Prediction of Clinical Course in COVID19 Patients

Prediction of Clinical Course in COVID19 Patients Using Unsupervised Classification Approaches of Clinical, Biological and the Multiparametric Signature of the Chest CT Scan Performed at Admission

Status

Completed

Phases

Unknown

Study type

Observational

Source

ClinicalTrials.gov

Registry ID

NCT04377685

Acronym

COVID-CTPRED

Enrollment

826

Registered

2020-05-06

Start date

2020-03-01

Completion date

2020-12-26

Last updated

2021-11-17

For informational purposes only — not medical advice. Sourced from public registries and may not reflect the latest updates. Terms

Conditions

COVID 19

Keywords

COVID 19, chest CT scan

Brief summary

In the context of the COVID19 pandemic and containment, chest CT is currently frequently performed on admission, looking for suggestive signs and basic abnormalities of COVID19 compatible viral pneumonitis pending confirmation of identification of viral RNA by reverse-transcription polymerase chain reaction(PCR), with a reported sensitivity of 56-88% in the first few days, slightly higher than PCR (60%) (1). Nevertheless, currently established radiological abnormalities are not specific for COVID19 and the specificity of the chest CT is \ 25% when PCR is used as a reference (1). Deconfinement and its consequences will complicate the triage of COVID patients and the role of the scanner, with the expected impact of a decrease in the prevalence of infection in the emergency department and an increase in the number of all-round patients, including patients with non-COVID viral infiltrates or pneumopathies. In addition, there are currently no imaging criteria to complement the clinical and biological data that can predict the progression of lung disease from the initial data.

Detailed description

In image processing, computational medical imaging has demonstrated its ability to predict a therapeutic response or a particular evolution after extracting relevant anatomical, functional or even non-visually perceptible information from the volume of images, making it possible to construct a powerful radiomic signature or to use robust anatomical/functional measurements to provide estimates of ventilation or vascular state. By combining these data extracted from the scanner with the standard clinical-biological data produced at admission during triage, our ambition is to build a predictive model using unsupervised classification approaches capable of helping predict clinical evolution with the aim of optimizing the management of the resource.

Interventions

OTHERCT-Scan

Chest CT scan on admission to the hospital

Study design

Observational model

COHORT

Time perspective

RETROSPECTIVE

Eligibility

Sex/Gender

ALL

Age

18 Years to No maximum

Inclusion criteria

* age ≥ 18 years * clinical suspicion of COVID-19 confirmed by RT-PCR * CT scan at ER admission * RT-PCR sampling

Exclusion criteria

* CT scan failure or loss of CT data * RT-PCR initial results unavailable

Design outcomes

Primary

Measure	Time frame	Description
diagnostic of COVID disease composite	On admission to the hospital	The diagnostoc of COVID disease is composite of: * CT features wich will include presence/location/laterality of morphological CT abonormal densities (ground glass opacities, consolidations, reticulations), * pulmonary vessels size, * distribution and abnormalities, * local / global CT-ventilation index (CT-VI) severity, * radiomic features (shape features, 1st-order and 2nd order statistics) Analysis of CT-Scan results.

Countries

France

Outcome results

None listed

Source: ClinicalTrials.gov · Data processed: Feb 4, 2026