Aims To assess if clinical factors predict admission for bronchiolitis.

Methods We conducted a prospective cohort study, recruiting babies less than 12 months old who presented to the Emergency Department (ED) with bronchiolitis (diagnosed clinically by ED doctors). Infants' clinical signs and symptom were recorded on presentation. These were analysed to determine those factors which predicted admission, both singly and by using multiple logistic regression.

Results 86 infants were recruited to the study. The mean age (±SE) was 22±1.5 weeks and 40% of infants were female. Five risk factors (shown in Tables - age, gender, heart rate, respiratory rate and oxygen saturation) predict admission in 88.4% of our cohort.

In the table (table 1), odds ratios for continuous variables reflect impact per unit change. Factors which did not predict admission included wheeze, crackles, absence of wheeze and crackles, increased work of breathing, poor feeding and apnoea.

Conclusion This supports but develops further the previous observations of Walsh et al1 that clinical factors can be used to predict admission in bronchiolitis; however our study identifies 5 key factors rather than the 12 that were previously identified. Factors were identified which did not predict admission although some of these may reflect the sample size. These factors can be used to develop algorithms to guide general practitioners to determine which babies need referral to secondary care, ED doctors to determine who requires hospital admission and help reduce unnecessary attendances at the ED. Further work needs to be done to develop appropriate tools and to assess their utility prospectively.