri-partum risk factors with the clinical status of the neonate, resulting in a stratification to low, medium or high risk of EONS. Application of the KPRC in some centres has been shown to reduce antibiotic administration by 50–75% without missing any cases of culture positive sepsis. Objectives To evaluate the performance of the KPRC in comparison to the NICE guideline in identifying babies at risk of developing EONS in a large UK maternity hospital. Methods In this prospective single-centre cohort study, we identified all neonates born at 34 weeks’ gestation who were commenced on empirical antibiotics, as per the CG149 guideline. We prospectively collected data on these babies including peri-partum risk factors and the clinical status of the baby at the time the decision was made to treat. This data was inputted into the KPRC, a risk score was calculated and the suggested management plan recorded. Our primary outcome measure was to observe for any babies who became significantly unwell with infection (defined as unwell as per KPRC criteria, maximal CRP > 50mg/L, raised CSF white cell count or positive blood or CSF culture) that would not have been started on empirical antibiotics had the KPRC been used. Results Over a 4-month period we identified 94 babies born at 34 weeks’ gestation receiving empirical antibiotics for suspected EONS as per CG149. One baby had a positive blood culture and one had a positive CSF culture. Of these 94 babies, only 19 (20.2%) would have been commenced on empirical antibiotics by the KPRC. Those 19 babies were more likely to have a CRP >20mg/L (p<0.05) and need a longer course of antibiotics (p<0.05). Interestingly, 27 out of the 94 babies (28.7%) were classed as significantly unwell (as defined above) and of these, only 15 (55.6%) would have been identified by the KPRC as needing antibiotics. 6 babies with a CRP >50 were not identified for antibiotics by the KPRC, including a baby with a CRP of 155. Conclusions Application of the KPRC to this cohort would have significantly reduced antibiotic administration rates. The safety and appropriateness of the KPRC for our population needs further assessment though, as some babies with significant evidence of infection went undetected by the calculator. In particular, the role of the KPRC in identifying clinically well babies with a very high CRP needs further consideration.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.