Are Nonconventional Ventilators Associated with Mortality in Covid Ards? A Multicenter Study

ABSTRACT

INTRODUCTION: Many first wave COVID ARDS patients who were intubated spent a prolonged time on non-conventional ventilators (NCV) (e.g. transport devices, BiPAP machines) that are not intended for long-term use. The impact of prolonged NCV use on mortality is unknown. We hypothesized that time spent on NCV, compared to conventional ventilators (CV), is associated with higher mortality among COVID ARDS patients. METHOD(S) This is a retrospective multicenter study of our health system's COVID ARDS database from 03/01/20 - 04/30/20. We included intubated adults with COVID ARDS, mechanically ventilated on either NCV or CV. We excluded patients who switched between ventilator types, palliative extubations, and deaths within 24 hours of intubation. Baseline demographics and confounders were recorded. The primary outcome was 90 day mortality, and secondary outcomes were 3 and 28 day mortality. The effect of time spent on a NCV on 90 day mortality was modeled using logistic regression while controlling for confounders. The effect of time on NCV on 90 day mortality was quantified as an odds ratio (OR) and compared to the null using the Wald test. This was repeated for 3 and 28 day mortality. RESULT(S) Of 2094 patients who met inclusion criteria, we excluded 113 deaths within 24 hours, 500 ventilator switches, and 426 missing data points, for n=1055. Mean age was 65 years, 317 (30%) female, and 384 (36%) Caucasian. NCVs were used in 345 (33%) cases. Median time intubated was 9.8 days for NCV and 11 days for CV groups. Mean adjusted tidal volumes (ATV) were 6.8 for NCV and 6.6 ml/kg for CV. Overall unadjusted 90, 3 and 28 day mortality were 68%, 9% and 62%, respectively. Charlson Comorbidity Index (CCI), BMI, respiratory compliance (RC), ATV, P/F ratio, time averaged PEEP, time on vasopressors, and steroid use were controlled for. CCI and vasopressors correlated with higher mortality (p< 0.05). RC, P/F and PEEP correlated with lower mortality (p< 0.05). Time on NCV did not correlate with 90 day (OR=1.27, p=0.12) or 3 day mortality (OR=1.28, p=0.31) and correlated with increased 28 day mortality (OR = 1.5, p = 0.006). CONCLUSION(S) Among patients with COVID ARDS in early 2020, mechanical ventilation with NCVs was associated with increased adjusted mortality at 28 days but not at 3 and 90 days compared with CVs.