What clinical signs, symptoms, and investigations are associated with poor prognosis in children with respiratory tract infections?
Demand for NHS services is at an all-time high and continues to rise. A key contributor is consultations for children with respiratory tract infections (RTI). Understanding which features are associated with poor prognosis in children with RTI is the first step in developing interventions for parents to reduce potentially unnecessary healthcare attendances in this patient group.
Systematic literature review
We found 27 studies met our inclusion criteria with a range of settings, populations, conditions, study designs, and outcomes. There were no studies in a home setting. In children presenting with signs of an RTI, we found that respiratory signs, vomiting, fever, signs of dehydration, and low oxygen saturations are associated with an increased odds of hospitalisation. Findings on ear examination was associated with a reduced odds of hospitalisation in this population. In children presenting with bronchiolitis, low oxygen saturation, reduced food and fluids intake, a high respiratory rate, an accessory muscle score ≥6/9, and tachycardia were associated with increased odds of hospitalisation. In patients presenting with laboratory confirmed influenza abnormal lung auscultation or x-ray, fatigue, tachycardia, respiratory distress, and shortness of breath were associated with increased odds of hospitalisation. Where associations were found, the absolute risk of hospitalisation remained small. Headache, congestion, and chills were association with a reduced odds of hospitalisation. There was little evidence of signs, symptoms, or investigations associated with death, re-consultation, antibiotic prescription, and prolonged symptoms. The available evidence was of a low to medium quality with a third of studies at risk of bias from confounding.
There is limited evidence for factors associated with poor prognosis in children with RTI, and no studies have explored factors which can be assessed by parents in the home. Evidence quality is limited by retrospective design of studies, lack of clarity with regard to prognostic factor measurement, and full reporting of data. Where evidence was available, the absolute risk of poor prognosis was small. Prospective studies asking whether symptoms, such as breathing difficulty or ‘a temperature’, can predict prognostic outcomes in the home setting would be the most relevant for parents.