Value of initial chest radiographs for predicting clinical outcomes in patients with severe acute respiratory syndrome

Radiographic patterns and severity scoring of COVID-19 pneumonia in children: a retrospective study

BACKGROUND

Chest radiography (CXR) is an adjunct tool in treatment planning and monitoring of the disease course of COVID-19 pneumonia. The purpose of the study was to describe the radiographic patterns and severity scores of abnormal CXR findings in children diagnosed with COVID-19 pneumonia.

METHODS

This retrospective study included children with confirmed COVID-19 by reverse transcriptase-polymerase chain reaction test who underwent CXR at the arrival. The CXR findings were reviewed, and modified radiographic scoring was assessed.

RESULTS

The number of abnormal CXR findings was 106 of 976 (10.9%). Ground-glass opacity (GGO) was commonly found in children aged > 9 years (19/26, 73.1%), whereas peribronchial thickening was predominantly found in children aged < 5 years (25/54, 46.3%). Overall, the most common radiographic finding was peribronchial thickening (54/106, 51%). The lower lung zone (56/106, 52.8%) was the most common affected area, and there was neither peripheral nor perihilar predominance (84/106, 79.2%). Regarding the severity of COVID-19 pneumonia based on abnormal CXR findings, 81 of 106 cases (76.4%) had mild lung abnormalities. Moderate and severe lung abnormalities were found in 21 (19.8%) and 4 (3.8%) cases, respectively. While there were no significant differences in the radiographic severity scores among the various pediatric age groups, there were significant disparities in severity scores in the initial CXR and medical treatments.

CONCLUSIONS

This study clarified the age distribution of radiographic features across the pediatric population. GGO was commonly found in children aged > 9 years, whereas peribronchial thickening was predominant in children aged < 5 years. The lower lung zone was the most common affected area, and the high severity lung scores required more medical treatments and oxygen support.

Evaluation of Patients with COVID-19 Followed Up in Intensive Care Units in the Second Year of the Pandemic: A Multicenter Point Prevalence Study

OBJECTIVE

A 1-day point prevalence study was planned to obtain country data by determining the clinical characteristics, follow-up and treatment methods of coronavirus disease 2019 (COVID-19) cases that required intensive care unit (ICU) treatment in the second year of the pandemic.

MATERIAL AND METHODS

All patients who were hospitalized in the ICUs due to COVID-19 between March 11, 2022, 08.00 am, and March 12, 2022, 08.00 am, were included in the study. Demographic characteristics, intensive care and laboratory data, radiological characteristics, and follow-up results of the patients were recorded.

RESULTS

A total of 811 patients from 59 centers were included in the study, 59% of the cases were male, and the mean age was 74 ± 14 years. At least one comorbid disease was present in 94% of the cases, and hypertension was the most common. When ICU weight scores were examined, Acute Physiology and Chronic Health Evaluation-II: 19 (15-27) and Sequential Organ Failure Assessment: 7 (4-10) were seen. Sepsis was present in 37% (n = 298) of cases. PaO2/FiO2 ratios of the patients were 190 the highest and 150 the lowest and 51% of the cases were followed via invasive mechanical ventilation. On the study day, 73% bilateral involvement was seen on chest x-ray, and ground-glass opacities (52%) were the most common on chest tomography. There was growth in culture in 40% (n = 318) of the cases, and the most common growth was in the tracheal aspirate (42%).

CONCLUSION

The clinical course of COVID-19 is variable, and ICU follow-up was required due to advanced age, comorbidity, presence of respiratory symptoms, and widespread radiological involvement. The need for respiratory support and the presence of secondary infection are important issues to be considered in the follow-up. Despite the end of the second year of the pandemic and vaccination, the high severity of the disease as well as the need for follow-up in ICUs has shown that COVID-19 is an important health problem.

Deep Learning-Based Time-to-Death Prediction Model for COVID-19 Patients Using Clinical Data and Chest Radiographs

Accurate estimation of mortality and time to death at admission for COVID-19 patients is important and several deep learning models have been created for this task. However, there are currently no prognostic models which use end-to-end deep learning to predict time to event for admitted COVID-19 patients using chest radiographs and clinical data. We retrospectively implemented a new artificial intelligence model combining DeepSurv (a multiple-perceptron implementation of the Cox proportional hazards model) and a convolutional neural network (CNN) using 1356 COVID-19 inpatients. For comparison, we also prepared DeepSurv only with clinical data, DeepSurv only with images (CNNSurv), and Cox proportional hazards models. Clinical data and chest radiographs at admission were used to estimate patient outcome (death or discharge) and duration to the outcome. The Harrel's concordance index (c-index) of the DeepSurv with CNN model was 0.82 (0.75-0.88) and this was significantly higher than the DeepSurv only with clinical data model (c-index = 0.77 (0.69-0.84), p = 0.011), CNNSurv (c-index = 0.70 (0.63-0.79), p = 0.001), and the Cox proportional hazards model (c-index = 0.71 (0.63-0.79), p = 0.001). These results suggest that the time-to-event prognosis model became more accurate when chest radiographs and clinical data were used together.

Development and Validation of a Multimodal-Based Prognosis and Intervention Prediction Model for COVID-19 Patients in a Multicenter Cohort

The ability to accurately predict the prognosis and intervention requirements for treating highly infectious diseases, such as COVID-19, can greatly support the effective management of patients, especially in resource-limited settings. The aim of the study is to develop and validate a multimodal artificial intelligence (AI) system using clinical findings, laboratory data and AI-interpreted features of chest X-rays (CXRs), and to predict the prognosis and the required interventions for patients diagnosed with COVID-19, using multi-center data. In total, 2282 real-time reverse transcriptase polymerase chain reaction-confirmed COVID-19 patients’ initial clinical findings, laboratory data and CXRs were retrospectively collected from 13 medical centers in South Korea, between January 2020 and June 2021. The prognostic outcomes collected included intensive care unit (ICU) admission and in-hospital mortality. Intervention outcomes included the use of oxygen (O2) supplementation, mechanical ventilation and extracorporeal membrane oxygenation (ECMO). A deep learning algorithm detecting 10 common CXR abnormalities (DLAD-10) was used to infer the initial CXR taken. A random forest model with a quantile classifier was used to predict the prognostic and intervention outcomes, using multimodal data. The area under the receiver operating curve (AUROC) values for the single-modal model, using clinical findings, laboratory data and the outputs from DLAD-10, were 0.742 (95% confidence interval [CI], 0.696−0.788), 0.794 (0.745−0.843) and 0.770 (0.724−0.815), respectively. The AUROC of the combined model, using clinical findings, laboratory data and DLAD-10 outputs, was significantly higher at 0.854 (0.820−0.889) than that of all other models (p < 0.001, using DeLong’s test). In the order of importance, age, dyspnea, consolidation and fever were significant clinical variables for prediction. The most predictive DLAD-10 output was consolidation. We have shown that a multimodal AI model can improve the performance of predicting both the prognosis and intervention in COVID-19 patients, and this could assist in effective treatment and subsequent resource management. Further, image feature extraction using an established AI engine with well-defined clinical outputs, and combining them with different modes of clinical data, could be a useful way of creating an understandable multimodal prediction model.

An overview of literature on COVID-19, MERS and SARS: Using text mining and latent Dirichlet allocation

The unprecedented outbreak of COVID-19 is one of the most serious global threats to public health in this century. During this crisis, specialists in information science could play key roles to support the efforts of scientists in the health and medical community for combatting COVID-19. In this article, we demonstrate that information specialists can support health and medical community by applying text mining technique with latent Dirichlet allocation procedure to perform an overview of a mass of coronavirus literature. This overview presents the generic research themes of the coronavirus diseases: COVID-19, MERS and SARS, reveals the representative literature per main research theme and displays a network visualisation to explore the overlapping, similarity and difference among these themes. The overview can help the health and medical communities to extract useful information and interrelationships from coronavirus-related studies.

Associations between CT pulmonary opacity score on admission and clinical characteristics and outcomes in patients with COVID-19

This study investigated associations between chest computed tomography (CT) pulmonary opacity score on admission and clinical features and outcomes in COVID-19 patients. The retrospective multi-center cohort study included 496 COVID-19 patients in Jiangsu province, China diagnosed as of March 15, 2020. Patients were divided into four groups based on the quartile of pulmonary opacity score: ≤ 5%, 6-20%, 21-40% and 41% +. CT pulmonary opacity score was independently associated with age, single onset, fever, cough, peripheral capillary oxygen saturation, lymphocyte count, platelet count, albumin level, C-reactive protein (CRP) level and fibrinogen level on admission. Patients with score ≥ 41% had a dramatic increased risk of severe or critical illness [odds ratio (OR), 15.58, 95% confidence interval (CI) 3.82-63.53), intensive care unit (ICU)] admission (OR, 6.26, 95% CI 2.15-18.23), respiratory failure (OR, 19.49, 95% CI 4.55-83.40), and a prolonged hospital stay (coefficient, 2.59, 95% CI 0.46-4.72) compared to those with score ≤ 5%. CT pulmonary opacity score on admission, especially when ≥ 41%, was closely related to some clinical characteristics and was an independent predictor of disease severity, ICU admission, respiratory failure and long hospital stay in patients with COVID-19.

Imaging of COVID-19

The novel coronavirus disease 2019 (COVID-19) emerged as the source of a global pandemic in late 2019 and early 2020 and quickly spread throughout the world becoming one of the worst pandemics in recent history. This chapter reviews the most up to date radiological literature and outlines the utility of thoracic imaging in COVID-19, defining both the common and the less typical imaging appearances during the acute and subacute phases of COVID-19. The short term complications and the long term sequela will also be discussed in the context of radiology, including pulmonary emboli, acute respiratory distress syndrome, superimposed infections, barotrauma, cardiac manifestations, pulmonary parenchymal scarring and fibrosis.

[Development of severity and mortality prediction models for covid-19 patients at emergency department including the chest x-ray]

Objectives

To develop prognosis prediction models for COVID-19 patients attending an emergency department (ED) based on initial chest X-ray (CXR), demographics, clinical and laboratory parameters.

Methods

All symptomatic confirmed COVID-19 patients admitted to our hospital ED between February 24th and April 24th 2020 were recruited. CXR features, clinical and laboratory variables and CXR abnormality indices extracted by a convolutional neural network (CNN) diagnostic tool were considered potential predictors on this first visit. The most serious individual outcome defined the three severity level: 0) home discharge or hospitalization ≤ 3 days, 1) hospital stay >3 days and 2) intensive care requirement or death. Severity and in-hospital mortality multivariable prediction models were developed and internally validated. The Youden index was used for the optimal threshold selection of the classification model.

Results

A total of 440 patients were enrolled (median 64 years; 55.9% male); 13.6% patients were discharged, 64% hospitalized, 6.6% required intensive care and 15.7% died. The severity prediction model included oxygen saturation/inspired oxygen fraction (SatO2/FiO2), age, C-reactive protein (CRP), lymphocyte count, extent score of lung involvement on CXR (ExtScoreCXR), lactate dehydrogenase (LDH), D-dimer level and platelets count, with AUC-ROC = 0.94 and AUC-PRC = 0.88. The mortality prediction model included age, SatO2/FiO2, CRP, LDH, CXR extent score, lymphocyte count and D-dimer level, with AUC-ROC = 0.97 and AUC-PRC = 0.78. The addition of CXR CNN-based indices did not improve significantly the predictive metrics.

Conclusion

The developed and internally validated severity and mortality prediction models could be useful as triage tools in ED for patients with COVID-19 or other virus infections with similar behaviour.

Interobserver Agreement in Semi-Quantitative Scale-Based Interpretation of Chest Radiographs in COVID-19 Patients

Background

The chest X-ray is the most available imaging modality enabling semi-quantitative evaluation of pulmonary involvement. Parametric evaluation of chest radiographs in patients with SARS-CoV-2 infection is crucial for triage and therapeutic management. The CXR Score (Brixia Score), SARI CXR Severity Scoring System, and Radiographic Assessment of Lung Edema (RALE), proposed to evaluate SARS-CoV-2 infiltration of the lungs, were analyzed for interobserver agreement.

Material/Methods

This study analyzed 200 chest X-rays from 200 consecutive patients with confirmed SARS-CoV-2 infection, hospitalized at the Central Clinical Hospital of the Ministry of the Interior and Administration in Warsaw. Radiographs were evaluated by 2 radiologists according to 3 scales: SARI, RALE, and CXR Score.

Results

The overall interobserver agreement for SARI ratings was good (κ=0.755; 95% CI, 0.817–0.694), for RALE scale assessments it was very good (κ=0.818; 95% CI, 0.844–0.793), and for CXR scale assessments it was very good (κ=0.844; 95% CI, 0.846–0.841). A moderate correlation was found between the radiological image assessed using each of the scales and the clinical condition of the patient in MEWS (Modified Early Warning Score) (r=0.425–0.591).

Conclusions

The analyzed scales are characterized by good or very good interobserver agreement of assessments of the extent of pulmonary infiltration. Since the CXR Score showed the strongest correlation with the clinical condition of the patient as expressed using the MEWS scale, it is the preferred scale for chest radiograph assessment of patients with COVID-19 in the light of data provided.

Review of Chest Radiograph Findings of COVID-19 Pneumonia and Suggested Reporting Language

The diagnosis of coronavirus disease 2019 (COVID-19) is confirmed by reverse transcription polymerase chain reaction. The utility of chest radiography (CXR) remains an evolving topic of discussion. Current reports of CXR findings related to COVID-19 contain varied terminology as well as various assessments of its sensitivity and specificity. This can lead to a misunderstanding of CXR reports and makes comparison between examinations and research studies challenging. With this need for consistency, we propose language for standardized CXR reporting and severity assessment of persons under investigation for having COVID-19, patients with a confirmed diagnosis of COVID-19, and patients who may have radiographic findings typical or suggestive of COVID-19 when the diagnosis is not suspected clinically. We recommend contacting the referring providers to discuss the likelihood of viral infection when typical or indeterminate features of COVID-19 pneumonia on CXR are present as an incidental finding. In addition, we summarize the currently available literature related to the use of CXR for COVID-19 and discuss the evolving techniques of obtaining CXR in COVID-19-positive patients. The recently published expert consensus statement on reporting chest computed tomography findings related to COVID-19, endorsed by the Radiological Society of North American (RSNA), the Society of Thoracic Radiology (STR), and American College of Radiology (ACR), serves as the framework for our proposal.

Prognostication of patients with COVID-19 using artificial intelligence based on chest x-rays and clinical data: a retrospective study

BACKGROUND

Chest x-ray is a relatively accessible, inexpensive, fast imaging modality that might be valuable in the prognostication of patients with COVID-19. We aimed to develop and evaluate an artificial intelligence system using chest x-rays and clinical data to predict disease severity and progression in patients with COVID-19.

METHODS

We did a retrospective study in multiple hospitals in the University of Pennsylvania Health System in Philadelphia, PA, USA, and Brown University affiliated hospitals in Providence, RI, USA. Patients who presented to a hospital in the University of Pennsylvania Health System via the emergency department, with a diagnosis of COVID-19 confirmed by RT-PCR and with an available chest x-ray from their initial presentation or admission, were retrospectively identified and randomly divided into training, validation, and test sets (7:1:2). Using the chest x-rays as input to an EfficientNet deep neural network and clinical data, models were trained to predict the binary outcome of disease severity (ie, critical or non-critical). The deep-learning features extracted from the model and clinical data were used to build time-to-event models to predict the risk of disease progression. The models were externally tested on patients who presented to an independent multicentre institution, Brown University affiliated hospitals, and compared with severity scores provided by radiologists.

FINDINGS

1834 patients who presented via the University of Pennsylvania Health System between March 9 and July 20, 2020, were identified and assigned to the model training (n=1285), validation (n=183), or testing (n=366) sets. 475 patients who presented via the Brown University affiliated hospitals between March 1 and July 18, 2020, were identified for external testing of the models. When chest x-rays were added to clinical data for severity prediction, area under the receiver operating characteristic curve (ROC-AUC) increased from 0·821 (95% CI 0·796-0·828) to 0·846 (0·815-0·852; p<0·0001) on internal testing and 0·731 (0·712-0·738) to 0·792 (0·780-0 ·803; p<0·0001) on external testing. When deep-learning features were added to clinical data for progression prediction, the concordance index (C-index) increased from 0·769 (0·755-0·786) to 0·805 (0·800-0·820; p<0·0001) on internal testing and 0·707 (0·695-0·729) to 0·752 (0·739-0·764; p<0·0001) on external testing. The image and clinical data combined model had significantly better prognostic performance than combined severity scores and clinical data on internal testing (C-index 0·805 vs 0·781; p=0·0002) and external testing (C-index 0·752 vs 0·715; p<0·0001).

INTERPRETATION

In patients with COVID-19, artificial intelligence based on chest x-rays had better prognostic performance than clinical data or radiologist-derived severity scores. Using artificial intelligence, chest x-rays can augment clinical data in predicting the risk of progression to critical illness in patients with COVID-19.

FUNDING

Brown University, Amazon Web Services Diagnostic Development Initiative, Radiological Society of North America, National Cancer Institute and National Institute of Biomedical Imaging and Bioengineering of the National Institutes of Health.

Utility of Screening Chest Radiographs in Patients with Asymptomatic or Minimally Symptomatic COVID-19 in Singapore

Background Singapore saw an escalation of coronavirus disease 2019 (COVID-19) cases from fewer than 4000 in April 2020 to more than 40 000 in June 2020, with most of these cases attributed to spread within shared facilities housing foreign workers. Appropriate triage and escalation of clinical care are crucial for this patient group managed in community care facilities (CCFs). Purpose To evaluate the imaging guideline recommendations for COVID-19 from the Fleischner Society and to analyze the clinical utility of screening chest radiography for asymptomatic or minimally symptomatic patients with COVID-19. Materials and Methods In this retrospective study, patients with reverse-transcription polymerase chain reaction-confirmed COVID-19 who were admitted to a designated CCF for continuation of their treatment during May 3-31, 2020, were identified. Upon admission, patients aged 36 years and older without any baseline chest images underwent chest radiography. All chest radiographs and clinical outcomes of patients, including those who were subsequently transferred to acute hospitals for escalation of care, were reviewed. Key proportions of patients with findings of pulmonary infection and those requiring further inpatient treatment were calculated, and 95% binomial proportion CIs were obtained using the Clopper-Pearson method. Results The study included 5621 patients. All patients were men (100%; 5621 of 5621), and the mean patient age was 37 years ± 8 (range, 17-60 years). A total of 1964 chest radiographs were obtained, of which normal images accounted for 98.0% (1925 of 1964 radiographs) and findings of pulmonary infection represented 2.0% (39 of 1964 radiographs). Only 0.2% of patients (four of 1964) with findings of pulmonary infection at chest radiography (all of whom were symptomatic) required supplemental oxygenation and inpatient treatment. None of the asymptomatic patients with findings of pulmonary infection required supplemental oxygenation, and they received only symptomatic treatment. Conclusion In accordance with Fleischner Society recommendations, screening chest radiography is not indicated in patients with coronavirus disease 2019 who are aged 17-60 years with mild or no symptoms unless there is risk of clinical deterioration. © RSNA, 2021 See also the editorial by Schaefer-Prokop and Prokop in this issue.

Initial chest radiographs and artificial intelligence (AI) predict clinical outcomes in COVID-19 patients: analysis of 697 Italian patients

OBJECTIVE

To evaluate whether the initial chest X-ray (CXR) severity assessed by an AI system may have prognostic utility in patients with COVID-19.

METHODS

This retrospective single-center study included adult patients presenting to the emergency department (ED) between February 25 and April 9, 2020, with SARS-CoV-2 infection confirmed on real-time reverse transcriptase polymerase chain reaction (RT-PCR). Initial CXRs obtained on ED presentation were evaluated by a deep learning artificial intelligence (AI) system and compared with the Radiographic Assessment of Lung Edema (RALE) score, calculated by two experienced radiologists. Death and critical COVID-19 (admission to intensive care unit (ICU) or deaths occurring before ICU admission) were identified as clinical outcomes. Independent predictors of adverse outcomes were evaluated by multivariate analyses.

RESULTS

Six hundred ninety-seven 697 patients were included in the study: 465 males (66.7%), median age of 62 years (IQR 52-75). Multivariate analyses adjusting for demographics and comorbidities showed that an AI system-based score ≥ 30 on the initial CXR was an independent predictor both for mortality (HR 2.60 (95% CI 1.69 - 3.99; p < 0.001)) and critical COVID-19 (HR 3.40 (95% CI 2.35-4.94; p < 0.001)). Other independent predictors were RALE score, older age, male sex, coronary artery disease, COPD, and neurodegenerative disease.

CONCLUSION

AI- and radiologist-assessed disease severity scores on CXRs obtained on ED presentation were independent and comparable predictors of adverse outcomes in patients with COVID-19.

TRIAL REGISTRATION

ClinicalTrials.gov NCT04318366 ( https://clinicaltrials.gov/ct2/show/NCT04318366 ).

KEY POINTS

• AI system-based score ≥ 30 and a RALE score ≥ 12 at CXRs performed at ED presentation are independent and comparable predictors of death and/or ICU admission in COVID-19 patients. • Other independent predictors are older age, male sex, coronary artery disease, COPD, and neurodegenerative disease. • The comparable performance of the AI system in relation to a radiologist-assessed score in predicting adverse outcomes may represent a game-changer in resource-constrained settings.

The prognostic value of chest X-ray in patients with COVID-19 on admission and when starting CPAP

OBJECTIVE

The objective was to explore if chest X-ray severity, assessed using a validated scoring system, predicts patient outcome on admission and when starting continuous positive pressure ventilation (CPAP) for COVID-19.

DESIGN

The study was a retrospective case-controlled study.

PARTICIPANTS

There were 163 patients with COVID-19 deemed candidates for CPAP on admission, including 58 who subsequently required CPAP.

OUTCOME MEASURES

On admission, we measured the proportion of patients meeting a composite 'negative' outcome of requiring CPAP, intubation or dying versus successful ward-based care. For those escalated to CPAP, 'negative' outcomes were intubation or death versus successful de-escalation of respiratory support.

RESULTS

Our results were stratified into tertiles, those with 'moderate' or 'severe' X-rays on admission had significantly higher odds of negative outcome versus 'mild' (odds ratio (OR) 2.32; 95% confidence interval (CI) 1.121-4.803; p=0.023; and OR 3.600; 95% CI 1.681-7.708; p=0.001, respectively). This could not be demonstrated in those commencing CPAP (OR 0.976; 95% CI 0.754-1.264; p=0.856).

CONCLUSIONS

We outline a scoring system to stratify X-rays by severity and directly link this to prognosis. However, we were unable to demonstrate this association in the patients commencing CPAP.

Model-based Prediction of Critical Illness in Hospitalized Patients with COVID-19

Background The prognosis of hospitalized patients with severe coronavirus disease 2019 (COVID-19) is difficult to predict, and the capacity of intensive care units was a limiting factor during the peak of the pandemic and is generally dependent on a country's clinical resources. Purpose To determine the value of chest radiographic findings together with patient history and laboratory markers at admission to predict critical illness in hospitalized patients with COVID-19. Materials and Methods In this retrospective study, which included patients from March 7, 2020, to April 24, 2020, a consecutive cohort of hospitalized patients with real-time reverse transcription polymerase chain reaction-confirmed COVID-19 from two large Dutch community hospitals was identified. After univariable analysis, a risk model to predict critical illness (ie, death and/or intensive care unit admission with invasive ventilation) was developed, using multivariable logistic regression including clinical, chest radiographic, and laboratory findings. Distribution and severity of lung involvement were visually assessed by using an eight-point scale (chest radiography score). Internal validation was performed by using bootstrapping. Performance is presented as an area under the receiver operating characteristic curve. Decision curve analysis was performed, and a risk calculator was derived. Results The cohort included 356 hospitalized patients (mean age, 69 years ± 12 [standard deviation]; 237 men) of whom 168 (47%) developed critical illness. The final risk model's variables included sex, chronic obstructive lung disease, symptom duration, neutrophil count, C-reactive protein level, lactate dehydrogenase level, distribution of lung disease, and chest radiography score at hospital presentation. The area under the receiver operating characteristic curve of the model was 0.77 (95% CI: 0.72, 0.81; P < .001). A risk calculator was derived for individual risk assessment: Dutch COVID-19 risk model. At an example threshold of 0.70, 71 of 356 patients would be predicted to develop critical illness, of which 59 (83%) would be true-positive results. Conclusion A risk model based on chest radiographic and laboratory findings obtained at admission was predictive of critical illness in hospitalized patients with coronavirus disease 2019. This risk calculator might be useful for triage of patients to the limited number of intensive care unit beds or facilities. © RSNA, 2020 Online supplemental material is available for this article.

Comparing a deep learning model's diagnostic performance to that of radiologists to detect Covid -19 features on chest radiographs

BACKGROUND

Whether the sensitivity of Deep Learning (DL) models to screen chest radiographs (CXR) for CoVID-19 can approximate that of radiologists, so that they can be adopted and used if real-time review of CXRs by radiologists is not possible, has not been explored before.

OBJECTIVE

To evaluate the diagnostic performance of a doctor-trained DL model (Svita_DL8) to screen for COVID-19 on CXR, and to compare the performance of the DL model with that of expert radiologists.

MATERIALS AND METHODS

We used a pre-trained convolutional neural network to develop a publicly available online DL model to evaluate CXR examinations saved in .jpeg or .png format. The initial model was subsequently curated and trained by an internist and a radiologist using 1062 chest radiographs to classify a submitted CXR as either normal, COVID-19, or a non-COVID-19 abnormal. For validation, we collected a separate set of 430 CXR examinations from numerous publicly available datasets from 10 different countries, case presentations, and two hospital repositories. These examinations were assessed for COVID-19 by the DL model and by two independent radiologists. Diagnostic performance was compared between the model and the radiologists and the correlation coefficient calculated.

RESULTS

For detecting COVID-19 on CXR, our DL model demonstrated sensitivity of 91.5%, specificity of 55.3%, PPV 60.9%, NPV 77.9%, accuracy 70.1%, and AUC 0.73 (95% CI: 0.86, 0.95). There was a significant correlation (r = 0.617, P = 0.000) between the results of the DL model and the radiologists' interpretations. The sensitivity of the radiologists is 96% and their overall diagnostic accuracy is 90% in this study.

CONCLUSIONS

The DL model demonstrated high sensitivity for detecting COVID-19 on CXR.

CLINICAL IMPACT

The doctor trained DL tool Svita_DL8 can be used in resource-constrained settings to quickly triage patients with suspected COVID-19 for further in-depth review and testing.

Prognostic value of initial chest CT findings for clinical outcomes in patients with COVID-19

Rationale: To identify whether the initial chest computed tomography (CT) findings of patients with coronavirus disease 2019 (COVID-19) are helpful for predicting the clinical outcome. Methods: A total of 224 patients with laboratory-confirmed COVID-19 who underwent chest CT examination within the first day of admission were enrolled. CT findings, including the pattern and distribution of opacities, the number of lung lobes involved and the chest CT scores of lung involvement, were assessed. Independent predictors of adverse clinical outcomes were determined by multivariate regression analysis. Adverse outcome were defined as the need for mechanical ventilation or death. Results: Of 224 patients, 74 (33%) had adverse outcomes and 150 (67%) had good outcomes. There were higher frequencies of more than four lung zones involved (73% vs 32%), both central and peripheral distribution (57% vs 42%), consolidation (27% vs 17%), and air bronchogram (24% vs 13%) and higher initial chest CT scores (8.6±3.4 vs 5.4±2.1) (P < 0.05 for all) in the patients with poor outcomes. Multivariate analysis demonstrated that more than four lung zones (odds ratio [OR] 3.93; 95% confidence interval [CI]: 1.44 to 12.89), age above 65 (OR 3.65; 95% CI: 1.11 to 10.59), the presence of comorbidity (OR 5.21; 95% CI: 1.64 to 19.22) and dyspnea on admission (OR 3.19; 95% CI: 1.35 to 8.46) were independent predictors of poor outcome. Conclusions: Involvement of more than four lung zones and a higher CT score on the initial chest CT were significantly associated with adverse clinical outcome. Initial chest CT findings may be helpful for predicting clinical outcome in patients with COVID-19.

Analysis of Chest CT Results of Coronavirus Disease 2019 (COVID-19) Patients at First Follow-Up

Objective

To investigate the dissipation and outcomes of pulmonary lesions at the first follow-up of patients who recovered from moderate and severe cases of COVID-19.

Methods

From January 21 to March 3, 2020, a total of 136 patients with COVID-19 were admitted to our hospital. According to inclusion and exclusion criteria, 52 patients who recovered from COVID-19 were included in this study, including 33 moderate cases and 19 severe cases. Three senior radiologists independently and retrospectively analyzed the chest CT imaging data of 52 patients at the last time of admission and the first follow-up after discharge, including primary manifestations, concomitant manifestations, and degree of residual lesion dissipation.

Results

At the first follow-up after discharge, 16 patients with COVID-19 recovered to normal chest CT appearance, while 36 patients still had residual pulmonary lesions, mainly including 33 cases of ground-glass opacity, 5 cases of consolidation, and 19 cases of fibrous strip shadow. The proportion of residual pulmonary lesions in severe cases (17/19) was statistically higher than in moderate cases (19/33) (χ ² = 5.759, P < 0.05). At the first follow-up, residual pulmonary lesions were dissipated to varying degrees in 47 cases, and lesions remained unchanged in 5 cases. There were no cases of increased numbers of lesions, enlargement of lesions, or appearance of new lesions. The dissipation of residual pulmonary lesions in moderate patients was statistically better than in severe patients (Z = -2.538, P < 0.05).

Conclusion

Clinically cured patients with COVID-19 had faster dissipation of residual pulmonary lesions after discharge, while moderate patients had better dissipation than severe patients. However, at the first follow-up, most patients still had residual pulmonary lesions, which were primarily ground-glass opacity and fibrous strip shadow. The proportion of residual pulmonary lesions was higher in severe cases of COVID-19, which required further follow-up.

Multi-organ Dysfunction in Patients with COVID-19: A Systematic Review and Meta-analysis

This study aimed to provide systematic evidence for the association between multiorgan dysfunction and COVID-19 development. Several online databases were searched for articles published until May 13, 2020. Two investigators independently selected trials, extracted data, and evaluated the quality of individual trials. Single-arm meta-analysis was performed to summarize the clinical features of confirmed COVID-19 patients. Fixed effects meta-analysis was performed for clinically relevant parameters that were closely related to the patients' various organ functions. A total of 73 studies, including 171,108 patients, were included in this analysis. The overall incidence of severe COVID-19 and mortality were 24% (95% confidence interval [CI], 20%-28%) and 2% (95% CI, 1%-3%), respectively. Patients with hypertension (odds ratio [OR] = 2.40; 95% CI, 2.08-2.78), cardiovascular disease (CVD) (OR = 3.54; 95% CI, 2.68-4.68), chronic obstructive pulmonary disease (COPD) (OR=3.70; 95% CI, 2.93-4.68), chronic liver disease (CLD) (OR=1.48; 95% CI, 1.09-2.01), chronic kidney disease (CKD) (OR = 1.84; 95% CI, 1.47-2.30), chronic cerebrovascular diseases (OR = 2.53; 95% CI, 1.84-3.49) and chronic gastrointestinal (GI) disease (OR = 2.13; 95% CI, 1.12-4.05) were more likely to develop severe COVID-19. Increased levels of lactate dehydrogenase (LDH), creatine kinase (CK), high-sensitivity cardiac troponin I (hs-cTnI), myoglobin, creatinine, urea, alanine aminotransferase (ALT), aspartate aminotransferase (AST), and total bilirubin were highly associated with severe COVID-19. The incidence of acute organ injuries, including acute cardiac injury (ACI); (OR = 11.87; 95% CI, 7.64-18.46), acute kidney injury (AKI); (OR=10.25; 95% CI, 7.60-13.84), acute respiratory distress syndrome (ARDS); (OR=27.66; 95% CI, 18.58-41.18), and acute cerebrovascular diseases (OR=9.22; 95% CI, 1.61-52.72) was more common in patients with severe COVID-19 than in patients with non-severe COVID-19. Patients with a history of organ dysfunction are more susceptible to severe conditions. COVID-19 can aggravate an acute multiorgan injury.

Clinical and Chest Radiography Features Determine Patient Outcomes in Young and Middle-aged Adults with COVID-19

Background Chest radiography has not been validated for its prognostic utility in evaluating patients with coronavirus disease 2019 (COVID-19). Purpose To analyze the prognostic value of a chest radiograph severity scoring system for younger (nonelderly) patients with COVID-19 at initial presentation to the emergency department (ED); outcomes of interest included hospitalization, intubation, prolonged stay, sepsis, and death. Materials and Methods In this retrospective study, patients between the ages of 21 and 50 years who presented to the ED of an urban multicenter health system from March 10 to March 26, 2020, with COVID-19 confirmation on real-time reverse transcriptase polymerase chain reaction were identified. Each patient's ED chest radiograph was divided into six zones and examined for opacities by two cardiothoracic radiologists, and scores were collated into a total concordant lung zone severity score. Clinical and laboratory variables were collected. Multivariable logistic regression was used to evaluate the relationship between clinical parameters, chest radiograph scores, and patient outcomes. Results The study included 338 patients: 210 men (62%), with median age of 39 years (interquartile range, 31-45 years). After adjustment for demographics and comorbidities, independent predictors of hospital admission (n = 145, 43%) were chest radiograph severity score of 2 or more (odds ratio, 6.2; 95% confidence interval [CI]: 3.5, 11; P < .001) and obesity (odds ratio, 2.4 [95% CI: 1.1, 5.4] or morbid obesity). Among patients who were admitted, a chest radiograph score of 3 or more was an independent predictor of intubation (n = 28) (odds ratio, 4.7; 95% CI: 1.8, 13; P = .002) as was hospital site. No significant difference was found in primary outcomes across race and ethnicity or those with a history of tobacco use, asthma, or diabetes mellitus type II. Conclusion For patients aged 21-50 years with coronavirus disease 2019 presenting to the emergency department, a chest radiograph severity score was predictive of risk for hospital admission and intubation. © RSNA, 2020 Online supplemental material is available for this article.

Imaging Findings in Patients With H1N1 Influenza A Infection

Background

Swine influenza (H1N1) is a very contagious respiratory infection and World Health Organization (WHO) has raised the alert level to phase 6 (pandemic). The study of clinical and laboratory manifestations as well as radiologic imaging findings helps in its early diagnosis.

Objectives

The aim of this study was to evaluate the imaging findings of patients with documented H1N1 infection referred to our center.

Patients and Methods

Thirty-one patients (16 men) with documented H1N1 infection were included in our study. The initial radiography obtained from the patients was reviewed regarding pattern (consolidation, ground glass, nodules and reticulation), distribution (focal, multifocal, and diffuse) and the lung zones involved. Computed tomography (CT) scans were also reviewed for the same abnormalities. The patient files were studied for their possible underlying diseases.

Results

The mean age was 37.97 ± 13.9 years. Seventeen (54.8%) patients had co-existing condition (eight respiratory, five cardiovascular, two immunodeficiency, two cancer, four others). Twelve (38.7%) patients required intensive care unit (ICU) admission. Five (16.1%) patients died. (25.8%) had normal initial radiographs. The most common abnormality was consolidation (12/31; 38.7%) in the peripheral region (11/31; 35.5%) followed by peribronchovascular areas (10/31; 32.3%) which was most commonly observed in the lower zone. The patients admitted to the ICU were more likely to have two or more lung zones involved (P = 0.005).

Conclusions

In patients with the novel swine flu infection, the most common radiographic abnormality observed was consolidation in the lower lung zones. Patients admitted to ICU were more likely to have two or more lung zones involved.

H1N1 influenza: initial chest radiographic findings in helping predict patient outcome

PURPOSE

To retrospectively evaluate whether findings on initial chest radiographs of influenza A (H1N1) patients can help predict clinical outcome.

MATERIALS AND METHODS

Institutional review board approval was obtained; informed consent was waived. All adult patients admitted to the emergency department (May to September 2009) with a confirmed diagnosis of H1N1 influenza who underwent frontal chest radiography within 24 hours were included. Radiologic findings were characterized by type and pattern of opacities and zonal distribution. Major adverse outcome measures were mechanical ventilation and death.

RESULTS

Of 179 H1N1 influenza patients, 97 (54%) underwent chest radiography at admission; 39 (40%) of these had abnormal radiologic findings likely related to influenza infection and five (13%) of these 39 had adverse outcomes. Fifty-eight (60%) of 97 patients had normal radiographs; two (3%) of these had adverse outcomes (P = .113). Characteristic imaging findings included the following: ground-glass (69%), consolidation (59%), frequently patchy (41%), and nodular (28%) opacities. Bilateral opacities were common (62%), with involvement of multiple lung zones (72%). Findings in four or more zones and bilateral peripheral distribution occurred with significantly higher frequency in patients with adverse outcomes compared with patients with good outcomes (multizonal opacities: 60% vs 6%, P = .01; bilateral peripheral opacities: 60% vs 15%, P = .049).

CONCLUSION

Extensive involvement of both lungs, evidenced by the presence of multizonal and bilateral peripheral opacities, is associated with adverse prognosis. Initial chest radiography may have significance in helping predict clinical outcome but normal initial radiographs cannot exclude adverse outcome.

A correlation between the severity of lung lesions on radiographs and clinical findings in patients with severe acute respiratory syndrome

Objective

The purpose of this study was to quantify lesions on chest radiographs in patients with severe acute respiratory syndrome (SARS) and analyze the severity of the lesions with clinical parameters.

Materials and Methods

Two experienced radiologists reviewed chest radiographs of 28 patients with SARS. Each lung was divided into upper, middle, and lower zones. A SARS-related lesion in each zone was scored using a four-point scale: zero to three. The mean and maximal radiographic scores were analyzed statistically to determine if the scorings were related to the laboratory data and clinical course.

Results

Forward stepwise multiple linear regression showed that the mean radiographic score correlated most significantly with the number of hospitalized days (p < 0.001). The second most significant factor was the absolute lymphocyte count (p < 0.001) and the third most significant factor was the number of days of intubation (p = 0.025). The maximal radiographic score correlated best with the percentage of lymphocytes in a leukocyte count (p < 0.001), while the second most significant factor was the number of hospitalized days (p < 0.001) and the third most significant factor was the absolute lymphocyte count (p = 0.013). The mean radiographic scores of the patients who died, with comorbidities and without a comorbidity were 11.1, 6.3 and 2.9, respectively (p = 0.032). The corresponding value for maximal radiographic scores were 17.7, 9.7 and 6.0, respectively (p = 0.033).

Conclusion

The severity of abnormalities quantified on chest radiographs in patients with SARS correlates with the clinical parameters.

Prognostication in severe acute respiratory syndrome: a retrospective time-course analysis of 1312 laboratory-confirmed patients in Hong Kong

Background and objective:  The temporal importance of prognostic indicators for severe acute respiratory syndrome (SARS) has not been studied. This study identified the various clinical prognostic factors for SARS and described the temporal evolution of these factors in the course of the SARS illness in Hong Kong in 2003.

Methods:  A retrospective analysis of the entire Hong Kong cohort of 1312 laboratory‐confirmed SARS patients aged 15–74 years was undertaken. Demographic, clinical and laboratory data at presentation and investigative data during the first 10 days of illness from the time of symptom onset were compiled. Two adverse outcomes were examined: hospital mortality and the development of oxygenation failure based on the estimated PaO₂/FiO₂ ratio of <200 mm Hg. Logistic regression was used to identify the association between these prognostic factors and outcomes.

Results:  Based on adjusted odds ratios with a P‐value of <0.05, older age, male gender, elevated pulse rate and elevated neutrophil count were all predictive of oxygenation failure and death during the 10‐day illness. Raised serum albumin and creatinine phosphokinase (CPK) levels were predictive of hospital mortality during this period. The presenting ALT and CPK level and the day 7 and day 10 platelet counts were predictive of oxygenation failure while the day 7 LDH was predictive of death. Contact exposure outside health‐care institutions also appeared to carry higher risk of death.

Conclusion:  This large‐scale analysis identified important discriminatory parameters related to the patients’ demographic profile (age and gender), severity of illness (pulse rate and neutrophil count), and multisystem derangement (platelet count, CPK, ALT and LDH), all of which prognosticated adverse outcomes during the SARS episode. While age, pulse rate and neutrophil count consistently remained significant prognosticators during the first 10 days of illness, the prognostic impact of other derangements was more time‐course dependent. Clinicians should be aware of the time‐course evolution of these prognosticators.

Adverse events associated with high-dose ribavirin: evidence from the Toronto outbreak of severe acute respiratory syndrome

Study Objectives. To distinguish adverse events related to ribavirin therapy from those attributable to severe acute respiratory syndrome (SARS), and to determine the rate of potential ribavirin‐related adverse events.

Design. Retrospective cohort study.

Setting. Hospitals in Toronto, Ontario, Canada.

Patients. A cohort of 306 patients with confirmed or probable SARS, 183 of whom received ribavirin and 123 of whom did not, between February 23, 2003, and July 1, 2003. Of the 183 treated patients, 155 (85%) received very high‐dose ribavirin; the other 28 treated patients received lower‐dose regimens.

Measurements and Main Results. Data on all patients with SARS admitted to hospitals in Toronto were abstracted from charts and electronic databases onto a standardized form by trained research nurses. Logistic regression was used to evaluate the association between ribavirin use and each adverse event (progressive anemia, hypomagnesemia, hypocalcemia, bradycardia, transaminitis, and hyperamylasemia) after adjusting for SARS‐related prognostic factors and corticosteroid use. In the primary logistic regression analysis, ribavirin use was strongly associated with anemia (odds ratio [OR] 3.0, 99% confidence interval [CI] 1.5–6.1, p<0.0001), hypomagnesemia (OR 21, 99% CI 5.8–73, p<0.0001), and bradycardia (OR 2.3, 99% CI 1.0–5.1, p=0.007). Hypocalcemia, transaminitis, and hyperamylasemia were not associated with ribavirin use. The risk of anemia, hypomagnesemia, and bradycardia attributable to ribavirin use was 27%, 45%, and 17%, respectively.

Conclusions. High‐dose ribavirin is associated with a high rate of adverse events. The use of high‐dose ribavirin is appropriate only for the treatment of infectious diseases for which ribavirin has proven clinical efficacy, or in the context of a clinical trial. Ribavirin should not be used empirically for the treatment of viral syndromes of unknown origin.

Cynomolgus macaque as an animal model for severe acute respiratory syndrome

BACKGROUND

The emergence of severe acute respiratory syndrome (SARS) in 2002 and 2003 affected global health and caused major economic disruption. Adequate animal models are required to study the underlying pathogenesis of SARS-associated coronavirus (SARS-CoV) infection and to develop effective vaccines and therapeutics. We report the first findings of measurable clinical disease in nonhuman primates (NHPs) infected with SARS-CoV.

METHODS AND FINDINGS

In order to characterize clinically relevant parameters of SARS-CoV infection in NHPs, we infected cynomolgus macaques with SARS-CoV in three groups: Group I was infected in the nares and bronchus, group II in the nares and conjunctiva, and group III intravenously. Nonhuman primates in groups I and II developed mild to moderate symptomatic illness. All NHPs demonstrated evidence of viral replication and developed neutralizing antibodies. Chest radiographs from several animals in groups I and II revealed unifocal or multifocal pneumonia that peaked between days 8 and 10 postinfection. Clinical laboratory tests were not significantly changed. Overall, inoculation by a mucosal route produced more prominent disease than did intravenous inoculation. Half of the group I animals were infected with a recombinant infectious clone SARS-CoV derived from the SARS-CoV Urbani strain. This infectious clone produced disease indistinguishable from wild-type Urbani strain.

CONCLUSIONS

SARS-CoV infection of cynomolgus macaques did not reproduce the severe illness seen in the majority of adult human cases of SARS; however, our results suggest similarities to the milder syndrome of SARS-CoV infection characteristically seen in young children.