Background In August 2012, new national guidance (National Institute of Health and Care Excellence (NICE) CG149) for management of early onset sepsis (EOS) was introduced in the UK. The guidance outlined a consistent approach for septic screens in newborn infants based on risk factors, and suggested biochemical and clinical parameters to guide management. In particular, it advised a second C-reactive protein level (CRP) 18–24 h into treatment to help determine length of antibiotic course, need for lumbar puncture (LP), and suggested review of blood culture at 36 h.
Objective We evaluated impact of this guidance in our neonatal unit.
Methods We compared two time periods, before and following the guidance. We evaluated length of stay, second CRP 18–24 h into treatment, percentage of babies having LP and duration of antibiotics.
Results Before NICE guidance, 38.1% of screened babies stayed <72 h. This reduced to 18.4% following guidance. Before guidance, 20.9% babies stayed >5 days, which increased to 27.7% following NICE recommendations. Repeat CRP measurements increased from 45% to 97%. In 58% of these babies, repeat CRPs influenced management and hospital stay. An increase in LPs performed from 14% to 23% was noted. There were no positive blood cultures or LP results.
Conclusions We envisaged shorter hospital stays with new NICE standards, particularly, with the aim of 36 h blood culture reporting. However, repeat CRP led to further investigations, increased LPs and longer durations of treatment and stay. This, in turn, impacted on workload and cost, and influenced parental experience in the first few days of life.
- Infectious Diseases
Statistics from Altmetric.com
What this study adds
Although National Institute of Health and Care Excellence (NICE) guidance improved consistency in management of EOS babies, repeat CRP led to increased investigations (eg, lumbar punctures).
More babies had longer courses of antibiotics and length of stay following new NICE recommendations.
What is already known on this topic
Early onset sepsis (EOS) is an important clinical entity.
Wide variation in management of well babies at risk of EOS between units.
Serial CRP is a sensitive marker of EOS.
In the UK, The National Institute of Health and Care Excellence (NICE) is an independent organisation, responsible for providing the national guidance to promote uniformity in quality and practice and is based on best available evidence. NICE CG149 was published in August 2012 to provide guidance on the management of infants with risk factors for early onset sepsis (EOS). The guidance outlined a strategy for septic screens based on risk factors, and used biochemical and clinical parameters to guide management. Key areas highlighted in the guidance were: (1) a second C-reactive protein level (CRP) 18–24 h into treatment to help determine length of antibiotic course and whether to perform lumbar puncture (LP) (consideration of lumbar puncture particularly if CRP >10) and (2) a 36 h blood culture reporting system.
We evaluated the impact of this guidance on our postnatal ward in a tertiary neonatal setting. We looked at the impact of repeat CRP during treatment, LP rates and length of postnatal stay in infants screened for sepsis.
We retrospectively evaluated infants who had septic screens for EOS risk factors before and after introduction of NICE guidance. We examined two time periods, one before guidance (Period 1) and one following NICE recommendations (Period 2). We identified 76 infants in Period 1 (Jul/Aug 2012); 66 infants were identified in Period 2 (Apr/May 2013). Hospital records and the neonatal database were used to identify babies. We evaluated the length of stay (LOS), repeat CRP 18–24 h into treatment, LP (% infants), and antibiotic duration of these babies.
Both the cohorts had similar demographic profiles (table 1). All babies were treated in postnatal ward with antibiotics, though nine (11.8%) babies in Period 1, and 12 (18.1%) babies in Period 2 had brief (<8 h) admission in neonatal high dependency unit. Prolonged rupture of membranes (>24 h) was the most common risk factors in both the groups.
The proportion of infants screened in both groups represented 8%–9% of total births with similar infant demographics. There were no positive blood cultures.
The new NICE guidance increased the number of second CRPs performed from 45% in Period 1% to 97% in Period 2. In Period 2, repeat CRP values were higher in 58% of babies, which influenced management (ie, LP, increased antibiotics and LOS). An increase in the number of LPs performed from 14% in Period 1 to 23% in Period 2 was also noted. In Period 1, before introduction of NICE guidance, 38.1% of screened babies stayed 72 h or less. This reduced to 18.4% in Period 2. The duration of hospital treatment and LOS of more than 5 days were noted in 20.9% in Period 1 and this increased to 27.7% in Period 2, post-NICE guidance. The LOS preguidance and postguidance is shown in figure 1.
Our data indicate that the new NICE guidance (CG149), although providing welcome consistency, may lead to increased investigations and increased length of hospital stay. Emphasis has been given to a second CRP 18–24 h into treatment to guide management. LP was proposed as an investigation to consider in well asymptomatic babies if CRP >10. Early blood culture reporting systems at 36 h instead of ‘conventional’ 48 h were highlighted.
In our unit, the second CRP was increased (and >10) in 58% of screened infants and led to increased antibiotic use, increased LP and greater LOS.
NICE Guideline Development Group (GDG) considered a list of potential biochemical markers to effectively diagnose and manage babies with risk for EOS. These markers were then qualitatively evaluated based on best available evidence, to assess sensitivity as a screening tool, costs incurred, and ease of implementation in a hospital setting. Serial CRP, particularly in the first 48 h, was proposed, as the marker of choice, in the absence of clinical symptoms. It has highest diagnostic accuracy and is most cost effective (health-economic analysis, NICE GDG, August 2012).
NICE gave special deliberation to LP, balancing the clinical imperative for prompt identification of bacterial meningitis to facilitate effective treatment. In period 2, we observed an increase in the number of LPs from 14% to 23%. The general consensus in our unit was a cut-off value of CRP >10 was too aggressive, and a threshold of CRP >20 for LP in otherwise asymptomatic babies has been adopted in our unit. It is interesting to note that consideration of LP in well-asymptomatic babies (with ‘cut-off’ values CRP >10), would have meant more invasive LPs in our cohort. Of note, all our LP culture results were sterile.
Thirty-six-hour blood culture reporting system was proposed as a means to assess discharge of these babies early to improve general parental experience and ensuring cost-effectiveness. Logistically, this was difficult to implement in our hospital as it meant reporting blood culture at 36 h for our neonatal unit alone. There are very limited data at present to support its use across all specialties and age groups. Reporting of blood culture at 36 h for a single specialty (neonatal) led to a delay in implementation in our unit.
NICE Guidance CG149 provides more consistency in management of babies at risk of EOS, and is due to be reviewed in 3 years time in 2015. NICE guidance did not increase the number of babies screened in our unit. However, repeat CRP led to further investigations, increased LPs and longer durations of treatment and stay. This, in turn, impacted on workload, costs and may have influenced parental experience in the first few days of life. This evaluation identifies the need for services to evaluate the impact of national guidelines on their service and the need for effective feedback mechanisms for guideline development.
The authors thank Dr Jenneh Tarawally (JT), Junior Trainee, Neonatal Unit, St Georges University Hospital.
Collaborators Nigel Kennea.
Contributors The proposal for an audit to assess efficacy of the NICE standards was done by DAD, LA and AM. Audit data was collected by AM, LD, LA and JT. Design, implementation, analysis and manuscript evaluation was done by AM, DAD and NK. AM compiled the evaluation and submitted the study.
Competing interests None.
Provenance and peer review Not commissioned; internally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.