How long is too long: A retrospective study evaluating the impact of the duration of noninvasive oxygenation support strategies (high flow nasal cannula & BiPAP) on mortality in invasive mechanically ventilated patients with COVID-19.

Abstract

We investigated the association of noninvasive oxygenation support [high flow nasal cannula (HFNC) and BiPAP], timing of invasive mechanical ventilation (IMV), and inpatient mortality among patients hospitalized with COVID-19. Retrospective chart review study of patients hospitalized with COVID-19 (ICD-10 code U07.1) and received IMV from March 2020-October 2021. Charlson comorbidity index (CCI) was calculated; Obesity defined as body mass index (BMI) ≥ 30 kg/m2; morbid obesity was BMI ≥ 40 kg/m2. Clinical parameters/vital signs recorded at time of admission. 709 COVID-19 patients underwent IMV, predominantly admitted from March-May 2020 (45%), average age 62±15 years, 67% male, 37% Hispanic, and 9% from group living settings. 44% had obesity, 11% had morbid obesity, 55% had type II diabetes, 75% had hypertension, and average CCI was 3.65 (SD = 3.11). Crude mortality rate was 56%. Close linear association of age with inpatient-mortality risk was found [OR (95% CI) = 1.35 (1.27-1.44) per 5 years, p<0.0001)]. Patients who died after IMV received noninvasive oxygenation support significantly longer: 5.3 (8.0) vs. 2.7 (SD 4.6) days; longer use was also independently associated with a higher risk of inpatient-mortality: OR = 3.1 (1.8-5.4) for 3-7 days, 7.2 (3.8-13.7) for ≥8 days (reference: 1-2 days) (p<0.0001). The association magnitude varied between age groups: 3-7 days duration (ref: 1-2 days), OR = 4.8 (1.9-12.1) in ≥65 years old vs. 2.1 (1.0-4.6) in <65 years old. Higher mortality risk was associated with higher CCI in patients ≥65 (P = 0.0082); among younger patients, obesity (OR = 1.8 (1.0-3.2) or morbid obesity (OR = 2.8;1.4-5.9) (p<0.05) were associated. No mortality association was found for sex or race. Time spent on noninvasive oxygenation support [as defined by high flow nasal cannula (HFNC) and BiPAP] prior to IMV increased mortality risk. Research for the generalizability of our findings to other respiratory failure patient populations is needed.