1. Using Risk-assessment tools to explore the scope
for safe outpatient antibiotic therapy in febrile infants
Christian Harkensee1,2, Sophie Vaughan2, Amy Moran3, Charlie Massey3, Seilesh Kadambari4, Michael Sharland1,4
1Department of Paediatric Infectious Diseases, St George’s Hospital, London, UK; 2Department of Paediatrics, Division of Paediatric Infectious Diseases, National University Hospital Singapore; 3Department of Paediatrics, St George’s Hospital, London, UK; 4Paediatric Infectious Diseases Research Group,
St George’s University, London, UK . Contact:
Abstract
Background: In the absence of a focus of infection, febrile illness in infants is often treated with empirical antibiotics and hospital admission until culture-negative. Serious bacterial infection is however present in only a minority of these patients. This study explores which patients may benefit from outpatient antibiotic therapy (OPAT).
Methods: 61 infants (≤ 6 months) admitted with suspected sepsis to a tertiary paediatric unit in the UK were prospectively assessed using the UK NICE fever guideline and Rochester screening criteria. Models of suitability for OPAT were devised based on patient age, investigation results and NICE and Rochester scores.
Results: 46 and 51 infants were classed as high risk by NICE and Rochester criteria, respectively. 17 infants had absolute criteria for admission; whereas 44 patients overall were clinically well with or without abnormal investigations (including 7 infants with urinary tract infection (UTI), one with bacterial sepsis). 5 (NICE) or 10 infants (Rochester) scored low risk and may not have required antibiotics. Using ≤3 months as a cut-off for admission (as encouraged by NICE), only 10 patients remained of which 9 had abnormal investigations. With no age limit, 34 patients were OPAT-suitable of which 10 had normal investigations.
Conclusions: NICE and Rochester criteria can support decision-making when using OPAT for infants with suspected sepsis. OPAT would be feasible for more than half of patients using a risk-based approach; early OPAT (<12h) for well patients with normal investigations, and short-term observation with delayed OPAT (12-24h) for patients with abnormal investigations or <3 months old.
potential OPAT patients were neonates, 6 (Rochester) and 7 (NICE) were between 1-3 months old, and 9 (Rochester) and 8 (NICE) patients (total 10) were older than 3 months. All bacterial blood stream and CNS infections (group-B streptococcus sepsis, group-B streptococcus meningitis, E. coli meningitis and UTI) were in neonates, and all but one UTI occuring in this study fell into the age group 0-3 months.
Of the 10 patients who were clinically well and older than 3 months, 9 had abnormal investigations: One had an abnormal urine microscopy with a culture proven UTI, 6 had abnormal CRP (none higher than 30 mg/L), and two had abnormal white cell counts (<5000 or >15,000 per mm3 ). None of these had SBI; five had proven viral infections (RSV, enterovirus), the remaining patients had no pathogen identified on culture or viral studies.
Patients with UTI were followed up with renal ultrasound, MCUG and DMSA scans, showing that two had vesico-ureteric reflux, and one had a defect on DMSA indicating a renal scar from a UTI involving the upper urinary tract. This patient was at presentation clinically well with low fever.
Introduction
Although febrile illness in neonates and young infants are common, only a minority of these (5-30%) are caused by serious bacterial infection (SBI), Because clinical findings and investigations are often non-specific, and the concern over evolving SBI is high especially in neonates and the youngest infants (0-3 months), many patients receive empirical antibiotic treatment and hospital admission, causing inconvenience and cost to the family and potentially exposing the patient to risk of hospital-acquired infection and antimicrobial resistance.
Models to address this problem in practice range from having a low threshold for antibiotics, but aiming for a short duration and OPAT, to withholding antibiotics in the first instance but admitting the patient for observation. This study is applying two scoring systems for SBI (Rochester criteria and NICE guideline for febrile illness in children) to stratify risk and potential eligibility for OPAT.
Study Design & Methodology
Infants between 0-6 months of age admitted to St George’s Hospital in London between July and December after presenting to the children’s A&E department were included in this prospective observational study. The study cohort includes all infants who had a full septic screen (blood/urine/CSF culture) over the six month period, and all infants admitted for a febrile illness during the month of December 2011 (regardless of completeness of septic screen). All children were investigated and treated according to existing hospital guidelines (which do not incorporate Rochester or NICE guideline criteria). After initial assessment ,laboratory tests, and treatment decisions all patients were scored by an independent clinicians using the Rochester and NICE criteria. Data on microbiological and clinical outcome were prospectively collected, and follow-up after discharge was monitored for a minimum of 3 months.
Patients were categorized as potential candidates for OPAT if they were clinically well (normal vital parameters AND judgement by the most senior clinician present (specialist registrar or consultant) with no focal signs of infection other than signs of viral upper respiratory tract infection, or lower tract UTI. Patients were categorised as unwell if they had vital signs outside the normal range, and/or focal signs of SBI, and/or were judged to be clinically unwell by the most senior clinician present. For the purpose of this study this group would receive an obligate full septic workup and iv antibiotics. Patients categorized as low risk (Rochester) or green (NICE) would have had no indication for antibiotic treatment.
Figure 1: Risk classification and identification of patients with a potential for OPAT from this study cohort
Results
A total of 61 patients were recruited; 9 of these had SBI (UTI=7, sepsis by group B streptococcus=1, meningitis by group B streptococcus=1, meningitis and UTI by E. coli=1). Using Rochester criteria, 51 patients were classified as high risk (of which 18 were clinically unwell), which included all SBI. By NICE criteria, 46 patients scored as high risk (‘red’), of which 25 were clinically unwell, which included all SBI except for 7 cases of UTI. 21 high-risk patients were clinically well and had the level of fever as the only as criterion for high risk , this group included all 7 urinary tract infections. By Rochester criteria, 10 patients were classified as low risk, by NICE criteria five patients fell into the low risk group (‘green’) whereas 10 patients were clinically well but had mild clinical symptoms, classified as intermediate risk (‘amber’).
From a risk-based perspective alone, 10 (Rochester) and 5 patients (NICE) may not have required antibiotics or admission, whereas clinically unwell patients (Rochester: 18, NICE: 25) would have qualified for antibiotics and inpatient management. 33 (Rochester) and 31 patients (NICE) could potentially benefit from OPAT management.
Dissecting this group further, 18 (Rochester) and 16 (NICE) of
Discussion & Conclusions
A risk-based approach demonstrates that potentially half of all febrile neonates and infants could be eligible to OPAT as they are clinically well or score low for SBI. Further factors, however, have to be taken into account: In neonates (up to 4 weeks of age) and very young infants (up to 3 months of age), concerns over sudden detoriation of SBI is high; and parents of such children may not feel comfortable with outpatient management. Any OPAT in this setting would have to be comprehensive and hospital back-up always available. We are proposing an approach of early discharge and OPAT for infants who are clinically well and have normal/mildly abnormal laboratory results or UTI. Neonates and infants <3 months old who are clinically well could qualify for short-term admission (12-24 hours) and discharged on OPAT if remaining stable or improving. Admission for inpatient antibiotic treatment could be reserved for unwell patients or patients who worsen during short-term admission or on OPAT.