Audit To Estimate the Incidence of Air Leak in Non-invasive Ventilation

ABSTRACT

Aim and objective: To determine the incidence of air leak in patients who are on non-invasive ventilation in a COVID ICU at a tertiary hospital. Materials and methods: Non-invasive ventilation (NIV) is a mode of providing ventilatory support without using an invasive airway. It has become increasingly popular in managing respiratory failures in recent times. The success of NIV is multifactorial. Among these, using an appropriate interface and ensuring patient co-operation is important as air leak at the interface is a major and a common cause of NIV failure. Through this audit, we aim to look at the incidence of air-leak among patients on NIV and assess if the incidence of air-leak can be reduced via the use of checklist. The audit included patients on non-invasive ventilation and excluded patients on NIV who had air leak from any source other than the interface also patients with facial anomalies were excluded. We monitored patients on NIV in a COVID ICU twice a day for a period of 1 week and recorded the percentage of air leak as calculated by the ventilator (Hamilton G5). For the purpose of this audit, we categorised air leak into mild (10%-< 30%), moderate (30%-<50%), and severe (≥50%). Following the first cycle of the audit, we tabulated the data and discussed the results with the respiratory therapist. A checklist was formulated which included hourly observations to ensure adequate mask seal and minimize air leak along with ensuring adequate sedation. Subsequently, we conducted two post-intervention cycles of the audit to check the efficiency of the checklist in reducing air leak in patients on NIV. Results: The first cycle of the audit showed that 27.45% of patients had severe air leak and 50.98% of patients had mild to moderate air leak on NIV. In addition, it was noted only 21.57% of patients had correct positioning of the mask with a permissible air leak of <10%. Therefore, we implemented the intervention checklist for the respiratory therapist to ensure minimal air leak and subsequently repeated the audit. Following which the percentage of severe air leak in patients on NIV dropped from 27.45% to 5.09%. The percentage of patients with permissible air leak marginally improved from 21.57% to 27.12%. A third cycle of the audit was done to ensure consistent results following the intervention and it did demonstrate consistent results similar to the second cycle of the audit with a percentage of patients with severe air leak being only 5.88% in comparison to 27.45% in the first cycle. Moreover, the percentage of patients with moderate air leak also dropped to 5.88% in comparison to 24.51% from the first audit and 22.03% from the second audit. Conclusion: A simple checklist based on hourly observations helped to improve ventilation in patients on NIV by decreasing the percentage of severe and moderate air leak and the method proved to be sustainable.