Network Graph Representation of COVID-19 Scientific Publications to Aid Knowledge Discovery (preprint)

IntroductionNumerous scientific journal articles have been rapidly published related to COVID-19 making navigation and understanding of relationships difficult. MethodsA graph network was constructed from the publicly available CORD-19 database of COVID-19-related publications using an engine leveraging medical knowledgebases to identify discrete medical concepts and an open source tool (Gephi) used to visualise the network. ResultsThe network shows connections between disease, medication and procedures identified from title and abstracts of 195,958 COVID-19 related publications (CORD-19 Dataset). Connections between terms with few publications, those unconnected to the main network and those irrelevant were not displayed. Nodes were coloured by knowledgebase and node size related to the number of publications containing the term. The dataset and visualisations made publicly accessible via a webtool. ConclusionKnowledge management approaches (text mining and graph networks) can effectively allow rapid navigation and exploration of entity interrelationships to improve understanding of diseases such as COVID-19.

Testing for coronavirus (SARS-CoV-2) infection in populations with low infection prevalence: the largely ignored problem of false positives and the value of repeat testing (preprint)

Background: Calls are increasing for widespread SARS-CoV-2 infection testing of people from populations with a very low prevalence of infection. We quantified the impact of less than perfect diagnostic test accuracy on populations, and on individuals, in low prevalence settings, focusing on false positives and the role of confirmatory testing. Methods: We developed a simple, interactive tool to assess the impact of different combinations of test sensitivity, specificity and infection prevalence in a notional population of 100,000. We derived numbers of true positives, true negatives, false positives and false negatives, positive predictive value (PPV, the percentage of test positives that are true positives) and overall test accuracy for three testing strategies: (1) single test for all; (2) add repeat testing in test positives; (3) add further repeat testing in those with discrepant results. We also assessed the impact on test results for individuals having one, two or three tests under these three strategies. Results: With sensitivity of 80%, infection prevalence of 1 in 2,000, and specificity 99.9% on all tests, PPV in the tested population of 100,000 will be only 29% with one test, increasing to >99.5% (100% when rounded to the nearest %) with repeat testing in strategies 2 or 3. More realistically, if specificity is 95% for the first and 99.9% for subsequent tests, single test PPV will be only 1%, increasing to 86% with repeat testing in strategy 2, or 79% with strategy 3 (albeit with 6 fewer false negatives than strategy 2). In the whole population, or in particular individuals, PPV increases as infection becomes more common in the population but falls to unacceptably low levels with lower test specificity. Conclusion: To avoid multiple unnecessary restrictions on whole populations, and in particular individuals, from widespread population testing for SARS-CoV-2, the crucial roles of extremely high test specificity and of confirmatory testing must be fully appreciated and incorporated into policy decisions.

Study Data Element Mapping: Feasibility of Defining Common Data Elements Across COVID-19 Studies (preprint)

Background: Numerous clinical studies are now underway investigating aspects of COVID-19. The aim of this study was to identify a selection of national and/or multicentre clinical COVID-19 studies in the United Kingdom to examine the feasibility and outcomes of documenting the most frequent data elements common across studies to rapidly inform future study design and demonstrate proof-of-concept for further subject-specific study data element mapping to improve research data management. Methods: 25 COVID-19 studies were included. For each, information regarding the specific data elements being collected was recorded. Data elements collated were arbitrarily divided into categories for ease of visualisation. Elements which were most frequently and consistently recorded across studies are presented in relation to their relative commonality. Results: Across the 25 studies, 261 data elements were recorded in total. The most frequently recorded 100 data elements were identified across all studies and are presented with relative frequencies. Categories with the largest numbers of common elements included demographics, admission criteria, medical history and investigations. Mortality and need for specific respiratory support were the most common outcome measures, but with specific studies including a range of other outcome measures. Conclusion: The findings of this study have demonstrated that it is feasible to collate specific data elements recorded across a range of studies investigating a specific clinical condition in order to identify those elements which are most common among studies. These data may be of value for those establishing new studies and to allow researchers to rapidly identify studies collecting data of potential use hence minimising duplication and increasing data re-use and interoperability