Diagnostic value of HRCT-Thorax for pandemic COVID-19 pneumonia in Pakistan

Background: In the scenario of, inadequate testing, the low sensitivity of the COVID-19-PCR test, limited availability of testing kits, and low detection rate, we aimed to investigate the usefulness of high-resolution computed tomography of chest (HRCT) for diagnosing pandemic coronavirus (COVID-19) pneumonia. Objective: To determine the diagnostic efficacy of HRCT thorax in Covid-19 pandemic pneumonia. Materials and Methods: This prospective, cross-sectional study was conducted in the Pulmonology–OPD of Gulab Devi Teaching Hospital, Lahore-Pakistan from 01-04-2020 to 15-07-2020. 121 patients with dry cough, fever, and dyspnea of sudden onset were included while patients with Bronchial Asthma, ILD, Tuberculosis, Bronchiectasis, COPD, and overt heart failure were excluded. Patients were investigated with chest x-ray, HRCT, COVID-PCR, and hematological tests. HRCT films were evaluated by a qualified and experienced radiologist. Findings were summarized, organized and statistical analysis was done by using SPSS-26 software to make an inference. Results: Five patients were diagnosed as non-covid. Out of 116-diagnosed covid-19 patients, 38(32.75%) showed sub-pleural consolidation, 19(16.37%) consolidation with air-bronchogram, 29(25.0%) crazy paving sign, one pleural effusion (0.86%) and 18 cases (15.51%) displayed reticulations. 11cases(9.48%) had isolated ground glass appearances, while all categories showed it to variable extent. 65 patients (56.03%) were PCR-positive while 51(43.96) patients with positive-HRCT findings for COVID-19 pneumonia had negative nasopharyngeal-PCR. HRCT-Thorax revealed sensitivity: 97.41 %, specificity: 80%, PPV: 99.12%, NPV: 57.14%, and diagnostic accuracy of 96.69% for COVID-19 pneumonia. Conclusion: HRCT-Thorax, having high sensitivity and adequate specificity, can provide foundations for evidence-based early diagnosis and quantification of coronavirus pneumonia. It can be tremendously useful for decision making in PCR-negative patients and anticipating respiratory improvement or decline by serial scans.


Introduction
The first case of coronavirus pneumonia was noticed in Wuhan, China in December 2019 and was later on recognized as coronavirus disease 2019 . 1 This epidemic spread worldwide like wildfire. WHO (World Health Organization) declared it as a Public Health Emergency of International Concern (PHEIC). 2 This disease is highly contagious and the number of cases multiplied rapidly over the globe. Current literature suggests that one infected patient may lead to more than two new patients. 3 Some-times, the disease is too drastic to be controlled and the outcome is a disaster only. This disease has challenged almost all fields of medicine and has affected not only the public rather very senior health care providers and eminent figures from the society have been lost to this lethal disease. At the mid of July, we are having 255,769 confirmed cases and 5,386 mortalities in the country, indicating the load and damage caused by this fatal disorder. The presentation of COVID-19 disease ranges from mild to an extensive disease requiring mechanical ventilation. Early diagnosis is crucial for the good management and effective control to preclude further dissemination. Usually patient presents with fever, dry cough, shortness of breath, and body aches and pains. The severity and acuteness of the symptoms, coupled with the propensity for an abrupt decline in respiratory function, necessitates the need to measure the magnitude of pulmonary involvement. Nasopharyngeal-PCR is the gold standard for diagnosis but its availability and sensitivity are quite low, around 60%. About 30-40% of cases are missed by this technique due to a high false-negative rate. 4 Furthermore, this test is not capable of calculating the magnitude and expected respiratory decline in patients. Chest X-ray PA view is limited in its ability to detect the early involvement of lung with COVID-19 infection. 5 In contrast, HRCT-Thorax has been reported as a good tool for evidence-based diagnosis in the current epidemic, providing immediate results with high sensitivity and adequate specificity for COVID-19 pneumonia and the extent of pulmonary disease. 6-7 HRCT-thorax can play a pivotal role by identifying the early phase lung infection and have been recommended as major evidence for clinical diagnosis in China. 8 Current researchers have reported the sensitivity of HRCT for COVID-19 infection more than 90% as compared to PCR around 60-70%. According to WHO, HRCT can be very useful in determining the diagnosis, progression, and severity of the COVID-19 disease. 9-10 The use of HRCT in Pakistan is very limited due to profound fear, contagiousness & inadequate understanding and is tried only in selected cases. [11][12] The Chinese literature is rich in evidence, advocating the use of HRCT-thorax while the European, British and American radiological societies which were not recommending this modality in the beginning, are changing their statements about the use of HRCT. [13][14][15] Radiological Society of Pakistan has also suggested, the appropriate use of CT scan in selected patients as a tool to triage, in the background of increased incidence of COVID-19 symptomatic patients. 16

Materials and Methods
This prospective study was conducted in Pulmonology-OPD of Gulab Devi Teaching Hospital, Lahore. (A 1500 bed-tertiary care hospital in the capital city of Punjab province). Ethical approval was obtained from the IRB of the hospital vide: No. Admin/GDEC/20/325. After informed consent, 121 OPD patients, from 1 st April 2020 to 15 July 2020 with a presumptive clinical diagnosis of COVID-19 infection were included. History of dry cough, fever, and shortness of breath of sudden onset were the main complaints while patients with Bronchial Asthma, ILD, Tuberculosis, Bronchiectasis, COPD, and overt heart failure were excluded. All patients underwent routine hematological tests including white blood cell count, lymphocyte count, and hypersensitive Creactive protein tests. The patients were subjected to an X-ray chest PA view, HRCT thorax, ECG, and echocardiography. The x-ray chest (CXR) & HRCT images were evaluated by a qualified radiologist with 20 years of experience at least. CXR was analyzed for haze and ill-defined consolidations with bilateral involvement. HRCT films were evaluated for radiological morphology and distribution patterns. ECG & Echocardiography explored cardiac status. All patients underwent the COVID-19 PCR-test by nasopharyngeal sample. Imaging findings, demographic and clinical data of the 121 patients were recorded on a preformed form. Sensitivity, specificity, and diagnostic accuracy were calculated by considering nasopharyngeal-PCR as a gold standard. Diagnostic yield was calculated on clinical, PCR, and HRCT grounds independently. Data was organized, summarized, tabulated and bio-statistical analysis was done to conclude. For statistical analysis, SPSS-26 software was utilized. Quantitative data were described by + SD and categorical variables were expressed by frequency (percentage). Fissure exact test was used for comparison and a p-value < 0.05 was considered as significant.

Results
This study included 121 patients age 18 to 67 years. The mean age was 42.64 years ± SD 17.62. Male gender was noted in 67 patients while 54 patients were female, male to female ratio was 1.2. All patients had contact history from their familial clusters. Patients presented with typical respiratory complaints (  The radiological morphology and radiographic appearance are described by Figure 1 and Figure 2. Not even a single case of cavitation or pneumothorax was found in 116 cases. Disease distribution pattern is depicted in Table 2.    Because this virus is highly contagious a very much unpredictable in behavior, its control needs early diagnosis for timely isolation or quarantine. In these circumstances, there is an urgent need for some alternative high sensitivity test, capable of capturing the disease untimely, for achieving good control in population. On the other hand, 87/116 patients were identified by x-ray chest (CXR) with a sensitivity of 75.0%. Although diagnostic yield is better than that of the gold standard (PCR) 25.0% patients were missed, while abnormal findings were noted on HRCT images of the same patients which are in agreement with the report of Van der Bruggen-Bogaarts BA1, Van der Bruggen HM. 23 The current study showed that HRCT thorax diagnosed viral pneumonia in 113 out of 116 cases with a sensitivity of 97.41%. Although it furnishes specific signs regarding COVID-19 pneumonia, it is not considered as specific as PCR, but according to the principles of differential diagnosis, we should always think about common issues. In the current on-going epidemic milieu, COVID-19 infection is the commonest issue, so HRCT can be utilized with confidence for its diagnosis. This study exhibited that both lungs were affected in 100% cases, the predominant affected part was the lower zone in 78.51 % of patients. GGO was the earliest and the commonest (82.75% cases) sign on HRCT. Out of 96 cases, 11 cases (11.45%) showed isolated GGO scattered in lung fields, 38/96 cases (39.58%) were associated with consolidation, and 29/96 cases (30.20%) were part of the crazy paving sign. Modern literature shows that in COVID-19 pneumonia, type II alveolar epithelial cell injury, proteinaceous exudate, focal hyperplasia of pneumocytes and patchy inflammatory cellular infiltration, contribute to this GGO. [24][25][26] Consolidation was noted in 57cases, 38/57 cases (66.67%) showed sub-pleural, multifocal consolidation, and 19/57 cases (33.33%) were associated with the airbronchogram sign. All cases displayed GGO to a variable extent. These findings are very much comparable with the published report of Cheng Z, Lu Y and co-authors. 27 When septal thickening occurs in addition to GGO, crazy paving appearance is manifested on HRCT which is found in moderate to severe COVID-19 Pneumonia. This study communicated 29 cases (25.0%) with this morphology while X. He, J. Zheng Dr., and co-workers have also described 25.0% of patients with the crazy paving sign. 28 These patients pose a practical challenge because of having a potential of transition into serious disease and anticipated mechanical ventilation, in the future. Many precious lives have been lost to this pattern. 29 We found only one case of pleural effusion in this study which was bilateral and cardiogenic, secondary to coronary bypass surgery with superimposed COVID-19 infection. Otherwise, pleural effusion is not a common sign of this disease. But Huang et al. reported that pleural effusion may be noted in severe COVID-19 pneumonia. 30 This study displayed the efficacy of HRCT-thorax as sensitivity: 97.41 %, specificity: 80%, PPV: 99.12 %, NPV: 57.14%, diagnostic accuracy: 96.69 % for pulmonary Covid-19 disease. The detailed efficacy of COVID-PCR is described in table-III. The p-value of 0.0006 is < 0.05 and is highly significant, showing a remarkable difference between the diagnostic potential of the two modalities. This value indicates that HRCT has excellent capabilities of diagnosing COVID-19 pulmonary disease in an epidemic. According to the current literature, the sensitivity of PCR is as low as 60%. 31 False-negative results can create the risk of inadvertent contamination of non-COVID-19 patients, wards, and the community. Lower respiratory symptoms require tracheal aspirates or BAL (broncho-alveolar lavage) for maximum diagnostic yield but there is a very high risk of aerosolization & disease transmission to healthcare workers during these procedures. 32 While HRCT is free of such drawbacks and was widely used in China as a diagnostic tool for COVID-19 pulmonary disease. HRCT can give rapid and valuable information for triaging isolation, and treatment procedures. It also determines the extent of lung involvement and is of tremendous help in anticipating the risk of rapid respiratory decline. No doubt, performing HRCT for every respiratory patient is neither cost-effective nor long-term sustainable. But this modality can be used during an epidemic, especially where high sensitivity reliable laboratory testing tool for COVID-19 is not available. In summary, HRCT features of COVID-19 pneumonia include ground-glass opacity, multi-focal, sub-pleural & basal consolidation, and crazy paving sign. Reticulation is found in patients with 2-3 weeks of history. We hope, the current study findings can facilitate early identification and good management of symptomatic and suspected COVID-19 pneumonia cases by adequate use of HRCT thorax, under adequate biosafety measures.

Limitations
The main limitation of our study is that it is a singlecenter study performed on limited sample size (121), so the study may not reflect a full range of radiological morphology and distribution pattern of COVID-19 pulmonary disease. To additionally elucidate, a study with larger sample size or preferably, a multi-center study can further explore the radiological behavior of the disease. The 15.51% cases with reticulations found in the current study need further interpretation by a longitudinal study whether these reticulations are going to resolve completely or going to develop into a new cluster of ILD in the community.

Conclusion
HRCT thorax, having excellent sensitivity and significant specificity for COVID-19 pulmonary infection, is a superb modality. It can play a positive role in early detection, quantification, and management of pulmonary coronavirus disease during a pandemic. If available, can be used with confidence for a re-evaluation of PCR-negative suspect for decision making about treatment and anticipating the risk of respiratory decline.