Identifying patient level factors of prone positioning therapy

Shewit Giovanni | 2020

Advisor: Daniel A. Enquobahrie

Research Area(s): Clinical Epidemiology, Epidemiologic Methods


BACKGROUND: Acute Respiratory Distress Syndrome (ARDS) is an important cause of respiratory failure associated with high mortality and morbidity. Prone positioning (proning) is a proven therapy for ARDS and is strongly recommended by the American Thoracic Society guidelines. Despite this, it is vastly underutilized in eligible patients. We sought to determine the patient-level factors associated with proning in moderate-to-severe ARDS. METHODS: We performed a secondary analysis of data collected as part of ROSE, a randomized controlled trial of patients with moderate-to-severe ARDS requiring intubation across 48 medical centers in the United States (U.S.). Patients who met criteria for moderate ARDS (PaO2:FIO2 ≤ 150 mmgHg) for less than 48 hours were included. We utilized data at baseline prior to enrollment, study day 1 – defined as the first day that patients met enrollment criteria – and study day 2. We excluded 31 patients who were already proned prior to the study period and one patient who was enrolled twice at two different sites. The outcome was proning during the initial 48 hours after moderate-to-severe ARDS diagnosis. The exposures of interest included persistent hypoxemia after 24 hours, severe ARDS (PaO2/FiO2) ≤ 100 mmHg) , presence of pneumonia, baseline mean arterial pressure (MAP) ≤ 65 mm hg, need for renal replacement therapy within initial 48 hours, baseline plateau pressure (Pplat) and baseline cardiovascular Sequential Organ Failure Assessment (SOFA) score. Multivariable logistical regression was used to estimate odds ratios and related 95% confidence intervals. RESULTS: Over 50% of the cohort were male and the median age was 56. Of the 974 patients, 13% of patients were proned in the first 48 hours after ARDS diagnosis. Proned patients were more likely to have pneumonia (86% vs. 76%), require renal replacement therapy (28% vs. 23%) and vasoactive agents on day 1 (72% vs. 66%) and were twice as likely to have persistent hypoxemia after 24 hours (28% vs. 14%). In multivariable analysis, pneumonia (OR 1.90, 95% CI 1.11-3.23), persistent hypoxemia after 24 hours (OR 2.32, 95% CI 1.45-3.71) and severe ARDS on initial diagnosis (OR 1.77, 95% CI 1.17-2.6) were significantly associated with proning. Markers of hemodynamic instability – hypotension and cardiovascular SOFA score – were not associated with proning. CONCLUSIONS: Proning is more likely to be utilized in patients with pneumonia, persistent hypoxemia, and severe ARDS. Better understanding of the site-level and provider-level factors can inform rational decision-making in clinical care and improve the current low rates of proning.