P202 Factors affecting survival-to-discharge in non-COPD patients receiving non-invasive ventilation (NIV) for acute hypercapnic respiratory failure (AHRF)

Introduction NIV is increasingly used outside critical care for patients presenting with AHRF due to non-COPD pathologies. However, there is little evidence to guide clinicians using NIV in these settings. The 2017 NCEPOD report into acute NIV reported inpatient mortality data on just 96 patients with non-COPD diagnoses. Our aim was to investigate the factors associated with inpatient mortality for these patients. We hypothesise that in addition to factors such as pH, which is known in COPD to be associated with mortality, the different diagnostic indications will affect inpatient mortality. Method This retrospective single-centre survey included patients who had acute NIV set up between February 2004 and July 2018 for their first recorded episode of NIV. Exclusion criteria included COPD diagnosis, domiciliary NIV on admission, lack of pre-NIV blood gas and pre-NIV pH ≥7.35. Pre-NIV pH was grouped by thresholds ( 7.25. Diagnosis was classified as shown in table 1 with comparison made with the ‘other’ diagnosis group. Factors (age, diagnosis and pre-NIV pH) were assessed in a multivariable binomial regression model if they were significant in a univariable binomial regression analysis. Results are shown for the multivariable analysis as (OR, 95% CI, p-value). Results 239 patients were included in the analysis. Obesity-related respiratory failure was the indication for 35% of all episodes (table 1). The mortality of obesity-related NIV was significantly lower when corrected for age and pH (0.059, 0.0.015–0.227, p Conclusion Our results suggest that NIV can be safely used outside of critical care for patients with obesity-related AHRF. We have also replicated previous work in COPD patients demonstrating that it may be safe to use in patients with a pH of between 7.15–7.25. However, further work is needed to characterise these thresholds in different diagnostic groups.