ROX Index Predicts Intubation in Patients with COVID-19 Pneumonia and Moderate to Severe Hypoxemic Respiratory Failure Receiving High Flow Nasal Therapy

Introduction Use of high flow nasal therapy (HFNT) to treat COVID-19 pneumonia has been greatly debated around the world due to concern for increased healthcare worker transmission and delays in invasive mechanical Ventilation (IMV). Herein we analyze the utility of the ROX index to predict the need and timing for IMV in a retrospective analysis of patients with COVID-19 with moderate to severe hypoxemic respiratory failure treated with HFNT. Methods This was a retrospective analysis of 129 consecutive patients with COVID-19 admitted to Temple University Hospital in Philadelphia, Pennsylvania, from March 10, 2020, to May 17, 2020 with moderate to severe hypoxemic respiratory failure treated with High Flow nasal therapy (HFNT). HFNT patients were divided into two groups: HFNT only and HFNT progressed to IMV. The primary outcome was the ability of the ROX index to predict the need of IMV. Secondary outcomes were mortality, rates of intubation, length of stay (LOS) and rates of nosocomial infections in our cohort treated with HFNT were also reported. Results 837 patients with COVID-19 were screened, 129 met inclusion criteria. The mean age was 60.8(+13.6) years, BMI 32.6(+8), 58(45 %) were female, 72(55.8%) were African American, 40 (31%) Hispanic. 48 (37.2%) were smokers. Of the 129, 89 were HFNT only group whereas 40 in the HFNT progressed to IMV group. Mean time to intubation was 2.5 days(+ 3.3). The 89 HFNT only patients had a significant improvement in ROX from initiation of HFNT at all recorded time points. In contrast, the ROX in HFNT progressed to IMV patients remained unchanged or decreased over time. ROX index of less than 5 at HFNT initiation was predictive of progression to IMV (OR = 2.137, p = 0,052). Any decrease in ROX index after HFNT initiation was predictive of intubation (OR= 14.67, p 0), peak D-dimer >4000 and admission GFR < 60 ml/min were very strongly predictive of need for IMV (ROC = 0.86, p=0.001). Mortality was 11.2% in HFNT only group versus 47.5% in the HFNT progressed to IMV group (p = 0.0001). Mortality and need for pulmonary vasodilators were higher in the HNFT progressed to IMV group. Conclusion ROX index is a valuable, noninvasive tool to evaluate patients with moderate to severe hypoxemic respiratory failure in COVID-19 treated with HFNT. ROX helps predicts need for IMV and thus limiting morbidity and mortality associated with IMV.