New ‘living search’ needed for global Covid-19 research

Research on Covid-19 is at risk of being ignored because academic publishing indexing and search functions are out-of-date, researchers from Goldsmiths, University of London and the University of Nottingham say.

Writing in the Journal of Clinical Epidemiology, Dr Tony Russell-Rose and Dr Farhad Shokraneh present the first “living search” strategy, designed to overcome some of the difficulties faced when searching for Covid-19 studies.

When the first case was reported in November 2019 neither the virus nor the disease it caused had a name, resulting in researchers in different countries referring to it using their own terms – from ‘Wuhan Pneumonia’ to ‘Coronavirus 2019’. A variety of terms are still in use.

The current Covid-19 open research dataset contains some 45,000 articles on the subject, published by researchers around the world, including scientific articles and those related to historic outbreaks of other coronaviruses and epidemics. Managing and curating the collection is a complex task.

Search systems are biased against publications written in languages other than English, with non-English papers harder to find and at greater risk of being under-valued, Dr Russell-Rose and Dr Shokraneh say.

Their letter to the Editor also notes that criticism is ongoing against the traditional long process of publishing peer-reviewed papers and delays in indexing them. At the time of writing, a quarter of studies on Covid-19 had not been published in a journal or indexed but may be searchable through separate ‘pre-print’ servers. 

Dr Russell-Rose and Dr Shokraneh explain: “Librarians and information specialists are frequently asked to design and run searches for literature on Covid-19.

“Apart from the fact that designing search strategies can be subjective and strategies may vary from one search expert to another, terminology is changing rapidly and search strategies that were accurate when first published can become unreliable and unable to retrieve relevant literature in a consistent manner.”

The reported terms used by researchers and the indexing terms used by indexing bodies have been insufficient for Covid-19, the authors argue, suggesting that “a standardized and up-to-date way to find literature on a particular concept” is needed. 

Dr Russell-Rose, Senior Lecturer in Computer Science at Goldsmiths, and Dr Shokraneh, Information Scientist at the School of Medicine, Nottingham, offer what may be the first living search strategy with live search results using the 2Dsearch platform, which was designed to solve complex search problems. 

Their search strategy is being updated and curated as the concept develops and as new terminology appears in MeSH – the international thesaurus of medical subject vocabulary produced by the National Library of Medicine in the US.

“The need for living search strategies has always been there, but Covid-19 has brought this need into sharper focus,” the authors write.

“To date, there has been no open platform to develop living search strategies and search experts have become accustomed to writing searches using word processing programs and sharing them in text documents (e.g. a spreadsheet or PDF) with all the limitations this implies.” 

Search experts such as librarians and other information professionals are now openly designing and sharing search strategies in different formats in an effort to share the latest research findings. 

“This is the time for both academic publishers and indexing bodies to consider whether they are letting science down by not being more inclusive,” the authors conclude.  

‘Lessons from COVID-19 to Future Evidence Synthesis Efforts: First Living Search Strategy and Out of Date Scientific Publishing and Indexing Industry’ by Farhad Shokraneh and Tony Rusell-Rose was published in the Journal of Clinical Epidemiology on 3 May 2020: https://doi.org/10.1016/j.jclinepi.2020.04.014