Chest CT Severity Score, CURB-65 Score and Their Relationship with in-Hospital-Mortality in COVID-19 Patients

Research Article

Austin J Radiol. 2021; 8(12): 1176.

Chest CT Severity Score, CURB-65 Score and Their Relationship with in-Hospital-Mortality in COVID-19 Patients

Azadbakht J¹* and Lajevardi ZS²

¹Department of Radiology, Faculty of Medicine, Kashan University of Medical Sciences, Kashan, Iran

²Faculty of Medicine, Kashan University of Medical Sciences, Kashan, Iran

*Corresponding author: Javid Azadbakht, Department of Radiology, Faculty of Medicine, Kashan University of Medical Sciences, Shahid Beheshti Hospital, Qotb Ravandi Blvd, 8715981151, Kashan, Iran

Received: November 01, 2021; Accepted: November 26, 2021; Published: December 03, 2021

Abstract

Objective: With every new strain of the SARS-CoV-2 spreading on a fast pace across the borders, an easy-to-calculate and reliable scoring system seems invaluable to identify high-risk patients. This study aims to investigate the relationship between CT Severity Score (CTSS) and CURB-65 score with mortality in COVID-19 patients.

Methods: This study was conducted on RT-PCR confirmed COVID-19 patients admitted to a tertiary teaching center during fifth national wave of disease in one of the early disease epicenters in the country. All enrolled patients underwent chest CT scan within first day of admission. CTSS and CURB-65 scores were calculated and assigned to patients, while radiologist was blinded to clinical and laboratory findings, and they were evaluated for their correlation with in-hospital mortality, additively and separately.

Results: Total number of 216 patients (140 males) with a mean age of 56.02 ± 17.34 years (ranging from 4 to 95) were enrolled. We found no significant relationship between CURB-65 score and CTSS (correlation coefficient: 0.065; P: 0.338). CURB-65 scores above 1 was predictive of in-hospital mortality with sensitivity of 56.4% and specificity of 81.9% (P: 0), those for CTSS above 11 were 79.5% and 4 51.5%, respectively (P: 0.001). CURB-65 score >1 and CTSS >11 predicted in-hospital mortality with sensitivity and specificity of 61.5% and 79.7% (P: 0.000). CURB-65 score and CTSS had a higher sensitivity and specificity to predict mortality comparing to each of those separately, but these enhanced statistics were not significant.

Conclusion: CURB-65 score is meaningfully stronger than CTSS to prognosticate in-hospital mortality in patients with COVID-19, and it is not significantly correlated with CTSS.

Keywords: Computed tomography; COVID-19; CT Severity Score; CURB-65; SARS CoV-2.

Introduction

With the global outbreak of COVID-19 and the rapid spread of the new variants of concern, all societies around the world are facing serious problems. On January 30, 2020, the World Health Organization declared the disease a global health emergency [1]. The disease infects the respiratory epithelial cells by targeting the human respiratory system, especially the lower airways [2]. COVID-19 can manifest with symptoms of the upper respiratory system such as coryza, sneezing and sore throat, despite the fact that it mainly involves the lower respiratory tract [3,4].

Covid-19 has no definitive cure to the moment, and this has led to the high prevalence and mortality of this disease which has put a lot of pressure on the world’s health care systems [5], specially countries with lower public vaccination coverage.

Reverse Transcription Polymerase Chain Reaction (RT-PCR) assays are widely used to confirm the infection as the standard diagnostic tool for COVID19, but due to the high rate of false positive results and its unavailability in the early stages of the outbreak, radiological examinations, especially chest CT scans, have played a more effective and practical role in early diagnosis and triage, as most pivotal steps to combat the infection. Chest CT can detect early lung infection, assess the severity of the disease and the extent of the chest involvement, and accordingly help in early triage and resource allocation/patient’s stratification [6-9].

In addition, the limitations of facilities such as diagnostic kits and the insufficient capacity of intensive care units double the importance of early identification of cases of COVID-19 who are prone to deterioration of general condition in the course of hospitalization. CT scan of the chest is highly sensitive to diagnose COVID-19 and more importantly it is available and fast in this era of resources shortage [10-14]. But CT scan alone cannot be used to rule out or rule in COVID-19 definitely [15].

CURB- 65 score determines the severity of pneumonia, and consists of five variables, each scored zero or 1 (with total score of 0-5), and is widely used to predict the 30-day mortality rate from community-acquired lung infections [16].

In an article by Gietema et al., It was found that by adding the CURB-65 score to the CTSS, the accuracy of CT scan in effectively diagnosing or rejecting pneumonia in patients clinically suspected of COVID-19 increases; as CURB-65 score greater than or equal to 3 in conjunction with a suggestive CT scan provides 100% sensitivity for COVID-19 detection [15].

The aim of this study was to determine the relationship between the CTSS and the CURB-65 score in COVID-19 patients and their individual and additive power to predict in-hospital mortality.

Material and Methods

This is an observational study and the data that support the findings of this investigation were collected retrospectively.

Study population

The hard copy and electronic records of all 216 participants with RT-PCR confirmed COVID-19 referred and admitted to our tertiary teaching center from April 2020 and September (fifth wave of outbreak in the country), who underwent on admission chest CT scan, were reviewed. Relevant positive and pertinent negative findings from history, physical examination, and laboratory data of studied patients were collected and recorded by the physician at the time of admission and all participants underwent Chest CT scan within the first 24 hours of admission. Missing pertinent data, including clinical symptoms, underlying diseases, etc., were obtained through telephone contact with patients. Patients with other lung diseases with possibility of presenting with similar manifestations on chest CT scan and potential of disturbing the CT severity scoring system (such as patients suspected for pulmonary edema [according to lesion distribution and opacities with dramatic response to diuretics] or lung contusion/alveolar hemorrhage [suggestive history]), patients with blood culture positive for either community-acquired or nosocomial pneumonia, and patients with artifactual chest CT scans were excluded from our study. In general, 216 subjects (140 men) remained eligible to go under investigation.

Chest CT protocols

All images were obtained on a same CT scanner (Toshiba, Canon, Alexia, Japan, 16-detector) and images were reconstructed in axial plane, with slice thickness of 3 mm, mAS of 100 and kvp of 120-100, while patient in supine position with raised hands. Images were taken at full inspiration (as tolerated by patient), reconstructed with sharp kernel, and reviewed in both mediastinal (WW: 400 HU, WL: 40 HU) and lung windows (WW: 1500 HU, WL: -500 HU).

Chest CT images interpretation

A radiologist confident and experienced in thoracic imaging (with 5 years of experience) interpreted the CT scan images adhering to a systematic approach, and findings were compared to previous reports. A CTSS was assigned to each participant, while radiologist was completely unaware of the clinical and laboratory findings.

Statistical analysis

Raw data was analyzed via SPSS software version 22, using both descriptive (frequency distribution and central indices and dispersion) and inferential statistics (t-test for comparing the mean of quantitative variables, and Chi-square test to assess the correlation between categorical variables). Significant predictors were then identified using the univariate model. In the next step, the multivariate conditional logistic regression model was used to design a model indicating the relationship between considered variables and patients mortality rate. Only variables with a p-value of less than 0.25 were included in the model. The results of the Omnibus test are acceptable model fit and significant at an error level of less than 0.001. After determining the significant predictor(s), the sensitivity and specificity (accuracy) of predicting mortality was measured for the CT-ss alone and with other model predictors through analyzing ROC curves.

Results

In this study, the hard copy and electronic records of 216 patients with rRT-PCR confirmed COVID-19 (140 men) with a mean age of 56.02 ± 17.34 years (ranging from 4 to 95 years) were reviewed. Our results showed that the median onset of symptoms and perform RTPCR was 5 days [3-7].

Diabetes mellitus (38%), hypertension (28.7%) and cardiovascular diseases (21.3%) were the most common underlying diseases among the subjects. In general, 12.5% of patients had a history of smoking cigarette or hookah. The most common blood group in the subjects was blood type O (42.3%).