Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Chest CT Scan Features to Predict COVID-19 Patients’ Outcome and Survival

Chest CT Scan Features to Predict COVID-19 Patients’ Outcome and Survival Hindawi Radiology Research and Practice Volume 2022, Article ID 4732988, 9 pages https://doi.org/10.1155/2022/4732988 Research Article Chest CT Scan Features to Predict COVID-19 Patients’ Outcome and Survival 1 1 1 Mohammad-Mehdi Mehrabi Nejad , Aminreza Abkhoo , Faeze Salahshour , 2 1 3 Mohammadreza Salehi , Masoumeh Gity , Hamidreza Komaki , and Shahriar Kolahi Department of Radiology, School of Medicine, Advanced Diagnostic and Interventional Radiology Research Center (ADIR), Imam Khomeini Hospital, Tehran University of Medical Sciences, Tehran, Iran Department of Infectious Diseases and Tropical Medicines, Tehran University of Medical Sciences, Tehran, Iran Brain Engineering Research Center, Institute for Research in Fundamental Sciences (IPM), Tehran, Iran Correspondence should be addressed to Faeze Salahshour; f-salahshour@sina.tums.ac.ir Received 11 April 2021; Revised 26 August 2021; Accepted 28 January 2022; Published 26 February 2022 Academic Editor: Sreekanth Kumar Mallineni Copyright © 2022 Mohammad-Mehdi Mehrabi Nejad et al. -is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background. Providing efficient care for infectious coronavirus disease 2019 (COVID-19) patients requires an accurate and accessible tool to medically optimize medical resource allocation to high-risk patients. Purpose. To assess the predictive value of on-admission chest CTcharacteristics to estimate COVID-19 patients’ outcome and survival time. Materials and Methods. Using a case-control design, we included all laboratory-confirmed COVID-19 patients who were deceased, from June to September 2020, in a tertiary-referral-collegiate hospital and had on-admission chest CT as the case group. -e patients who did not die and were equivalent in terms of demographics and other clinical features to cases were considered as the control (survivors) group. -e equivalency evaluation was performed by a fellowship-trained radiologist and an expert radiologist. Pulmonary involvement (PI) was scored (0–25) using a semiquantitative scoring tool. -e PI density index was calculated by dividing the total PI score by the number of involved lung lobes. All imaging parameters were compared between case and control group members. Survival time was recorded for the case group. All demographic, clinical, and imaging variables were included in the survival analyses. Results. After evaluating 384 cases, a total of 186 patients (93 in each group) were admitted to the studied setting, consisting of 126 (67.7%) male patients with a mean age of 60.4± 13.6 years. -e PI score and PI density index in the case vs. the control group were on average 8.9± 4.5 vs. 10.7± 4.4 (p value: 0.001) and 2.0± 0.7 vs. 2.6± 0.8 (p value: 0.01), respectively. Axial distribution (p value: 0.01), cardiomegaly (p value: 0.005), pleural effusion (p value: 0.001), and pericardial effusion (p value: 0.04) were mostly observed in deceased patients. Our survival analyses demonstrated that PI score≥ 10 (p value: 0.02) and PI density index≥ 2.2 (p value: 0.03) were significantly associated with a lower survival rate. Conclusion. On-admission chest CT features, particularly PI score and PI density index, are potential great tools to predict the patient’s clinical outcome. respiratory distress syndrome (ARDS) [2]. -e global death- 1. Introduction to-case ratio is estimated to be 3.5% [3]. However, it varies Coronavirus disease 2019 (COVID-19), caused by severe geographically probably due to the local preventive measures acute respiratory syndrome coronavirus-2 (SARS-CoV-2), and medical resources. -erefore, we need to improve the was officially announced as a pandemic by the World Health admitted patients’ initial triage not only to optimize the Organization on March 11, 2020 [1]. Even though most of allocation of the medical resources to high-risk patients and the patients experience mild symptoms, some may develop a minimize the mortality rate accordingly but also to more severe type of disease and it may progress to acute accurately and reliably predict the outcome of the patients. 2 Radiology Research and Practice Chest computed tomography (CT) scan, as the con- characteristics, age and sex; (b) on-admission vital signs, ventional and relatively accessible imaging modality for temperature (T-Celsius), oxygen saturation (SpO ), heart pneumonia diagnosis and follow-up, is confirmed to have rate (HR-per minute), respiratory rate (RR-per minute), and high diagnostic and prognostic values in the current out- blood pressure (BP-mmHg); (c) survival time, days from break of COVID-19 [4, 5]. Chest CT findings mainly consist admission to death in nonsurvived patients; (d) underlying of ground-glass opacities (GGOs), multifocal patchy con- diseases, hypertension (HTN), diabetes (DM), respiratory solidation, and interstitial changes with a peripheral dis- disease (asthma, COPD, ILD, or bronchiectasis), malig- tribution [6–8]. Efforts have been made to underpin the nancies (solid or hematological malignancies), immuno- predictive factors for mortality [9]. Among all, the main compromised conditions (chemoradiation therapy and factors consist of age, underlying disease (i.e., immuno- long-term corticosteroid usage), and hypothyroidism; and compromised patients and preexisting cardiovascular and (f) laboratory findings, white blood cell including neutrophil pulmonary disorders), laboratory findings (i.e., D-dimer and lymphocyte counts, hemoglobin, platelet, creatinine, level, neutrophil-to-lymphocyte ratio (NLR), urea, international normalized ratio (INR), partial throm- and lymphocyte count), and most recently, imaging features boplastin time (PTT), D-dimer, lactate dehydrogenase [10–13]. However, there is still no consensus on the factor (LDH),C-reactive protein (CRP), and pro-b-type natriuretic with the highest predictive value. Besides, almost all previous peptide (Pro-BNP). studies were retrospective cross-sectional and nonsurvivors comprised only a small population [14, 15], which further 2.2.2. Image Acquisition. All chest CT images were acquired limit the application of the findings. In this regard, we aimed to utilize the vastly used mo- at the time of admission, in the supine position, with full inspiration with no contrast injection. Examinations were dality-chest CT scan, and we hypothesize that on-admission chest CT findings could serve as a potential deterministic performed on either the Siemens Somatom Emotion (16 slices, Erlangen, Germany) or the Lightspeed 64-detector CT factor for risk stratification in hospitalized COVID-19 pa- tients. In order to evaluate the proposed notion, we con- (GE Healthcare, Milwaukee, USA) MDCT scanner. -e imaging parameters were set at 5–6 mm section thickness, ducted a case-control study and assessed the adjusted beam collimation of 0.6–2 mm, 120 kVp tube voltage, tube predictive value of CT scan in terms of mortality rate for admitted COVID-19 patients. current of 150–250 mAs, tube rotation speed of 0.75 seconds, and gantry rotation time of 0.5–0.75 s, reconstructed with a mediastinum B20f smooth kernel and a lung B70f sharp 2. Materials and Methods kernel (Siemens Healthineers, Erlangen, Germany); coronal 2.1. Study Design and Participants. -is case-control study and sagittal multiplanar reconstructions were also available was reviewed and approved by the Institutional Review with a reconstructed slice thickness of 1.2 mm. Board of our university. Given the retrospective design of the study and anonymous use of medical records, informed consent requirement was waived by the ethics committee of 2.2.3. Image Interpretation. Two fellowship-trained diag- our institute (IR.TUMS.VCR.REC.1399.054). -e current nostic imaging radiologists, with respective 9 and 13 years of case-control study was carried out in a referral tertiary experience in thoracic radiology and blinded to patients’ university hospital from June to September 2020. -is study outcomes, independently interpreted chest CT images’ evaluated individuals with the following conditions: (a) all findings. All CT images were reviewed on both lung- and hospitalized COVID-19 patients in whom COVID-19 di- mediastinal-window settings. -e intraclass correlation agnosis was confirmed by positive real-time reverse tran- coefficient (ICC) was calculated to assess inter-rater reli- scription-polymerase chain reaction (rRT-PCR) assay on ability. If ICC was less than 0.8, any disagreement in image nasopharyngeal or oropharyngeal swap or endotracheal interpretation for the case was discussed until resolved. If aspirate samples; (b) age equal or greater than 18 years; and ICC was greater than or equal to 0.8, the value reported by (c) a definite outcome of either death or hospital discharge. the radiologist with higher experience was recorded. Case patients consisted of patients who met the inclusion Chest CT scan findings were recorded according to the criteria and were deceased. Control patients included dis- Fleischner Society glossary and published literature on viral charged patients with equivalent demographic (age and sex) pneumonia [16]. Chest CT scan features included the fol- and clinical (underlying diseases and laboratory findings) lowing: (a) predominant pattern, ground-glass opacification/ features. -e equivalency evaluation was performed by a opacity (GGO) (Figure 2), consolidation (Figure 3), and fellowship-trained radiologist and an expert epidemiologist. mixed; (b) dominant distribution pattern, peripheral (pe- -e flow diagram of the study is presented in Figure 1. Of ripheral one-third of the lung) (Figure 3), axial (medial two- note, admission, discharge criteria, and treatment of all thirds of the lung), and diffuse (Figure 2); (c) number of patients were based on the national protocol of COVID-19. involved lobes; (d) other morphologies, parenchymal band, crazy paving, reverse halo sign, or intralesional traction bronchiectasis; and (e) additional findings, cardiomegaly 2.2. Data Collection (Figure 2), pleural effusion (unilateral or bilateral) (Fig- ure 2), pericardial effusion, emphysema, dilated pulmonary 2.2.1. Population Characteristics. -e recorded attributes of trunk, pleural thickening, and mosaic attenuation. patients consisted of the following: (a) demographic Radiology Research and Practice 3 Admitted patients with COVID-19 between June and September 2020 = 384 < 18 years old = 13 Patients with on-admission CT scan = 323 Incomplete medical record = 43 Patients with definite outcome = 247 No Matching = 61 Case group Control group (deceased) (discharged) N = 93 N = 93 Figure 1: Flow diagram of the study. Figure 2: A 76-year-old male patient with diabetes deceased due to COVID-19. Chest CT scan ((a) coronal view, (b) axial view-lung window, and (c) axial view-mediastinal window) showed predominancy of ground-glass opacity (GGO) with the bilateral and diffuse distribution. Pulmonary involvement (PI) and PI density scores were 16 and 3.2, respectively. Additional findings included cardiomegaly and bilateral pleural effusion. He was considered a high-risk patient. 2.2.4. Pulmonary Involvement (PI) Scoring System. To assess lobes’ scores. -e PI score ranged from 0 (no involvement) PI, a semiquantitative scoring tool was proposed and used to 25 (maximum involvement). Finally, the PI density index [17]. All five lung lobes (right upper lobe (RUL), right middle was calculated by dividing the total PI score by the number of involved lobes. lobe (RML), right lower lobe (RLL), left upper lobe (LUL), and left lower lobe (LLL)) were visually reviewed for GGO and consolidation. -en, a score from 0 to 5 was assigned to each lobe according to involvement percentage (0: no in- 2.3. Statistical Analysis. We performed the analyses in SPSS volvement; 1: ≤5%; 2: 6–25%; 3: 26–50%, 4: 51–75%; and 5: for Windows ver. 18 (Chicago, IL, USA). Descriptive data are ≥76%). -e total PI score was calculated as the sum of all five presented as mean with standard deviations (SD) for 4 Radiology Research and Practice Figure 3: A 61-year-old female patient with hypertension and diabetes. Pulmonary involvement: predominancy of GGO with peripheral, pleural-based distribution. Total pulmonary involvement (PI) score and PI density index were 6 and 1.2, respectively, and she was stratified as a low-risk patient in death predictive models. continuous variables and as frequency and percentage of the 3. Results population for categorical variables. In order to evaluate 3.1. Study Population Characteristics. After evaluating 384 whether the recorded data have a normal distribution, we cases, a total of 186 patients (93 in each group) were ad- used the Kolmogorov–Smirnov test. We conducted the mitted to the studied setting, consisting of 126 (67.7%) male comparisons by (a) the independent two-tailed sample t-test patients with a mean age of 60.4± 13.6 years (Figure 1). -e for continuous variables with normal distribution and the most common underlying diseases were HTN (38.7%) and relevant degree of freedom; (b) the Mann–Whitney U test DM (34.9%) (Table 1). for the continuous variables with significant lack of nor- Two groups were almost equivalent for demographic mality variables; and (c) the Chi-square test for nominal (age and sex) and clinical (underlying diseases and labo- variables. All p values less than 0.05 were considered sta- ratory findings) variables. Table 1 further illustrates the tistically significant. homogeneity of cases and controls in the aforementioned We used multivariate logistic regression to predict characteristics. All ICCs for inter-rater reliability were >0.8 mortality probability with all imaging parameters as the for all imaging parameters. independent variables. As previously stated, the statistically significant threshold was considered as a p value less than 0.05. To define optimum cutoff values for PI score and PI density index in outcome prediction, receiver-operating 3.2. Chest CT Scan Findings. -e most common CT features characteristic (ROC) curves were drawn and Youden’s J among survivors and nonsurvivors were GGO (65.6% and index [18] was calculated. -e area under the ROC curve 68.8%), multilobar (95.7% and 98.9%), bilateral lobe in- (AUC) was considered as the indicator for ROC analysis volvement (93.5% and 94.6%), and lower lobe (RLL and/or efficacy. LLL) involvement (94.6% and 98.9%), respectively. Pre- To determine the impact of any independent variable on dominant distribution patterns among the two groups were survival time, we implemented univariate Cox regressions peripheral (47.3% in survivors) and axial (57.0% in (in nonsurvivors) considering all demographic, clinical, and nonsurvivors). imaging parameters as covariates in the model. Multivariate -e mean PI score and PI density index in survivors vs. backward Cox regression was performed to find the final nonsurvivors were 8.9± 4.5 vs. 10.7± 4.4 (p value: 0.001) and model in terms of which variables to include. Kaplan–Meier 2.0± 0.7 vs. 2.6± 0.8 (p value: 0.01), respectively. In addition, survival analysis was performed to calculate survival, and the nonsurvived patients had a higher involvement score for all log-rank test was used to compare the survival distribution single lung lobes (except RLL) and axial distribution (37.6% of two subgroups of interest. vs. 57.0%, p value: 0.01). However, the number of involved Radiology Research and Practice 5 Table 1: Details of demographic and clinical data of patients and differences between two groups. Variables All patients, N � 186 Survivors, N � 93 Nonsurvivors, N � 93 P value Demographic data Age 60.4 (13.6) 58.7 (13.9) 62.2 (13.1) 0.08 Gender Male 126 (67.7) 61 (65.6) 65 (69.9) 0.53 Female 60 (32.3) 32 (34.4) 28 (30.1) Clinical data Vital signs RR 15.1 (9.1) 25.3 (5.1) 24.9 (11.8) 0.79 RR>24 102 (54.8) 51 (54.8) 51 (54.8) 1 Systolic BP 122.2 (20.1) 124.3 (18.0) 120.1 (21.8) 0.15 Diastolic BP 75.2 (12.7) 78.0 (10.9) 72.5 (13.7) 0.03 PR 96.9 (18.8) 97.4 (18.6) 96.4 (19.2) 0.70 Temperature 37.5 (0.8) 37.4 (0.8) 37.5 (0.9) 0.19 Hospitalization duration Total admission days 10.6 (8.9) 9.6 (8.0) 11.6 (9.7) 0.11 ICU days 5.2 (8.1) 3.0 (5.8) 7.4 (9.5) <0.001 Underlying disease HTN 72 (38.7) 32 (34.4) 40 (43.0) 0.29 DM 65 (34.9) 36 (38.7) 29 (31.2) 0.35 Respiratory disease 14 (7.5) 9 (9.7) 5 (5.4) 0.40 Immunocompromised 18 (9.7) 7 (7.5) 11 (11.8) 0.45 Hypothyroidism 13 (7.0) 9 (9.7) 4 (4.3) 0.24 Laboratory findings WBC 8.9 (4.7) 8.3 (4.3) 9.3 (5.0) 0.28 Neutrophil 7.0 (3.9) 6.5 (3.9) 7.4 (3.8) 0.24 Lymphocyte 1.4 (3.9) 1.2 (0.7) 1.5 (2.8) 0.52 Hemoglobin 12.5 (2.7) 12.8 (2.4) 12.3 (2.9) 0.40 Platelet 209.3 (95.2) 219.2 (110.9) 202.2 (82.4) 0.39 Cr 1.7 (1.6) 1.7 (1.3) 1.7 (1.8) 0.99 Urea 58.3 (60.8) 49.8 (36.1) 64.5 (73.4) 0.24 INR 1.3 (0.8) 1.4 (1.1) 1.2 (0.5) 0.51 PTT 41.4 (22.2) 43.4 (25.1) 40.0 (19.8) 0.47 D-dimer 3364.1 (3210.0) 3450.7 (3344.9) 3147.7 (3316.3) 0.88 LDH 661.9 (287.3) 676.6 (269.7) 649.1 (305.7) 0.72 CRP 123.8 (73.0) 113.3 (72.1) 130.6 (73.4) 0.24 Pro-BNP 5329.9 (10289.0) 4505.8 (10687.2) 6085.2 (10324.3) 0.72 Mean (standard deviation); all other variables reported as N (%). RR � respiratory rate; BP � blood pressure; PR � pulse rate; HTN � hypertension; DM � diabetes; ICU � intensive care unit; WBC � white blood cell; Cr � creatinine; INR � international normalized ratio; PTT �partial thromboplastin time; LDH � lactate dehydrogenase; CR � C-reactive protein; Pro-BNP � pro-b-type natriuretic peptide. Table 2: Radiologic findings in all patients and differences between the two groups. Variables All patients, N � 186 Survivors, N � 93 Nonsurvivors, N � 93 P value PI scores RUL total score 2.0 (1.1) 1.8 (1.0) 2.2 (1.1) 0.02 RML total score 1.5 (1.0) 1.3 (0.9) 1.6 (1.0) 0.03 RLL total score 2.3 (1.1) 2.1 (1.1) 2.4 (1.1) 0.15 LUL total score 2.0 (1.1) 1.7 (1.1) 2.2 (1.1) 0.02 LLL total score 2.2 (1.2) 2.0 (1.2) 2.3 (1.2) 0.03 Total lung GGO score 6.7 (4.5) 6.2 (4.2) 7.2 (4.8) 0.13 Total lung consolidation score 3.1 (3.2) 2.6 (3.0) 3.6 (3.5) 0.05 Total PI score 9.8 (4.5) 8.9 (4.5) 10.7 (4.4) 0.001 PI density index 2.1 (0.8) 2.0 (0.7) 2.6 (0.8) 0.01 Predominant pattern GGO 125 (67.2) 61 (65.6) 64 (68.8) 0.63 Consolidation 61 (32.8) 32 (34.4) 29 (31.2) Dominant distribution of lesions 6 Radiology Research and Practice Table 2: Continued. Variables All patients, N � 186 Survivors, N � 93 Nonsurvivors, N � 93 P value Peripheral 70 (37.6) 44 (47.3) 26 (28.0) Axial 88 (47.3) 35 (37.6) 53 (57.0) 0.01 Diffuse 28 (15.1) 14 (15.1) 14 (15.1) No. of involved lobes 4.5 (1.0) 4.5 (1.1) 4.6 (0.9) 0.49 Monolobar 1 5 (2.7) 4 (4.3) 1 (1.1) Multilobar 2 5 (2.7) 1 (1.1) 4 (4.3) 0.05 3 12 (6.5) 3 (3.2) 9 (9.7) 4 15 (8.1) 11 (11.8) 4 (4.3) 5 149 (80.1) 74 (79.6) 75 (80.6) Laterality Unilateral 11 (6.1) 6 (6.9) 5 (5.4) 0.76 Bilateral 175 (94.1) 87 (93.5) 88 (94.6) Lower lobe involvement Yes 180 (96.8) 88 (94.6) 92 (98.9) 0.09 No 6 (3.2) 5 (5.4) 1 (1.1) Additional findings Cardiomegaly 85 (45.7) 33 (35.5) 52 (55.9) 0.005 Mosaic attenuation 7 (3.7) 3 (3.2) 4 (4.3) 1 Pleural effusion 35 (18.8) 9 (9.7) 26 (28.0) 0.001 Unilateral 12 (6.5) 3 (3.2) 9 (9.7) 0.07 Bilateral 23 (12.4) 6 (6.5) 17 (18.3) 0.01 Pericardial effusion 13 (7.0) 3 (3.2) 10 (10.8) 0.04 Emphysema 7 (3.8) 3 (3.2) 4 (4.3) 1 Dilated pulmonary trunk 23 (12.4) 5 (5.4) 18 (19.4) 0.004 Pleural thickening 34 (18.3) 3 (3.2) 31 (33.3) <0.001 Subsegmental atelectasis 64 (34.4) 24 (25.8) 40 (43.0) 0.01 Other morphologies Parenchymal band 83 (44.6) 42 (45.2) 41 (44.1) 0.88 Crazy paving 69 (37.1) 29 (31.2) 40 (43.0) 0.09 Reverse halo 14 (7.5) 7 (7.5) 7 (7.5) 1 In-lesion bronchiectasis 11 (5.9) 7 (7.5) 4 (4.3) 0.35 Mean (standard deviation); all other variables reported as N (%). PI � pulmonary involvement; RUL � right upper lobe; RML � right middle lobe; RLL � right lower lobe; LUL � left upper lobe; LLL � left lower lobe; GGO � ground-glass opacity. lobes, predominant pattern, and bilateral and lower lobe (AUC: 0.61 [0.52–0.69], Youden’s J: 0.19, p value: 0.01) involvement did not show any significant difference in showed the best accuracy in predicting death (Figures 2–3). comparison (p values: 0.49, 0.63, 0.76, and 0.09) (Table 2). Of additional findings and other morphologies, car- diomegaly (35.5% vs. 55.9%, p value: 0.005) (Figure 2), 3.3. Survival Analysis. -e survival analyses were limited to nonsurvived patients. -e mean survival was 4.1 ± 3.5 days, and pleural effusion (9.7%% vs. 28.0%, p value: 0.001) (Figure 2), the median survival was four days. -e Kaplan–Meier survival pericardial effusion (3.2% vs. 10.8%, p value: 0.04), dilated function for death is illustrated in Figure 4. To determine the pulmonary trunk (5.4% vs. 19.4%, p value: 0.004), and impact of independent variables on survival, we implemented pleural thickening (3.2% vs. 33.3%, p value:<0.001) were significantly more prevalent in the nonsurvived group. univariate Cox regressions, considering demographic, clinical, and imaging variables as independent variables. Statistically Table 2 demonstrates the details of patients’ characteristics associated with death. significant variables included PI score≥ 10 (p value: 0.02) and PI density≥ 2.2(p value: 0.03). Afterward, we fitted a multi- We exploited backward multivariate logistic regressions that included death as the outcome of interest and all im- variate backward Cox regression on the variables with a sig- nificant univariate association: PI score≥ 10 (p value: 0.03) aging parameters with significant association in Table 2 as the independent variables. PI score (Ex(B): 1.10 (95% CI: remained significant at the final step (Table 3). Further analyses using the Kaplan–Meier survival 1.03–1.17), p value: 0.006)) and PI density index (Ex(B): 1.64 function and the log-rank test revealed that survival was (95% CI: 1.10–2.43), p value: 0.02)) independently remained as significant variables in the models. -e PI score with a significantly different in two subgroups according to PI score≥ 10 (χ cutoff value of 10 (AUC: 0.62 [0.54–0.70], Youden’s J: 0.23, p � 7.05, p value: 0.008) and PI density≥ 2.2 value: 0.005) and PI density index with a cutoff value of 2.2 (χ � 6.58, p value: 0.01) (Figure 4). Radiology Research and Practice 7 1.0 1.0 1.0 0.8 0.8 0.8 0.6 0.6 0.6 0.4 0.4 0.4 0.2 0.2 0.2 0.0 0.0 0.0 0.00 2.00 4.00 6.00 8.00 10.0012.00 14.00 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 survival days survival days survival days PI PI score density <10 <2.2 ≥10 ≥2.2 (a) (b) (c) Figure 4: Kaplan–Meier survival curve. Estimated survival rate in (a) all deceased patients and comparisons based on (b) PI score and (c) PI density score. Table 3: Univariate and multivariate Cox regression. Variable Coefficient of variable in the model Exp(B) (95% CI) p value of variables p value of models Univariate Cox regression PI score≥ 10 0.60 1.8 [1.09–3.01] 0.02 0.02 PI density score≥ 2.2 0.57 1.77 [1.10–2.91] 0.03 0.03 Multivariate backward Cox regression PI score≥ 10 0.56 1.75 [1.05–2.9] 0.03 0.01 PI: pulmonary involvement; CI: confidence interval. used a 20-scale scoring system in 50 patients (23 deceased, 4. Discussion 27 survived) and suggested a cutoff value of 12 with 0.79 -ere is still no consensus on the clinical and imaging factors AUC in predicting mortality [21]. Moreover, a study used a that are associated with and affect COVID-19 patients’ 72-scale CT score and reported 85% sensitivity and speci- outcomes and survival. Our findings contribute to the ficity in predicting the mortality of patients with a cutoff existing literature by illustrating that patients with higher PI value of 24.5 [14]. In general, all previous studies were scores, PI density index, axial distribution, cardiomegaly, or retrospective and nonsurvivors mostly comprised a small pleural or pericardial effusion are more likely to decease. -e proportion of the studied population. Hence, differences in data PI density score is our novel suggested index to distinguish sampling and study design (case-control vs. cross-sectional), as between patients with the same PI score but a different well as group matching, in our study remarkably improved number of involved lobes because it showed a stronger the validity of results. Furthermore, details of chest CT mortality prediction power. findings were not fully reported in previous studies, and Previously reported chest CT features to predict there is no consistency on the value of additional findings, COVID-19 patients’ mortality mainly included crude lung including crazy paving and pleural effusion in predicting the involvement score, number of involved lobes, bilateral or outcome [19, 21, 25], which are both depicted in this study. lower lobe involvement, and diffuse pattern [14, 19–25]. In Further analyses on patients’ survival demonstrated that line with our findings, yet in contrast to previous reports, PI score <10 and PI density index <2.2 were significantly Yuan et al. [14] reported no significant association of lower associated with higher survival. To the best of our knowl- lobe or bilateral lung involvement with death. Nevertheless, edge, imaging factors associated with the survival days of a significantly higher CT score was the most common re- COVID-19 patients were addressed only in one retrospec- ported feature to predict death [14, 19–23, 25]. Limited tive study on 20 deceased patients [23]. In line with our previous studies also reported cutoff values using totally findings, they found significantly lower survival days in different scoring systems [14, 21, 23]. For instance, Francone patients with higher (≥18) vs. lower (<18) CT scores over a et al. [23] retrospectively evaluated 130 COVID-19 patients 24-day follow-up period [23]. (20 patients deceased) and recommended the CT score≥ 18 Considering the limitations of the study, we still (out of 25) as the predictive factor for death. Another study performed the largest case-control investigation on Cum Survival Cum Survival Cum Survival 8 Radiology Research and Practice evaluating the clinical and imaging factors to predict Acknowledgments COVID-19 patients’ outcomes and survival. We provided -e authors are thankful to the patients and hospital staff for a novel semiquantitative scale with a defined optimum their collaboration. cutoff, which could serve to better identify high-risk pa- tients and recognize patients with a higher demand for critical care but similar clinical conditions compared to References other patients. We posit that patients with a PI score≥ 10 [1] WHO, Novel Coronavirus (2019-nCoV) Situation Report-51, and a PI density score≥ 2.2 should be carefully cared for https://www.who.int/docs/default-source/coronaviruse/ and receive aggressive treatment, as they are highly sus- situation-reports/20200311-sitrep-51-covid-19.pdf, WHO, ceptible to death. Hence, radiologists are expected to re- Geneva, Switzerland, 2020, https://www.who.int/docs/ port PI and PI density scores in their everyday practice to default-source/coronaviruse/situation-reports/20200311- help clinical physicians manage patients more effectively sitrep-51-covid-19.pdf. and efficiently and consequently minimize COVID-19- [2] N. Chen, M. Zhou, X. Dong et al., “Epidemiological and related mortality. Nevertheless, our findings must be clinical characteristics of 99 cases of 2019 novel coronavirus interpreted in light of some limitations. Firstly, the study pneumonia in Wuhan, China: a descriptive study,” 7e context is prone to a biased selection of cases and controls. Lancet, vol. 395, no. 10223, pp. 507–513, 2020. To elaborate, the setting is a tertiary and referral hospital [3] V. J. Munster, M. Koopmans, N. van Doremalen, D. van Riel, and E. de Wit, “A novel coronavirus emerging in China - key for COVID-19 patients, and the patients who get admitted questions for impact assessment,” New England Journal of are probably more clinically severe than the cases in the Medicine, vol. 382, no. 8, pp. 692–694, 2020. general population with similar paraclinical and demo- [4] Y. Fang, H. Zhang, and J. Xie, “Sensitivity of chest CT for graphic features. On the other hand, there are reports on COVID-19: comparison to RT-PCR,” 7e Radiologist, factors, such as immunological factors and predispositions vol. 296, 2020. in patients, that might have an impact on the survival rate [5] A. Jalali, E. Karimialavijeh, and P. Babaniamansour, “Pre- of the COVID-19 cases [26,27]. However, we did not have dicting the 30-day adverse outcomes of non-critical new- other available data to incorporate in the study. In addi- onset COVID-19 patients in emergency departments based tion, the admitted patients are not equivalent to outpatient on their lung CT scan findings; a pilot study for derivation an cases in the predicted mortality rate. For example, out- emergency scoring tool,” Frontiers in Emergency Medicine, patient cases might have an even lower mortality rate with vol. 5, p. e40, 2021. [6] E. Lanza, R. Muglia, I. Bolengo et al., “Quantitative chest CT a similar PI score because they are in a relatively better analysis in COVID-19 to predict the need for oxygenation clinical condition. Another limitation is that we did not support and intubation,” European Radiology, vol. 30, no. 12, follow the discharged patients to evaluate whether they pp. 6770–6778, 2020. were deceased. -e interval time between symptom onset [7] R. Han, L. Huang, and H. Jiang, “Early clinical and CT and admission was not exactly the same in all patients; manifestations of coronavirus disease 2019 (COVID-19) however, we used the on-admission CT scans to minimize pneumonia,” American Journal of Roentgenology, vol. 61, this bias. Although all patients were treated based on a pp. 1–6, 2020. unique national protocol, minor treatment differences [8] T. Ai, Z. Yang, and H. Hou, “Correlation of chest CT and RT- based on physician clinical judgment could affect mortality PCR testing in coronavirus disease 2019 (COVID-19) in and survival, which were not assessed in the present study. China: a report of 1014 cases,” 7e Radiologist, vol. 22, Article ID 200642, 2020. -is study was conducted before the national COVID-19 [9] F. Salahshour, M.-M. Mehrabinejad, and M. N. Toosi, vaccination program and further investigations on a larger “Clinical and chest CT features as a predictive tool for population after vaccination are recommended to confirm COVID-19 clinical progress: introducing a novel semi- our findings [28]. quantitative scoring system,” European Radiology, vol. 31, In conclusion, patients with a higher PI score, PI density pp. 1–11, 2021. index, axial distribution, cardiomegaly, or pleural or peri- [10] X. Zhao, B. Zhang, and P. Li, “Incidence, clinical charac- cardial effusion are more susceptible to poor prognosis. teristics and prognostic factor of patients with covid-19: Survival analyses revealed that a PI score≥ 10 and PI density asystematic review and meta-analysis,” medRxiv, vol. 20, 2020. score≥ 2.2 were significantly associated with lower survival [11] S. Babaniamansour, A. Atarodi, and P. Babaniamansour, and those patients should be prioritized for higher medical “Clinical findings and prognosis of COVID-19 patients with attention. Initial chest CT examination may be able to help benign prostatic hyperplasia: acase series,” Journal of Bio- medical Research&EnvironmentalSciences, vol. 2, 2021. in predicting the patient’s outcome and survival. [12] S. Goudarzi, F. Dehghani Firouzabadi, M. Dehghani Fir- ouzabadi, and N Rezaei, “Cutaneous lesions and COVID-19: Data Availability cystic painful lesion in a case with positive SARS-CoV-2,” Dermatologic 7erapy, vol. 33, no. 6, p. e14266, 2020. All the data used in this manuscript can be made available [13] H. Dehghanbanadaki, H. Aazami, M. Shabani et al., “A upon request to the corresponding author. systematic review and meta-analysis on the association be- tween lymphocyte subsets and the severity of COVID-19,” Immunopathologia Persa, 2022. Conflicts of Interest [14] M. Yuan, W. Yin, Z. Tao, W. Tan, and Y. Hu, “Association of -e authors declare that they have no conflicts of interest. radiologic findings with mortality of patients infected with Radiology Research and Practice 9 2019 novel coronavirus in Wuhan, China,” PLoS One, vol. 15, no. 3, Article ID e0230548, 2020. [15] S. Richardson, J. S. Hirsch, M. Narasimhan et al., “Presenting characteristics, comorbidities, and outcomes among 5700 patients hospitalized with COVID-19 in the New York City area,” JAMA, vol. 323, no. 20, pp. 2052–2059, 2020. [16] D. M. Hansell, A. A. Bankier, H. MacMahon, T. C. McLoud, N. L. Muller, ¨ and J. Remy, “Fleischner Society: glossary of terms for thoracic imaging,” Radiology, vol. 246, no. 3, pp. 697–722, 2008. [17] A. Abkhoo, E. Shaker, and M.-M. Mehrabinejad, “Factors predicting outcome in intensive care unit-admitted COVID- 19 patients: using clinical, laboratory, and radiologic char- acteristics,” Critical Care Research and Practice, vol. 2021, Article ID 9941570, 2021. [18] W. J. Youden, “Index for rating diagnostic tests,” Cancer, vol. 3, no. 1, pp. 32–35, 1950. [19] F. Pan, C. Zheng, T. Ye et al., “Different computed tomog- raphy patterns of Coronavirus Disease 2019 (COVID-19) between survivors and non-survivors,” Scientific Reports, vol. 10, pp. 11336–11339, 2020. [20] Y. Li, Z. Yang, and T. Ai, “Association of “initial CT” findings with mortality in older patients with coronavirus disease 2019 (COVID-19),” European Radiology, vol. 30, pp. 1–8, 2020. [21] M. Mirza-Aghazadeh-Attari, A. Zarrintan, and N. Nezami, “Predictors of coronavirus disease 19 (COVID-19) pneu- monitis outcome based on computed tomography (CT) imaging obtained prior to hospitalization: a retrospective study,” EmergencyRadiology, vol. 27, pp. 1–9, 2020. [22] L. Li, L. Yang, S. Gui et al., “Association of clinical and ra- diographic findings with the outcomes of 93 patients with COVID-19 in Wuhan, China,” 7eranostics, vol. 10, no. 14, pp. 6113–6121, 2020. [23] M. Francone, F. Iafrate, and G. M. Masci, “Chest CT score in COVID-19 patients: correlation with disease severity and short-term prognosis,” European Radiology, vol. 30, pp. 1–10, [24] Y. Hu, C. Zhan, C. Chen, T. Ai, and L. Xia, “Chest CT findings related to mortality of patients with COVID-19: a retro- spective case-series study,” PLoS One, vol. 15, no. 8, Article ID e0237302, 2020. [25] K. Li, D. Chen, S. Chen et al., “Predictors of fatality including radiographic findings in adults with COVID-19,” Respiratory Research, vol. 21, pp. 146–210, 2020. [26] A. Kronbichler, M. Effenberger, M. Eisenhut, K. H. Lee, and J. I. Shin, “Seven recommendations to rescue the patients and reduce the mortality from COVID-19 infection: an immu- nological point of view,” Autoimmunity Reviews, vol. 19, no. 7, Article ID 102570, 2020. [27] M. D. Firouzabadi, S. Goudarzi, and F. D. Firouzabadi, “Complete heart block and itchy rash in a patient with COVID-19,” Caspian Journal of Internal Medicine, vol. 11, p. 569, 2020. [28] M.-M. Mehrabi Nejad, F. Moosaie, H. Dehghanbanadaki et al., “Immunogenicity of COVID-19 mRNA vaccines in immunocompromised patients: a systematic review and meta-analysis,” European Journal of Medical Research, vol. 27, no. 1, p. 23, 2022. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Radiology Research and Practice Hindawi Publishing Corporation

Loading next page...
 
/lp/hindawi-publishing-corporation/chest-ct-scan-features-to-predict-covid-19-patients-outcome-and-0igAqmOYbK
Publisher
Hindawi Publishing Corporation
Copyright
Copyright © 2022 Mohammad-Mehdi Mehrabi Nejad et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ISSN
2090-1941
eISSN
2090-195X
DOI
10.1155/2022/4732988
Publisher site
See Article on Publisher Site

Abstract

Hindawi Radiology Research and Practice Volume 2022, Article ID 4732988, 9 pages https://doi.org/10.1155/2022/4732988 Research Article Chest CT Scan Features to Predict COVID-19 Patients’ Outcome and Survival 1 1 1 Mohammad-Mehdi Mehrabi Nejad , Aminreza Abkhoo , Faeze Salahshour , 2 1 3 Mohammadreza Salehi , Masoumeh Gity , Hamidreza Komaki , and Shahriar Kolahi Department of Radiology, School of Medicine, Advanced Diagnostic and Interventional Radiology Research Center (ADIR), Imam Khomeini Hospital, Tehran University of Medical Sciences, Tehran, Iran Department of Infectious Diseases and Tropical Medicines, Tehran University of Medical Sciences, Tehran, Iran Brain Engineering Research Center, Institute for Research in Fundamental Sciences (IPM), Tehran, Iran Correspondence should be addressed to Faeze Salahshour; f-salahshour@sina.tums.ac.ir Received 11 April 2021; Revised 26 August 2021; Accepted 28 January 2022; Published 26 February 2022 Academic Editor: Sreekanth Kumar Mallineni Copyright © 2022 Mohammad-Mehdi Mehrabi Nejad et al. -is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background. Providing efficient care for infectious coronavirus disease 2019 (COVID-19) patients requires an accurate and accessible tool to medically optimize medical resource allocation to high-risk patients. Purpose. To assess the predictive value of on-admission chest CTcharacteristics to estimate COVID-19 patients’ outcome and survival time. Materials and Methods. Using a case-control design, we included all laboratory-confirmed COVID-19 patients who were deceased, from June to September 2020, in a tertiary-referral-collegiate hospital and had on-admission chest CT as the case group. -e patients who did not die and were equivalent in terms of demographics and other clinical features to cases were considered as the control (survivors) group. -e equivalency evaluation was performed by a fellowship-trained radiologist and an expert radiologist. Pulmonary involvement (PI) was scored (0–25) using a semiquantitative scoring tool. -e PI density index was calculated by dividing the total PI score by the number of involved lung lobes. All imaging parameters were compared between case and control group members. Survival time was recorded for the case group. All demographic, clinical, and imaging variables were included in the survival analyses. Results. After evaluating 384 cases, a total of 186 patients (93 in each group) were admitted to the studied setting, consisting of 126 (67.7%) male patients with a mean age of 60.4± 13.6 years. -e PI score and PI density index in the case vs. the control group were on average 8.9± 4.5 vs. 10.7± 4.4 (p value: 0.001) and 2.0± 0.7 vs. 2.6± 0.8 (p value: 0.01), respectively. Axial distribution (p value: 0.01), cardiomegaly (p value: 0.005), pleural effusion (p value: 0.001), and pericardial effusion (p value: 0.04) were mostly observed in deceased patients. Our survival analyses demonstrated that PI score≥ 10 (p value: 0.02) and PI density index≥ 2.2 (p value: 0.03) were significantly associated with a lower survival rate. Conclusion. On-admission chest CT features, particularly PI score and PI density index, are potential great tools to predict the patient’s clinical outcome. respiratory distress syndrome (ARDS) [2]. -e global death- 1. Introduction to-case ratio is estimated to be 3.5% [3]. However, it varies Coronavirus disease 2019 (COVID-19), caused by severe geographically probably due to the local preventive measures acute respiratory syndrome coronavirus-2 (SARS-CoV-2), and medical resources. -erefore, we need to improve the was officially announced as a pandemic by the World Health admitted patients’ initial triage not only to optimize the Organization on March 11, 2020 [1]. Even though most of allocation of the medical resources to high-risk patients and the patients experience mild symptoms, some may develop a minimize the mortality rate accordingly but also to more severe type of disease and it may progress to acute accurately and reliably predict the outcome of the patients. 2 Radiology Research and Practice Chest computed tomography (CT) scan, as the con- characteristics, age and sex; (b) on-admission vital signs, ventional and relatively accessible imaging modality for temperature (T-Celsius), oxygen saturation (SpO ), heart pneumonia diagnosis and follow-up, is confirmed to have rate (HR-per minute), respiratory rate (RR-per minute), and high diagnostic and prognostic values in the current out- blood pressure (BP-mmHg); (c) survival time, days from break of COVID-19 [4, 5]. Chest CT findings mainly consist admission to death in nonsurvived patients; (d) underlying of ground-glass opacities (GGOs), multifocal patchy con- diseases, hypertension (HTN), diabetes (DM), respiratory solidation, and interstitial changes with a peripheral dis- disease (asthma, COPD, ILD, or bronchiectasis), malig- tribution [6–8]. Efforts have been made to underpin the nancies (solid or hematological malignancies), immuno- predictive factors for mortality [9]. Among all, the main compromised conditions (chemoradiation therapy and factors consist of age, underlying disease (i.e., immuno- long-term corticosteroid usage), and hypothyroidism; and compromised patients and preexisting cardiovascular and (f) laboratory findings, white blood cell including neutrophil pulmonary disorders), laboratory findings (i.e., D-dimer and lymphocyte counts, hemoglobin, platelet, creatinine, level, neutrophil-to-lymphocyte ratio (NLR), urea, international normalized ratio (INR), partial throm- and lymphocyte count), and most recently, imaging features boplastin time (PTT), D-dimer, lactate dehydrogenase [10–13]. However, there is still no consensus on the factor (LDH),C-reactive protein (CRP), and pro-b-type natriuretic with the highest predictive value. Besides, almost all previous peptide (Pro-BNP). studies were retrospective cross-sectional and nonsurvivors comprised only a small population [14, 15], which further 2.2.2. Image Acquisition. All chest CT images were acquired limit the application of the findings. In this regard, we aimed to utilize the vastly used mo- at the time of admission, in the supine position, with full inspiration with no contrast injection. Examinations were dality-chest CT scan, and we hypothesize that on-admission chest CT findings could serve as a potential deterministic performed on either the Siemens Somatom Emotion (16 slices, Erlangen, Germany) or the Lightspeed 64-detector CT factor for risk stratification in hospitalized COVID-19 pa- tients. In order to evaluate the proposed notion, we con- (GE Healthcare, Milwaukee, USA) MDCT scanner. -e imaging parameters were set at 5–6 mm section thickness, ducted a case-control study and assessed the adjusted beam collimation of 0.6–2 mm, 120 kVp tube voltage, tube predictive value of CT scan in terms of mortality rate for admitted COVID-19 patients. current of 150–250 mAs, tube rotation speed of 0.75 seconds, and gantry rotation time of 0.5–0.75 s, reconstructed with a mediastinum B20f smooth kernel and a lung B70f sharp 2. Materials and Methods kernel (Siemens Healthineers, Erlangen, Germany); coronal 2.1. Study Design and Participants. -is case-control study and sagittal multiplanar reconstructions were also available was reviewed and approved by the Institutional Review with a reconstructed slice thickness of 1.2 mm. Board of our university. Given the retrospective design of the study and anonymous use of medical records, informed consent requirement was waived by the ethics committee of 2.2.3. Image Interpretation. Two fellowship-trained diag- our institute (IR.TUMS.VCR.REC.1399.054). -e current nostic imaging radiologists, with respective 9 and 13 years of case-control study was carried out in a referral tertiary experience in thoracic radiology and blinded to patients’ university hospital from June to September 2020. -is study outcomes, independently interpreted chest CT images’ evaluated individuals with the following conditions: (a) all findings. All CT images were reviewed on both lung- and hospitalized COVID-19 patients in whom COVID-19 di- mediastinal-window settings. -e intraclass correlation agnosis was confirmed by positive real-time reverse tran- coefficient (ICC) was calculated to assess inter-rater reli- scription-polymerase chain reaction (rRT-PCR) assay on ability. If ICC was less than 0.8, any disagreement in image nasopharyngeal or oropharyngeal swap or endotracheal interpretation for the case was discussed until resolved. If aspirate samples; (b) age equal or greater than 18 years; and ICC was greater than or equal to 0.8, the value reported by (c) a definite outcome of either death or hospital discharge. the radiologist with higher experience was recorded. Case patients consisted of patients who met the inclusion Chest CT scan findings were recorded according to the criteria and were deceased. Control patients included dis- Fleischner Society glossary and published literature on viral charged patients with equivalent demographic (age and sex) pneumonia [16]. Chest CT scan features included the fol- and clinical (underlying diseases and laboratory findings) lowing: (a) predominant pattern, ground-glass opacification/ features. -e equivalency evaluation was performed by a opacity (GGO) (Figure 2), consolidation (Figure 3), and fellowship-trained radiologist and an expert epidemiologist. mixed; (b) dominant distribution pattern, peripheral (pe- -e flow diagram of the study is presented in Figure 1. Of ripheral one-third of the lung) (Figure 3), axial (medial two- note, admission, discharge criteria, and treatment of all thirds of the lung), and diffuse (Figure 2); (c) number of patients were based on the national protocol of COVID-19. involved lobes; (d) other morphologies, parenchymal band, crazy paving, reverse halo sign, or intralesional traction bronchiectasis; and (e) additional findings, cardiomegaly 2.2. Data Collection (Figure 2), pleural effusion (unilateral or bilateral) (Fig- ure 2), pericardial effusion, emphysema, dilated pulmonary 2.2.1. Population Characteristics. -e recorded attributes of trunk, pleural thickening, and mosaic attenuation. patients consisted of the following: (a) demographic Radiology Research and Practice 3 Admitted patients with COVID-19 between June and September 2020 = 384 < 18 years old = 13 Patients with on-admission CT scan = 323 Incomplete medical record = 43 Patients with definite outcome = 247 No Matching = 61 Case group Control group (deceased) (discharged) N = 93 N = 93 Figure 1: Flow diagram of the study. Figure 2: A 76-year-old male patient with diabetes deceased due to COVID-19. Chest CT scan ((a) coronal view, (b) axial view-lung window, and (c) axial view-mediastinal window) showed predominancy of ground-glass opacity (GGO) with the bilateral and diffuse distribution. Pulmonary involvement (PI) and PI density scores were 16 and 3.2, respectively. Additional findings included cardiomegaly and bilateral pleural effusion. He was considered a high-risk patient. 2.2.4. Pulmonary Involvement (PI) Scoring System. To assess lobes’ scores. -e PI score ranged from 0 (no involvement) PI, a semiquantitative scoring tool was proposed and used to 25 (maximum involvement). Finally, the PI density index [17]. All five lung lobes (right upper lobe (RUL), right middle was calculated by dividing the total PI score by the number of involved lobes. lobe (RML), right lower lobe (RLL), left upper lobe (LUL), and left lower lobe (LLL)) were visually reviewed for GGO and consolidation. -en, a score from 0 to 5 was assigned to each lobe according to involvement percentage (0: no in- 2.3. Statistical Analysis. We performed the analyses in SPSS volvement; 1: ≤5%; 2: 6–25%; 3: 26–50%, 4: 51–75%; and 5: for Windows ver. 18 (Chicago, IL, USA). Descriptive data are ≥76%). -e total PI score was calculated as the sum of all five presented as mean with standard deviations (SD) for 4 Radiology Research and Practice Figure 3: A 61-year-old female patient with hypertension and diabetes. Pulmonary involvement: predominancy of GGO with peripheral, pleural-based distribution. Total pulmonary involvement (PI) score and PI density index were 6 and 1.2, respectively, and she was stratified as a low-risk patient in death predictive models. continuous variables and as frequency and percentage of the 3. Results population for categorical variables. In order to evaluate 3.1. Study Population Characteristics. After evaluating 384 whether the recorded data have a normal distribution, we cases, a total of 186 patients (93 in each group) were ad- used the Kolmogorov–Smirnov test. We conducted the mitted to the studied setting, consisting of 126 (67.7%) male comparisons by (a) the independent two-tailed sample t-test patients with a mean age of 60.4± 13.6 years (Figure 1). -e for continuous variables with normal distribution and the most common underlying diseases were HTN (38.7%) and relevant degree of freedom; (b) the Mann–Whitney U test DM (34.9%) (Table 1). for the continuous variables with significant lack of nor- Two groups were almost equivalent for demographic mality variables; and (c) the Chi-square test for nominal (age and sex) and clinical (underlying diseases and labo- variables. All p values less than 0.05 were considered sta- ratory findings) variables. Table 1 further illustrates the tistically significant. homogeneity of cases and controls in the aforementioned We used multivariate logistic regression to predict characteristics. All ICCs for inter-rater reliability were >0.8 mortality probability with all imaging parameters as the for all imaging parameters. independent variables. As previously stated, the statistically significant threshold was considered as a p value less than 0.05. To define optimum cutoff values for PI score and PI density index in outcome prediction, receiver-operating 3.2. Chest CT Scan Findings. -e most common CT features characteristic (ROC) curves were drawn and Youden’s J among survivors and nonsurvivors were GGO (65.6% and index [18] was calculated. -e area under the ROC curve 68.8%), multilobar (95.7% and 98.9%), bilateral lobe in- (AUC) was considered as the indicator for ROC analysis volvement (93.5% and 94.6%), and lower lobe (RLL and/or efficacy. LLL) involvement (94.6% and 98.9%), respectively. Pre- To determine the impact of any independent variable on dominant distribution patterns among the two groups were survival time, we implemented univariate Cox regressions peripheral (47.3% in survivors) and axial (57.0% in (in nonsurvivors) considering all demographic, clinical, and nonsurvivors). imaging parameters as covariates in the model. Multivariate -e mean PI score and PI density index in survivors vs. backward Cox regression was performed to find the final nonsurvivors were 8.9± 4.5 vs. 10.7± 4.4 (p value: 0.001) and model in terms of which variables to include. Kaplan–Meier 2.0± 0.7 vs. 2.6± 0.8 (p value: 0.01), respectively. In addition, survival analysis was performed to calculate survival, and the nonsurvived patients had a higher involvement score for all log-rank test was used to compare the survival distribution single lung lobes (except RLL) and axial distribution (37.6% of two subgroups of interest. vs. 57.0%, p value: 0.01). However, the number of involved Radiology Research and Practice 5 Table 1: Details of demographic and clinical data of patients and differences between two groups. Variables All patients, N � 186 Survivors, N � 93 Nonsurvivors, N � 93 P value Demographic data Age 60.4 (13.6) 58.7 (13.9) 62.2 (13.1) 0.08 Gender Male 126 (67.7) 61 (65.6) 65 (69.9) 0.53 Female 60 (32.3) 32 (34.4) 28 (30.1) Clinical data Vital signs RR 15.1 (9.1) 25.3 (5.1) 24.9 (11.8) 0.79 RR>24 102 (54.8) 51 (54.8) 51 (54.8) 1 Systolic BP 122.2 (20.1) 124.3 (18.0) 120.1 (21.8) 0.15 Diastolic BP 75.2 (12.7) 78.0 (10.9) 72.5 (13.7) 0.03 PR 96.9 (18.8) 97.4 (18.6) 96.4 (19.2) 0.70 Temperature 37.5 (0.8) 37.4 (0.8) 37.5 (0.9) 0.19 Hospitalization duration Total admission days 10.6 (8.9) 9.6 (8.0) 11.6 (9.7) 0.11 ICU days 5.2 (8.1) 3.0 (5.8) 7.4 (9.5) <0.001 Underlying disease HTN 72 (38.7) 32 (34.4) 40 (43.0) 0.29 DM 65 (34.9) 36 (38.7) 29 (31.2) 0.35 Respiratory disease 14 (7.5) 9 (9.7) 5 (5.4) 0.40 Immunocompromised 18 (9.7) 7 (7.5) 11 (11.8) 0.45 Hypothyroidism 13 (7.0) 9 (9.7) 4 (4.3) 0.24 Laboratory findings WBC 8.9 (4.7) 8.3 (4.3) 9.3 (5.0) 0.28 Neutrophil 7.0 (3.9) 6.5 (3.9) 7.4 (3.8) 0.24 Lymphocyte 1.4 (3.9) 1.2 (0.7) 1.5 (2.8) 0.52 Hemoglobin 12.5 (2.7) 12.8 (2.4) 12.3 (2.9) 0.40 Platelet 209.3 (95.2) 219.2 (110.9) 202.2 (82.4) 0.39 Cr 1.7 (1.6) 1.7 (1.3) 1.7 (1.8) 0.99 Urea 58.3 (60.8) 49.8 (36.1) 64.5 (73.4) 0.24 INR 1.3 (0.8) 1.4 (1.1) 1.2 (0.5) 0.51 PTT 41.4 (22.2) 43.4 (25.1) 40.0 (19.8) 0.47 D-dimer 3364.1 (3210.0) 3450.7 (3344.9) 3147.7 (3316.3) 0.88 LDH 661.9 (287.3) 676.6 (269.7) 649.1 (305.7) 0.72 CRP 123.8 (73.0) 113.3 (72.1) 130.6 (73.4) 0.24 Pro-BNP 5329.9 (10289.0) 4505.8 (10687.2) 6085.2 (10324.3) 0.72 Mean (standard deviation); all other variables reported as N (%). RR � respiratory rate; BP � blood pressure; PR � pulse rate; HTN � hypertension; DM � diabetes; ICU � intensive care unit; WBC � white blood cell; Cr � creatinine; INR � international normalized ratio; PTT �partial thromboplastin time; LDH � lactate dehydrogenase; CR � C-reactive protein; Pro-BNP � pro-b-type natriuretic peptide. Table 2: Radiologic findings in all patients and differences between the two groups. Variables All patients, N � 186 Survivors, N � 93 Nonsurvivors, N � 93 P value PI scores RUL total score 2.0 (1.1) 1.8 (1.0) 2.2 (1.1) 0.02 RML total score 1.5 (1.0) 1.3 (0.9) 1.6 (1.0) 0.03 RLL total score 2.3 (1.1) 2.1 (1.1) 2.4 (1.1) 0.15 LUL total score 2.0 (1.1) 1.7 (1.1) 2.2 (1.1) 0.02 LLL total score 2.2 (1.2) 2.0 (1.2) 2.3 (1.2) 0.03 Total lung GGO score 6.7 (4.5) 6.2 (4.2) 7.2 (4.8) 0.13 Total lung consolidation score 3.1 (3.2) 2.6 (3.0) 3.6 (3.5) 0.05 Total PI score 9.8 (4.5) 8.9 (4.5) 10.7 (4.4) 0.001 PI density index 2.1 (0.8) 2.0 (0.7) 2.6 (0.8) 0.01 Predominant pattern GGO 125 (67.2) 61 (65.6) 64 (68.8) 0.63 Consolidation 61 (32.8) 32 (34.4) 29 (31.2) Dominant distribution of lesions 6 Radiology Research and Practice Table 2: Continued. Variables All patients, N � 186 Survivors, N � 93 Nonsurvivors, N � 93 P value Peripheral 70 (37.6) 44 (47.3) 26 (28.0) Axial 88 (47.3) 35 (37.6) 53 (57.0) 0.01 Diffuse 28 (15.1) 14 (15.1) 14 (15.1) No. of involved lobes 4.5 (1.0) 4.5 (1.1) 4.6 (0.9) 0.49 Monolobar 1 5 (2.7) 4 (4.3) 1 (1.1) Multilobar 2 5 (2.7) 1 (1.1) 4 (4.3) 0.05 3 12 (6.5) 3 (3.2) 9 (9.7) 4 15 (8.1) 11 (11.8) 4 (4.3) 5 149 (80.1) 74 (79.6) 75 (80.6) Laterality Unilateral 11 (6.1) 6 (6.9) 5 (5.4) 0.76 Bilateral 175 (94.1) 87 (93.5) 88 (94.6) Lower lobe involvement Yes 180 (96.8) 88 (94.6) 92 (98.9) 0.09 No 6 (3.2) 5 (5.4) 1 (1.1) Additional findings Cardiomegaly 85 (45.7) 33 (35.5) 52 (55.9) 0.005 Mosaic attenuation 7 (3.7) 3 (3.2) 4 (4.3) 1 Pleural effusion 35 (18.8) 9 (9.7) 26 (28.0) 0.001 Unilateral 12 (6.5) 3 (3.2) 9 (9.7) 0.07 Bilateral 23 (12.4) 6 (6.5) 17 (18.3) 0.01 Pericardial effusion 13 (7.0) 3 (3.2) 10 (10.8) 0.04 Emphysema 7 (3.8) 3 (3.2) 4 (4.3) 1 Dilated pulmonary trunk 23 (12.4) 5 (5.4) 18 (19.4) 0.004 Pleural thickening 34 (18.3) 3 (3.2) 31 (33.3) <0.001 Subsegmental atelectasis 64 (34.4) 24 (25.8) 40 (43.0) 0.01 Other morphologies Parenchymal band 83 (44.6) 42 (45.2) 41 (44.1) 0.88 Crazy paving 69 (37.1) 29 (31.2) 40 (43.0) 0.09 Reverse halo 14 (7.5) 7 (7.5) 7 (7.5) 1 In-lesion bronchiectasis 11 (5.9) 7 (7.5) 4 (4.3) 0.35 Mean (standard deviation); all other variables reported as N (%). PI � pulmonary involvement; RUL � right upper lobe; RML � right middle lobe; RLL � right lower lobe; LUL � left upper lobe; LLL � left lower lobe; GGO � ground-glass opacity. lobes, predominant pattern, and bilateral and lower lobe (AUC: 0.61 [0.52–0.69], Youden’s J: 0.19, p value: 0.01) involvement did not show any significant difference in showed the best accuracy in predicting death (Figures 2–3). comparison (p values: 0.49, 0.63, 0.76, and 0.09) (Table 2). Of additional findings and other morphologies, car- diomegaly (35.5% vs. 55.9%, p value: 0.005) (Figure 2), 3.3. Survival Analysis. -e survival analyses were limited to nonsurvived patients. -e mean survival was 4.1 ± 3.5 days, and pleural effusion (9.7%% vs. 28.0%, p value: 0.001) (Figure 2), the median survival was four days. -e Kaplan–Meier survival pericardial effusion (3.2% vs. 10.8%, p value: 0.04), dilated function for death is illustrated in Figure 4. To determine the pulmonary trunk (5.4% vs. 19.4%, p value: 0.004), and impact of independent variables on survival, we implemented pleural thickening (3.2% vs. 33.3%, p value:<0.001) were significantly more prevalent in the nonsurvived group. univariate Cox regressions, considering demographic, clinical, and imaging variables as independent variables. Statistically Table 2 demonstrates the details of patients’ characteristics associated with death. significant variables included PI score≥ 10 (p value: 0.02) and PI density≥ 2.2(p value: 0.03). Afterward, we fitted a multi- We exploited backward multivariate logistic regressions that included death as the outcome of interest and all im- variate backward Cox regression on the variables with a sig- nificant univariate association: PI score≥ 10 (p value: 0.03) aging parameters with significant association in Table 2 as the independent variables. PI score (Ex(B): 1.10 (95% CI: remained significant at the final step (Table 3). Further analyses using the Kaplan–Meier survival 1.03–1.17), p value: 0.006)) and PI density index (Ex(B): 1.64 function and the log-rank test revealed that survival was (95% CI: 1.10–2.43), p value: 0.02)) independently remained as significant variables in the models. -e PI score with a significantly different in two subgroups according to PI score≥ 10 (χ cutoff value of 10 (AUC: 0.62 [0.54–0.70], Youden’s J: 0.23, p � 7.05, p value: 0.008) and PI density≥ 2.2 value: 0.005) and PI density index with a cutoff value of 2.2 (χ � 6.58, p value: 0.01) (Figure 4). Radiology Research and Practice 7 1.0 1.0 1.0 0.8 0.8 0.8 0.6 0.6 0.6 0.4 0.4 0.4 0.2 0.2 0.2 0.0 0.0 0.0 0.00 2.00 4.00 6.00 8.00 10.0012.00 14.00 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 survival days survival days survival days PI PI score density <10 <2.2 ≥10 ≥2.2 (a) (b) (c) Figure 4: Kaplan–Meier survival curve. Estimated survival rate in (a) all deceased patients and comparisons based on (b) PI score and (c) PI density score. Table 3: Univariate and multivariate Cox regression. Variable Coefficient of variable in the model Exp(B) (95% CI) p value of variables p value of models Univariate Cox regression PI score≥ 10 0.60 1.8 [1.09–3.01] 0.02 0.02 PI density score≥ 2.2 0.57 1.77 [1.10–2.91] 0.03 0.03 Multivariate backward Cox regression PI score≥ 10 0.56 1.75 [1.05–2.9] 0.03 0.01 PI: pulmonary involvement; CI: confidence interval. used a 20-scale scoring system in 50 patients (23 deceased, 4. Discussion 27 survived) and suggested a cutoff value of 12 with 0.79 -ere is still no consensus on the clinical and imaging factors AUC in predicting mortality [21]. Moreover, a study used a that are associated with and affect COVID-19 patients’ 72-scale CT score and reported 85% sensitivity and speci- outcomes and survival. Our findings contribute to the ficity in predicting the mortality of patients with a cutoff existing literature by illustrating that patients with higher PI value of 24.5 [14]. In general, all previous studies were scores, PI density index, axial distribution, cardiomegaly, or retrospective and nonsurvivors mostly comprised a small pleural or pericardial effusion are more likely to decease. -e proportion of the studied population. Hence, differences in data PI density score is our novel suggested index to distinguish sampling and study design (case-control vs. cross-sectional), as between patients with the same PI score but a different well as group matching, in our study remarkably improved number of involved lobes because it showed a stronger the validity of results. Furthermore, details of chest CT mortality prediction power. findings were not fully reported in previous studies, and Previously reported chest CT features to predict there is no consistency on the value of additional findings, COVID-19 patients’ mortality mainly included crude lung including crazy paving and pleural effusion in predicting the involvement score, number of involved lobes, bilateral or outcome [19, 21, 25], which are both depicted in this study. lower lobe involvement, and diffuse pattern [14, 19–25]. In Further analyses on patients’ survival demonstrated that line with our findings, yet in contrast to previous reports, PI score <10 and PI density index <2.2 were significantly Yuan et al. [14] reported no significant association of lower associated with higher survival. To the best of our knowl- lobe or bilateral lung involvement with death. Nevertheless, edge, imaging factors associated with the survival days of a significantly higher CT score was the most common re- COVID-19 patients were addressed only in one retrospec- ported feature to predict death [14, 19–23, 25]. Limited tive study on 20 deceased patients [23]. In line with our previous studies also reported cutoff values using totally findings, they found significantly lower survival days in different scoring systems [14, 21, 23]. For instance, Francone patients with higher (≥18) vs. lower (<18) CT scores over a et al. [23] retrospectively evaluated 130 COVID-19 patients 24-day follow-up period [23]. (20 patients deceased) and recommended the CT score≥ 18 Considering the limitations of the study, we still (out of 25) as the predictive factor for death. Another study performed the largest case-control investigation on Cum Survival Cum Survival Cum Survival 8 Radiology Research and Practice evaluating the clinical and imaging factors to predict Acknowledgments COVID-19 patients’ outcomes and survival. We provided -e authors are thankful to the patients and hospital staff for a novel semiquantitative scale with a defined optimum their collaboration. cutoff, which could serve to better identify high-risk pa- tients and recognize patients with a higher demand for critical care but similar clinical conditions compared to References other patients. We posit that patients with a PI score≥ 10 [1] WHO, Novel Coronavirus (2019-nCoV) Situation Report-51, and a PI density score≥ 2.2 should be carefully cared for https://www.who.int/docs/default-source/coronaviruse/ and receive aggressive treatment, as they are highly sus- situation-reports/20200311-sitrep-51-covid-19.pdf, WHO, ceptible to death. Hence, radiologists are expected to re- Geneva, Switzerland, 2020, https://www.who.int/docs/ port PI and PI density scores in their everyday practice to default-source/coronaviruse/situation-reports/20200311- help clinical physicians manage patients more effectively sitrep-51-covid-19.pdf. and efficiently and consequently minimize COVID-19- [2] N. Chen, M. Zhou, X. Dong et al., “Epidemiological and related mortality. Nevertheless, our findings must be clinical characteristics of 99 cases of 2019 novel coronavirus interpreted in light of some limitations. Firstly, the study pneumonia in Wuhan, China: a descriptive study,” 7e context is prone to a biased selection of cases and controls. Lancet, vol. 395, no. 10223, pp. 507–513, 2020. To elaborate, the setting is a tertiary and referral hospital [3] V. J. Munster, M. Koopmans, N. van Doremalen, D. van Riel, and E. de Wit, “A novel coronavirus emerging in China - key for COVID-19 patients, and the patients who get admitted questions for impact assessment,” New England Journal of are probably more clinically severe than the cases in the Medicine, vol. 382, no. 8, pp. 692–694, 2020. general population with similar paraclinical and demo- [4] Y. Fang, H. Zhang, and J. Xie, “Sensitivity of chest CT for graphic features. On the other hand, there are reports on COVID-19: comparison to RT-PCR,” 7e Radiologist, factors, such as immunological factors and predispositions vol. 296, 2020. in patients, that might have an impact on the survival rate [5] A. Jalali, E. Karimialavijeh, and P. Babaniamansour, “Pre- of the COVID-19 cases [26,27]. However, we did not have dicting the 30-day adverse outcomes of non-critical new- other available data to incorporate in the study. In addi- onset COVID-19 patients in emergency departments based tion, the admitted patients are not equivalent to outpatient on their lung CT scan findings; a pilot study for derivation an cases in the predicted mortality rate. For example, out- emergency scoring tool,” Frontiers in Emergency Medicine, patient cases might have an even lower mortality rate with vol. 5, p. e40, 2021. [6] E. Lanza, R. Muglia, I. Bolengo et al., “Quantitative chest CT a similar PI score because they are in a relatively better analysis in COVID-19 to predict the need for oxygenation clinical condition. Another limitation is that we did not support and intubation,” European Radiology, vol. 30, no. 12, follow the discharged patients to evaluate whether they pp. 6770–6778, 2020. were deceased. -e interval time between symptom onset [7] R. Han, L. Huang, and H. Jiang, “Early clinical and CT and admission was not exactly the same in all patients; manifestations of coronavirus disease 2019 (COVID-19) however, we used the on-admission CT scans to minimize pneumonia,” American Journal of Roentgenology, vol. 61, this bias. Although all patients were treated based on a pp. 1–6, 2020. unique national protocol, minor treatment differences [8] T. Ai, Z. Yang, and H. Hou, “Correlation of chest CT and RT- based on physician clinical judgment could affect mortality PCR testing in coronavirus disease 2019 (COVID-19) in and survival, which were not assessed in the present study. China: a report of 1014 cases,” 7e Radiologist, vol. 22, Article ID 200642, 2020. -is study was conducted before the national COVID-19 [9] F. Salahshour, M.-M. Mehrabinejad, and M. N. Toosi, vaccination program and further investigations on a larger “Clinical and chest CT features as a predictive tool for population after vaccination are recommended to confirm COVID-19 clinical progress: introducing a novel semi- our findings [28]. quantitative scoring system,” European Radiology, vol. 31, In conclusion, patients with a higher PI score, PI density pp. 1–11, 2021. index, axial distribution, cardiomegaly, or pleural or peri- [10] X. Zhao, B. Zhang, and P. Li, “Incidence, clinical charac- cardial effusion are more susceptible to poor prognosis. teristics and prognostic factor of patients with covid-19: Survival analyses revealed that a PI score≥ 10 and PI density asystematic review and meta-analysis,” medRxiv, vol. 20, 2020. score≥ 2.2 were significantly associated with lower survival [11] S. Babaniamansour, A. Atarodi, and P. Babaniamansour, and those patients should be prioritized for higher medical “Clinical findings and prognosis of COVID-19 patients with attention. Initial chest CT examination may be able to help benign prostatic hyperplasia: acase series,” Journal of Bio- medical Research&EnvironmentalSciences, vol. 2, 2021. in predicting the patient’s outcome and survival. [12] S. Goudarzi, F. Dehghani Firouzabadi, M. Dehghani Fir- ouzabadi, and N Rezaei, “Cutaneous lesions and COVID-19: Data Availability cystic painful lesion in a case with positive SARS-CoV-2,” Dermatologic 7erapy, vol. 33, no. 6, p. e14266, 2020. All the data used in this manuscript can be made available [13] H. Dehghanbanadaki, H. Aazami, M. Shabani et al., “A upon request to the corresponding author. systematic review and meta-analysis on the association be- tween lymphocyte subsets and the severity of COVID-19,” Immunopathologia Persa, 2022. Conflicts of Interest [14] M. Yuan, W. Yin, Z. Tao, W. Tan, and Y. Hu, “Association of -e authors declare that they have no conflicts of interest. radiologic findings with mortality of patients infected with Radiology Research and Practice 9 2019 novel coronavirus in Wuhan, China,” PLoS One, vol. 15, no. 3, Article ID e0230548, 2020. [15] S. Richardson, J. S. Hirsch, M. Narasimhan et al., “Presenting characteristics, comorbidities, and outcomes among 5700 patients hospitalized with COVID-19 in the New York City area,” JAMA, vol. 323, no. 20, pp. 2052–2059, 2020. [16] D. M. Hansell, A. A. Bankier, H. MacMahon, T. C. McLoud, N. L. Muller, ¨ and J. Remy, “Fleischner Society: glossary of terms for thoracic imaging,” Radiology, vol. 246, no. 3, pp. 697–722, 2008. [17] A. Abkhoo, E. Shaker, and M.-M. Mehrabinejad, “Factors predicting outcome in intensive care unit-admitted COVID- 19 patients: using clinical, laboratory, and radiologic char- acteristics,” Critical Care Research and Practice, vol. 2021, Article ID 9941570, 2021. [18] W. J. Youden, “Index for rating diagnostic tests,” Cancer, vol. 3, no. 1, pp. 32–35, 1950. [19] F. Pan, C. Zheng, T. Ye et al., “Different computed tomog- raphy patterns of Coronavirus Disease 2019 (COVID-19) between survivors and non-survivors,” Scientific Reports, vol. 10, pp. 11336–11339, 2020. [20] Y. Li, Z. Yang, and T. Ai, “Association of “initial CT” findings with mortality in older patients with coronavirus disease 2019 (COVID-19),” European Radiology, vol. 30, pp. 1–8, 2020. [21] M. Mirza-Aghazadeh-Attari, A. Zarrintan, and N. Nezami, “Predictors of coronavirus disease 19 (COVID-19) pneu- monitis outcome based on computed tomography (CT) imaging obtained prior to hospitalization: a retrospective study,” EmergencyRadiology, vol. 27, pp. 1–9, 2020. [22] L. Li, L. Yang, S. Gui et al., “Association of clinical and ra- diographic findings with the outcomes of 93 patients with COVID-19 in Wuhan, China,” 7eranostics, vol. 10, no. 14, pp. 6113–6121, 2020. [23] M. Francone, F. Iafrate, and G. M. Masci, “Chest CT score in COVID-19 patients: correlation with disease severity and short-term prognosis,” European Radiology, vol. 30, pp. 1–10, [24] Y. Hu, C. Zhan, C. Chen, T. Ai, and L. Xia, “Chest CT findings related to mortality of patients with COVID-19: a retro- spective case-series study,” PLoS One, vol. 15, no. 8, Article ID e0237302, 2020. [25] K. Li, D. Chen, S. Chen et al., “Predictors of fatality including radiographic findings in adults with COVID-19,” Respiratory Research, vol. 21, pp. 146–210, 2020. [26] A. Kronbichler, M. Effenberger, M. Eisenhut, K. H. Lee, and J. I. Shin, “Seven recommendations to rescue the patients and reduce the mortality from COVID-19 infection: an immu- nological point of view,” Autoimmunity Reviews, vol. 19, no. 7, Article ID 102570, 2020. [27] M. D. Firouzabadi, S. Goudarzi, and F. D. Firouzabadi, “Complete heart block and itchy rash in a patient with COVID-19,” Caspian Journal of Internal Medicine, vol. 11, p. 569, 2020. [28] M.-M. Mehrabi Nejad, F. Moosaie, H. Dehghanbanadaki et al., “Immunogenicity of COVID-19 mRNA vaccines in immunocompromised patients: a systematic review and meta-analysis,” European Journal of Medical Research, vol. 27, no. 1, p. 23, 2022.

Journal

Radiology Research and PracticeHindawi Publishing Corporation

Published: Feb 26, 2022

References