*Correspondence to: Jirong LI, E-mail:304353448@qq.com
[Abstract] Objective: To develop a pneumonia risk prediction model for SARS-CoV-2 infected patients to reduce unnecessary chest CT scans; Materials and Methods:Retrospective analysis was performed on the clinical data of SARS-CoV-2-positive patients who visited outpatient and emergency clinics and underwent chest CT scans at the Mawangdui Branch of Hunan Provincial People’s Hospital from 20 December 2022 to 23 December 2022 and at the Tianxinge Branch of Hunan Provincial People’s Hospital from 1 January 2023 to 4 January 2023. A retrospective analysis of imaging and clinical data from 205 cases (training cohort) and 94 cases (validation cohort) of SARS-CoV-2-positive patients who visited outpatient and emergency clinics was conducted. The predictor variables were screened using the ”univariate and then multivariate logistic regression” and ”least absolute shrinkage and selection operator (LASSO)” approaches, and the predictive model was constructed using multifactorial logistic regression and represented as a nomogram. The diagnostic effectiveness of the pneumonia risk model was evaluated using receiver operating characteristic (ROC) curves; the Delong test and Integrated Discrimination Improvement Index (IDI) were used to compare the AUC of the pneumonia risk model with the AUCs for predictors incorporated in the model alone. The calibration of the pneumonia risk model was assessed using calibration curves; Decision curve analysis (DCA) was used to evaluate the clinical validity of the pneumonia risk model. In addition, a smoothed curve was fitted using a generalized additive model (GAM) to explore the relationship between the pneumonia grade and the model’s predicted probability of pneumonia; Results:”univariate and then multivariate logistic regression ” and Lasso regression together show that age, natural log-transformed value (InCRP), Monocytes percentage (%Mon) are valid predictors of pneumonia risk; the AUC of the pneumonia risk model was 0.7820 (95% CI: 0.7254-0.8439) in the training cohort and 0.8432 (95% CI: 0.7588-0.9151) in the validation cohort; at the cut-off value of 0.5, the sensitivity and specificity of the pneumonia risk model were 70.75%, 66.33% (training cohort), 76.09%, and 73.91% (validation cohort), the calibration curves showed that the pneumonia risk model has good calibration accuracy. The decision curve analysis showed that the pneumonia risk model has high clinical value in predicting the probability of pneumonia in SARS-CoV-2 infected patients.Conclusion: The pneumonia risk prediction model developed in this study can be used to predict the risk of pneumonia in SARS-CoV-2 infected patients diagnostically.
【Key words】 SARS-CoV-2;COVID-19;prediction model;nomogram;pneumonia
INTRODUCTION
SARS-CoV-2 infection has spread globally since 2020, leading many countries to impose recurring quarantines, significantly impacting public health and the global economy1-2. Globally, as of 10 February 2023, there have been 755,385,709 confirmed cases of COVID-19, including 6,833,388 deaths, reported to WHO. Omicron, the mutant strain, entered the community in November 2021 and is far more contagious and escape-resistant than the previous variants of concern (VOC), like Delta3-8. At the beginning of 2022, the Omicron version quickly surpasses the Delta variant as the prevalent strain worldwide9.
During the early period of the COVID-19 pandemic, SARS-CoV-2 primarily affected the lung and caused pneumonia10-13. As one of the most representative and accurate diagnostic methods for COVID-1914, chest computed tomography (CT) scans are widely used in mainland China.
However, recent studies have demonstrated that the most recent VOC Omicron variant is much less likely to cause pulmonary infections3-5,15-16, suggesting potential implications for adapting management strategies for these infections.
In clinical practice, we found that due to the apprehension of contracting severe pneumonia from the SARS-CoV-2, many people with mild symptoms are choosing to receive CT scans, causing excessive CT scans and putting a strain on the availability of healthcare resources, which is particularly true when SRAS-CoV-2 localized epidemic outbreaks occur. Consequently, a strategy to evaluate the risk of pneumonia among recently infected people is essential to ensure the efficient use of healthcare resources and decrease unnecessary exposure to electromagnetic radiation.
With the aim of improving the classification of the risk of pneumonia in individuals with the most recent VOC of SARS-CoV-2 infections, reducing the overuse of CT scans, reducing non-essential ionizing radiation in individuals, as well as reducing the associated financial burden on patients, and optimizing the allocation of healthcare resources, we have developed and externally validated a pneumonia risk prediction model based on general patient data and blood routine test, to meet the needs of the new phase of the COVID-19 epidemic control.
Material and Methods
Materials
A retrospective analysis was performed on the clinical data of SARS-CoV-2-positive patients who visited outpatient and emergency clinics and underwent chest CT scans at the Mawangdui Branch of Hunan Provincial People’s Hospital from 20 December 2022 to 23 December 2022 and at the Tianxinge Branch of Hunan Provincial People’s Hospital from 1 January 2023 to 4 January 2023.Inclusion criteria: (1) Attendance as an outpatient or emergency (not including inpatients); (2) Patients had completed chest CT scans, and CT image quality meets diagnostic requirements; (3) SARS-CoV-2 infection positive was diagnosed by antigen test or nucleic acid test within 3 days before the current chest CT; (4) Complete blood routine examination results. Exclusion criteria: (1) Inflammation of a body part other than the lungs has been diagnosed at the time of the current blood routine tests;(2) the Patient was already on antiviral medication at the time of the visit. The patient recruitment pathway is detailed in FIGURE 1. The study complies with the Declaration of Helsinki. It was approved by the Medical Ethics Committee of Hunan Provincial People’s Hospital (The First Affiliated Hospital of Hunan Normal University), exempting the subjects from informed consent.
Methods
Device parameters and image analysis
The Mawangdui Branch (training cohort) used CT scans with a field of view (FOV) of 230 mm × 230 mm, a layer thickness of 5 mm, and layer spacing of 5 mm using the United Imaging uCT 760GE 128-slice CT; the Tianxinge Branch (validation cohort) used CT scans with a field of view (FOV) of 230 mm × 230 mm, a layer thickness of 5 mm, and layer spacing of 5 mm using the United Imaging uCT 860 160-slice CT or United Imaging uCT 960+ 640-slice CT. Two attending radiologists performed image analysis separately, and the final decision in case of a dispute was determined by consultation between the two physicians. CT Diagnosis of COVID-19 was referred to the report published by the RSNA17; typical findings were as follows: peripheral distribution, ground-glass opacity, fine reticular opacity, vascular thickening, and reverse halo sign. Patients with pneumonia were also classified into grades 0, 1, 2, 3, and 4 according to the extent and distribution of lung involvement (no lung involvement was categorized as grade 0).
Statistical analysis & construction and evaluation of predictive models
Statistical analysis was performed using Empower Stats, version 5.0 (http://www.empowerstats.com, X&Y Solutions, Inc., Boston, MA, USA), R statistical software, version 4.2.0 (http://www.R-project.org, The R Foundation), and the SPSS statistical software, version 27.0 (SPSS Inc., Chicago, IL, USA) with continuity variables expressed as medians (min, max) and categorical variables expressed as frequencies (percentages). Kruskal Wallis rank sum test or Fisher exact probability test was used to compare differences between groups of continuity variables. Chi-square tests are used for comparisons of categorical variables. After the natural log transformation of some continuity variables, to reduce irrelevant and redundant information, the predictor variables of the training cohort are filtered by both ” univariate and then multivariate logistic regression ” and ” least absolute shrinkage and selection operator (LASSO)” methods, the variables selected by both screening methods were used as the final predictor variables. The prediction model was constructed using multivariate logistic regression and presented in a nomogram. The ROC curves were used, and 500 in eternal resamples were performed by Bootstrap to evaluate the discrimination of the pneumonia risk model between the training and validation cohorts. Delong test and Integrated Discrimination Improvement Index (IDI) were used to compare the AUC of the pneumonia risk model with the AUCs for predictors incorporated in the model alone. Calibration curves were plotted to assess the calibration of the model. The clinical validity of the model was evaluated by the net benefit of DCA at different threshold probabilities; in addition, a smoothed curve was fitted using a generalized additive model (GAM) to explore the relationship between the pneumonia grade and the model’s predicted probability of pneumonia; a difference of p < 0.05 was considered statistically significant.
Results
General information.
A total of 205 cases were enrolled in the training cohort, with a median age of 47 years, the youngest being 14 years and the oldest 97 years, of which 105 cases (51.22%) were female and 100 cases (48.78%) were male, 99 cases (48.29%) without pneumonia and 106 cases (51.71%) with pneumonia; a total of 94 cases were enrolled in the validation cohort, with a median age of 56 years, the youngest being 2 years and the oldest 89 years, of which 60 (63.83%) were female and 34 (36.17%) were male, 47 (50.00%) were without pneumonia, and 47 (50.00%) were with pneumonia; the distribution of the remaining baseline indicators is shown in TABLE 1.