Back

Characterising Long Covid: a living systematic review update with controlled studies

Michelen, M.; Sigfrid, L.; Kartsonaki, C.; Shemilt, I.; Hastie, C.; O'Hara, M. E.; Suett, J. C.; Stelson, E. A.; Bugaeva, P.; Dahmash, D.; Rigby, I.; Munblit, D.; Harriss, E.; Burls, A.; Cheng, V.; Scott, J. T.; Carson, G.; Olliaro, P. L.; Stavropoulou, C.

2022-08-30 infectious diseases
10.1101/2022.08.29.22279338 medRxiv
Show abstract

StatementThe authors have withdrawn their manuscript owing to a reporting error that we have identified in the conduct of this updated systematic review. On 7th March 2022 we made changes to the original protocol published in F1000, to highlight the need to keep up with the evolving landscape of Covid-19 and Covid-19 research and focus on comparative, controlled studies. The revisions included also changes in data extraction and analysis that would be needed to do so. These changes were done in consultation with members of the Long COVID Support Group. However, we omitted to publicize these changes in an accessible, modified version of our protocol, and/or by updating our preregistered PROSPERO record (CRD42020211131) and F1000 publication, prior to commencing and completing these stages of the process. In this manuscript, which reports the revised analyses, we have consequently also omitted to explicitly describe these changes as important deviations from the published protocol. Additionally, following rapid, sustained, and continuous growth in the conduct and reporting of eligible primary studies, it has become clear that, with or without the changes described above, we do not have sufficient capacity or resources to transition our baseline systematic review to being regularly updated using a living systematic review approach. Therefore, the authors do not wish this work to be cited as reference for the project. If you have any questions, please contact the corresponding author.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.