139 Do neuraminidase inhibitors improve outcomes in critically unwell children with influenza?

BackgroundInfluenza is a common seasonal acute respiratory viral illness. Children, especially those with co-morbidities, are at risk of complications and ICU admission. No specific guidelines have been formulated about starting neuraminidase inhibitors (NAI) in critically ill children with influenza, but the Health Protection Agency and American Academy of Pediatrics have both stated that antiviral therapy should be initiated as soon as possible in this cohort.ObjectivesWe aimed to evaluate the evidence supporting the early initiation of NAI in critically ill children with influenza, by conducting a literature search to establish whether NAI improved survival and shortened intensive care admissions in children critically ill with influenza.MethodsWe searched the literature for articles on the use of NAI treatment in critically unwell children or children in PICU diagnosed with influenza. We excluded articles with adult patients only, or with a mixture of adults and children where the results were not stratified by age. We also excluded articles with children in outpatient settings, or hospitalised on low-dependency units.ResultsOut of 369 articles (Cochrane Library: 2, PUBMED: 328, NHS Evidence: 39), twelve studies published between 2010 and 2017 were included, seven of which were cohort studies (Level 3 evidence) and five case series (Level 4 evidence), with a total of over 7,000 critically ill children with influenza worldwide.Six cohort studies compared mortality in children receiving NAI and children who did not: five of these demonstrated a trend towards decreased mortality with the use of NAI, and one showed no difference between treated and untreated groups. Only one study reached statistical significance, with p = 0.01 for association of NAI treatment with survival. We noted that the two studies which reported on NAI-related adverse events reported none.There is some evidence in our data that early NAI within 48 hours of symptom onset and/or admission has additional benefits compared to late NAI, although the largest cohort study did not demonstrate this effect.The studies included here had limitations. Patient cohorts were heterogeneous, with some having had RT-PCR confirmed influenza and others just a clinical diagnosis. Few studies reported on concurrent treatment with antibiotics or steroids. Several studies stressed that patients receiving NAI were more likely to have co-morbidities and very severe influenza requiring mechanical ventilation a