ABSTRACT
CONTEXT: As the burden of COVID-19 enhances, the need of a fast and reliable screening method is imperative. Chest radiographs plays a pivotal role in rapidly triaging the patients. Unfortunately, in low-resource settings, there is a scarcity of trained radiologists. AIM: This study evaluates and compares the performance of an artificial intelligence (AI) system with a radiologist in detecting chest radiograph findings due to COVID-19. SUBJECTS AND METHODS: The test set consisted of 457 CXR images of patients with suspected COVID-19 pneumonia over a period of three months. The radiographs were evaluated by a radiologist with experience of more than 13 years and by the AI system (NeuraCovid, a web application that pairs with the AI model COVID-NET). Performance of AI system and the radiologist were compared by calculating the sensitivity, specificity and generating a receiver operating characteristic curve. RT-PCR test results were used as the gold standard. RESULTS: The radiologist obtained a sensitivity and specificity of 44.1% and 92.5%, respectively, whereas the AI had a sensitivity and specificity of 41.6% and 60%, respectively. The area under curve for correctly classifying CXR images as COVID-19 pneumonia was 0.48 for the AI system and 0.68 for the radiologist. The radiologist's prediction was found to be superior to that of the AI with a P VALUE of 0.005. CONCLUSION: The specificity and sensitivity of detecting lung involvement in COVID-19, by the radiologist, was found to be superior to that by the AI system.
ABSTRACT
The importance of this study is the efficacy of "symptoms only" approach at a screening clinic for coronavirus disease 2019 (COVID-19) diagnosis in low- and middle-income countries (LMIC) setting. The objective of this study was to assess how efficiently primary care physicians at the screening clinic were able to predict whether a patient had COVID-19 or not, based on their symptom-based assessment alone. The current study is a cross-sectional retrospective observational study. This study was conducted at a single-center, tertiary care setting with a dedicated COVID-19 facility in a metropolitan city in eastern India. Participants are all suspected COVID-19 patients who presented themselves to this center during the outbreak from 1 August 2020 to 30 August 2020. Patients were referred to the Cough Clinic from the various outpatient departments of the hospital or from smaller satellite centers located in different parts of the city and other dependent geographical areas. The main outcome(s) and measure(s) is to study whether outcome of confirmatory test results can be predicted accurately by history taking alone. From 01 August 2020 to 30 Aug 2020, 511 patients with at least one symptom suggestive of COVID-19 reported to screening clinic. Out of these, 65.4% were males and 34.6% were females. Median age was 45 years with range being 01 to 92 years. Fever was seen in 70.4% while cough was present in 22% of cases. Overall positivity for SARS-CoV-2 during this period in this group was 54.21%. At 50% pre-test probability, the sensitivity of trained doctors working at the clinic, in predicting positive cases based on symptoms alone, was approximately 74.7%, and specificity for the same was 58.12%. The positive predictive value of the doctors' assessment was 67.87%, and the negative predictive value was 66.02%. Rapid triaging for confirmatory diagnosis of COVID-19 is feasible at screening clinic based on history taking alone by training of primary care physicians. This is particularly relevant in LMIC with scarce healthcare resources to overcome COVID-19 pandemic.