Construction of a Composite Clinical-echo Score Predictive of a Risk of Short-term Aggravation of Respiratory Impairment in Patients Suspected of Covid-19

With the influx of patients suspected of Covid-19 and the limited number of hospital beds, there is a need for sensitive triage to detect patients at risk of pulmonary complications and therefore requiring hospitalization, but also specific triage to safely discharge patients without risk factors or signs of clinical or ultrasound severity. The use of pulmonary ultrasound in addition to clinical assessment seems appropriate. Indeed, it allows early detection of signs of pneumopathy which, in the current context, most often correspond to Covid-19. These signs include B-lines, which indicate interstitial pulmonary oedema, and an anfractuous and thickened pleural line, or even centimetric parenchymal condensations with a low level of pleural effusion. Conversely, the presence of a medium to large pleural effusion is not very suggestive of the diagnosis of Covid-19. In addition, a lung ultrasound score has been developed and validated to assess the severity of acute respiratory distress and predict the occurrence of acute respiratory distress syndrome. It is based on the performance of a 12-point (6 per hemi-thorax) pulmonary ultrasound with the collection of the presence of B-lines, condensation or pleural effusion. In the hands of a trained operator, this examination takes only a few minutes. The aim of the study is to develop a score based on clinical and ultrasound evidence to allow early and safer referral than that based on clinical evidence alone. To do this, the study will retrospectively collect clinical and lung ultrasound data from departments that use this technique on a daily basis.

The objective of the VIRUS research is therefore to develop an innovative clinical-echo score (VIRUScore), based on risk factors, clinical elements and ultrasound data. This predictive score corresponds to a probabilistic measure of the individual risk of aggravation and is intended to become the cornerstone of a decision-making algorithm for triaging/managing COVID-19 patients (VIRUS algorithm). Initial VIRUScore and evolution of clinical and/or clinico-biological signs will thus have to arbitrate different patient pathway scenarios with 2 major objectives: first, to reduce hospital tension and desaturate emergency departments and COVID-19 units to ensure maximum monitoring of moderate forms, some of which are likely to evolve towards severe forms. The negative predictive value of severe short-term aggravation (H48) should therefore be maximized for these patient profiles invited to return home and/or transfer to non-specialized hospitals or clinics. And secondly, it should be sensitive in detecting and predicting the most severe forms with a high risk of resuscitative escalation and/or death, allowing prioritization (access to CT, access to research protocols) and grading the intensity of clinical surveillance, for anticipation of resuscitative resources.