Increase in use of non-invasive ventilation for infants with severe bronchiolitis is associated with decline in intubation rates over a decade

Purpose

To redress the paucity of studies evaluating non-invasive respiratory support in bronchiolitis patients.

Methods

Following ethics committee approval, the clinical database of a tertiary 23-bed paediatric intensive care unit (PICU) was reviewed for bronchiolitis admissions from January 2000 to December 2009. Length of stay (LOS), ventilatory requirements and risk factors, including prematurity, respiratory syncytial virus (RSV) status, chronic lung, neuromuscular, immune and congenital heart disease, were analysed.

Results

Of 8,288 admissions, 520 (6.27?%) had bronchiolitis with 343 (65.9?%) having RSV. Median (±SD) age and LOS were 2.78?months and 2.68 (±4.32)?days. One (0.2?%) patient died. Assisted ventilation was required for 399 (76.7?%) patients. A total of 114 (28.6?%) patients were intubated directly and 285 (71.4?%) had a trial of non-invasive ventilation (NIV). Significant increase in the use of NIV was seen (2.8?%/year) with decline in intubation rates (1.9?%/ year) (p?=?0.002). Of NIV patients, 237 (83.2?%) needed only NIV and 48 (16.8?%) failed and therefore needed intubation. The median LOS was shorter in those who succeeded NIV (2.38?±?2.43?days) compared to those with invasive ventilation (5.19?±?6.34?days) and those who failed NIV (8.41?±?3.44?days). Presence of a risk factor increased the chances of failing NIV from 6 to 10?%.

Conclusion

NIV was successful in the vast majority of patients, particularly in those without risk factors and halved the LOS in intensive care. Failure of NIV was associated with increased duration of invasive ventilation and PICU LOS. A prospective study comparing different techniques of NIV will be helpful in defining the risks of failure of NIV.