Prone during pandemic: development and implementation of a quality-based protocol for proning severe COVID-19 hypoxic lung failure patients in situationally or historically low resource hospitals

BACKGROUND: Intermittent Prone Positioning (IPP) for Acute Respiratory Distress Syndrome (ARDS) decreases mortality. We present a program for IPP using expedient materials for settings of significant limitations in both overwhelmed established ICUs and particularly in low- and middle-income countries (LMICs) treating ARDS due to COVID-19 caused by SARS CoV-2. METHODS: The proning program evolved based on the principles of High Reliability Organizations (HROs) and Crew Resource Management (CRM). Patients with severe ARDS [PaO(2):FiO(2) ratio (PFr) ≤ 150 on FiO(2) ≥ 0.6 and PEEP ≥ 5 cm H(2)O] received IPP. Patients were placed prone 16 h each day. When PFr was ≥ 200 for > 8 h supine IPP ceased. IPP used available materials without requiring additional work from the bedside team. Changes in PFr, PaCO(2), and the SaO(2):FiO(2) ratio (SaFr) positionally were evaluated using t-statistics and ANOVA with Bonferroni correction (p < 0.017). RESULTS: Between 14APR2020 and 09MAY2020, at the peak of deaths in New York, there were 202 IPPs in 29 patients. Patients were 58.5 ± 1.7 years of age (37, 73), 76% male and had a body mass index (BMI) of 27.8 ± 0.8 (21, 38). Pressor agents were used in 76% and 17% received dialysis. The PFr prior to IPP was 107.5 ± 5.6 and 1 h after IPP was 155.7 ± 11.2 (p < 0.001 compared to pre-prone). PFr after the patients were placed supine was 131.5 ± 9.1 (p = 0.02). Pre-prone PaCO(2) was 60.0 ± 2.5 and the 1-h post-prone PaCO(2) was 67.2 ± 3.1 (p = 0.02). Supine PaCO(2) after IPP was 60.4 ± 3.4 (p = 0.90). The SaFr prior to IPP was 121.3 ± 4.2 and the SaFr 1 h after positioning was 131.5 ± 5.1 (p = 0.03). The post-IPP supine SaFr was 139.7 ± 5.9 (p < 0.001). With ANOVA and Bonferroni correction there were statistically significant changes in PFr (p < 0.001) and SaFr (p < 0.001) and no significant changes in PaCO(2) over the four time points measured. Using regression coefficients, the SaFrs predicted by PFrs of 150 and 200 at baseline are 133.2 and 147.3, respectively. CONCLUSIONS: An IPP program for patients with COVID-19 ARDS can be instituted rapidly, safely, and effectively during an overwhelming mass casualty scenario. This approach may be equally applicable in both traditionally austere environments in LMICs and in otherwise capable centers facing situational resource limitations.