Baseline Variation in Lung Point-of-Care Ultrasound Cohorts With COVID: Implications for Prognostication

ABSTRACT

Study Objectives: Lung point-of-care ultrasound (L-POCUS) is a novel, radiation-free diagnostic tool that could aid in COVID-19 prognosis in non-critically ill patients. Prognostication requires capturing presenting symptoms and outcomes that may change over time. Variations of environment, presenting symptomatology and follow up can introduce uncontrolled heterogeneity impacting outcome. The purpose of our study was to examine demographic, clinical, and 40-day follow up patterns between two national sites enrolling ambulatory COVID patients for the purpose of determining the association between hypoxia at day 40 and initial L-POCUS findings. Method(s) This was a cross sectional study design of patients at two tertiary care institutions in the Northeast (NE) and Midwest (MW) from January 1st, 2021-April 30th, 2022. We included subjects with respiratory complaints who tested positive for COVID-19 and maintained oxygen saturation >=92% for two hours after presentation to the emergency department as part of a larger project focused on describing L-POCUS prognostic characteristics in non-critically ill COVID patients. Initial vital signs and diagnostic data were collected. Blinded L-POCUS operators recorded seven lung windows (two anterior, two lateral and three posterior per lung field). We utilized a rubric that ranged from zero to six with zero being normal lung and six indicating severe lung pathology from COVID to score each image. Pleural findings included indentation, thickening (each one point), or discontinuity (two points). Parenchymal abnormalities included B lines (1-3 B lines =1 point, >3 B lines =2 points, coalescing or "waterfall" B lines=3 points). Subpleural consolidations scored an automatic six points out of a maximum of 42 per lung. Subjects received pulse oximetry use training and were followed by structured chart review or telephone interview 40-days following presentation. Telephone follow up included highest and lowest pulse oximetry at rest and on 60 second ambulatory test and a structured chart review at any health care visit documented evidence of hypoxia. Hypoxia was defined at <=92% 40 days from index visit. We present descriptive data and corresponding parametric or non-parametric statistic. Result(s) We enrolled 154 subjects (MW 122 (80%), NE 32 (21%). The NE population was more likely to be Hispanic (55% vs 18%, p=<.05) while the MW site was more likely to be African American (76% vs 42%, p<.05). There were no sex differences (NE, 63% female, MW 56% female). There were no significant differences between age (NE 40 years (IQR 31-54), MW 42 years (IQR, 30-56), or Body Mass Index (NE 29 (IQR 25-33), MW 29 (IQR, 24-35). CXR was ordered for 128 (83%) subjects and CT for 18 (12%) but there was no difference between sites (NE CXR 27(93%), CT 5 (17%), MW CXR 101 (83%), CT 13 (11%)). Median L-POCUS scores were 6 (IQR 5-12) and differed by site (NE 14, (IQR 13-27);MW 2 (IQR 2-10, p<.0001). Forty day telephone follow-up was 40% (59/154) and did not differ by site. We identifed 40 (26%) cases of subsequent hypoxia within 40 days of index visit. Outcome did not differ by site (NE 5/32 (16%) MW35/122 (29%), P=0.18). Conclusion(s) There were no meaningful clinical differences between cohorts at distinct geographical locations although NE subjects score higher on initial L-POCUS. Telephone follow up rates were low at both sites. Prognostication may need to account for L-POCUS scoring variability. No, authors do not have interests to disclose Copyright © 2022