Assessment of the SpO2/FiO2 ratio as a tool for hypoxemia screening in the emergency department.

OBJECTIVE: We assessed the performance of the ratio of peripheral arterial oxygen saturation to the inspired fraction of oxygen (SpO2/FiO2) to predict the ratio of partial pressure arterial oxygen to the fraction of inspired oxygen (PaO2/FiO2) among patients admitted to our emergency department (ED) during the SARS-CoV-2 outbreak. METHODS: We retrospectively studied patients admitted to an academic-level ED in France who were undergoing a joint measurement of SpO2 and arterial blood gas. We compared SpO2 with SaO2 and evaluated performance of the SpO2/FiO2 ratio for the prediction of 300 and 400 mmHg PaO2/FiO2 cut-off values in COVID-19 positive and negative subgroups using receiver-operating characteristic (ROC) curves. RESULTS: During the study period from February to April 2020, a total of 430 arterial samples were analyzed and collected from 395 patients. The area under the ROC curves of the SpO2/FiO2 ratio was 0.918 (CI 95% 0.885-0.950) and 0.901 (CI 95% 0.872-0.930) for PaO2/FiO2 thresholds of 300 and 400 mmHg, respectively. The positive predictive value (PPV) of an SpO2/FiO2 threshold of 350 for PaO2/FiO2 inferior to 300 mmHg was 0.88 (CI95% 0.84-0.91), whereas the negative predictive value (NPV) of the SpO2/FiO2 threshold of 470 for PaO2/FiO2 inferior to 400 mmHg was 0.89 (CI95% 0.75-0.96). No significant differences were found between the subgroups. CONCLUSIONS: The SpO2/FiO2 ratio may be a reliable tool for hypoxemia screening among patients admitted to the ED, particularly during the SARS-CoV-2 outbreak.

Authors' response: Novel criteria for dyspnea patients

Dear Editor, We would like to thank the authors of the recent letter “Novel Criteria for Dyspnea Patients” for sharing their concerns after reading our manuscript. In this study, we discussed the possibility of improving early patient assessment and identifying the presence of significant hypoxemia by using the SpO2/FiO2 ratio for all patients admitted for respiratory symptoms. Patients characteristics N = 395 Age, median (IQR) 60 (44–78) Gender, male, N (%) 193 (48.9%) Comorbidities Cardiac, N (%) 91 (23.0%) Heart failure, N (%) 71 (18.0%) Atrial fibrillation, N (%) 40 (10.1%) Coronary artery disease, N (%) 30 (7.6%) Other cardiac disease, N (%) 33 (8.4%) Pulmonary, N (%) 118 (29.9%) Asthma, N (%) 48 (12.2%) Chronic obstructive pulmonary disease(COPD), N (%) 51 (12.9%) Lung carcinosis, N (%) 10 (2.53%) Restrictive lung disease, N (%) 16 (4.05%) Other, N (%) 10 (2.53%) Acute kidney failure, N (%) 33 (8.4%) Chronic kidney disease, N (%) 29 (7.34%) COVID-19 status Positive, N (%) 90 (22.8%) Negative, N (%) 305 (77.2%) Diagnosis at discharge Pulmonary disease, N (%) 291 (73.7%) SARS-CoV-2, N (%) 90 (22.8%) COPD exacerbation, N (%) 21 (5.32%) Acute asthma, N (%) 18 (4.56%) Pneumonia, N (%) 49 (12.4%) Cardiogenic pulmonary oedema, N (%) 21 (5.32%) Respiratory acute viral syndrome (undetermined), N (%) 72 (18.3%) Others, N (%) 20 (5.06%) Extra-respiratory disease, N (%) 104 (26.3%) Vital signs Heart rate, median (IQR) 88 (77–101) Systolic blood pressure, median (IQR) 129 (115–144) Diastolic blood pressure, median (IQR) 77 (68–87) SpO2, median (IQR) 97 (95–99) Ventilatory rate, median (IQR) 22 (18–28) Temperature (°C), median (IQR) 37 (36.6–37.5) Blood gas analysis N = 430 Sample from COVID-19 patient, N (%) 94 (21.9%) FiO2 (%), median (IQR) 21 (21–29) PaO2 (mm Hg), median (IQR) 83.3 (71.3–97.5) SaO2 (%), median (IQR) 97.9 (96.4–98.9) PaCO2 (mm Hg), median (IQR) 36.0 (32.3–40.5) pH, median (IQR) 7.44 (7.41–7.47) PaO2/FiO2 (mm Hg), median (IQR) 364.3 (269.5–450) PaO2/FiO2 > 400, N (%) 164 (38.1%) PaO2/FiO2 300–400, N (%) 132 (30.7%) PaO2/FiO2 < 300, N (%) 134 (31.2%) SaO2/FiO2, median (IQR) 461.2 (339.0–470.5) SpO2/FiO2, median (IQR) 452.4 (337.9–466.7) Table 1 Characteristics of study subjects and samples.

Trends in reasons for emergency calls during the COVID-19 crisis in the department of Gironde, France using artificial neural network for natural language classification.

OBJECTIVES: During periods such as the COVID-19 crisis, there is a need for responsive public health surveillance indicators in order to monitor both the epidemic growth and potential public health consequences of preventative measures such as lockdown. We assessed whether the automatic classification of the content of calls to emergency medical communication centers could provide relevant and responsive indicators. METHODS: We retrieved all 796,209 free-text call reports from the emergency medical communication center of the Gironde department, France, between 2018 and 2020. We trained a natural language processing neural network model with a mixed unsupervised/supervised method to classify all reasons for calls in 2020. Validation and parameter adjustment were performed using a sample of 39,907 manually-coded free-text reports. RESULTS: The number of daily calls for flu-like symptoms began to increase from February 21, 2020 and reached an unprecedented level by February 28, 2020 and peaked on March 14, 2020, 3 days before lockdown. It was strongly correlated with daily emergency room admissions, with a delay of 14 days. Calls for chest pain and stress and anxiety, peaked 12 days later. Calls for malaises with loss of consciousness, non-voluntary injuries and alcohol intoxications sharply decreased, starting one month before lockdown. No noticeable trends in relation to lockdown was found for other groups of reasons including gastroenteritis and abdominal pain, stroke, suicide and self-harm, pregnancy and delivery problems. DISCUSSION: The first wave of the COVID-19 crisis came along with increased levels of stress and anxiety but no increase in alcohol intoxication and violence. As expected, call related to road traffic crashes sharply decreased. The sharp decrease in the number of calls for malaise was more surprising. CONCLUSION: The content of calls to emergency medical communication centers is an efficient epidemiological surveillance data source that provides insights into the societal upheavals induced by a health crisis. The use of an automatic classification system using artificial intelligence makes it possible to free itself from the context that could influence a human coder, especially in a crisis situation. The COVID-19 crisis and/or lockdown induced deep modifications in the population health profile.