Post COVID-19 – lessons from a living systematic literature search on treatment and rehabilitation
Background In 2021, the Swedish Agency for Health Technology Assessment and Assessment of Social Services (SBU) was commissioned by the Swedish Government to continuously evaluate the scientific evidence for treatment and rehabilitation of patients with Post COVID-19. A living systematic review was launched with regularly updated literature searches and the results published on the SBU website on a weekly basis. Objectives To describe the ad hoc methodology used for a continuous and systematic search for the Post COVID-19 literature. Methods To find the most recent published research studies, a comprehensive search in Medline was performed every week via alerts. Every month, five additional databases were searched. Also, reference lists and citations for relevant primary studies and reviews were screened. The searches were performed from April 2021 up to June 2022. The project group also continuously tracked COVID-19-specific resources. The Long COVID search filter in PubMed Clinical Queries was validated against our set of included articles. The sensitivity of the Long COVID search filter, mainly focused on the concept for Post COVID-19 and not including specific symptom categories, was quite low. Results A total of 24,729 references were screened; 536 articles were read in full text, and 19 were deemed relevant. In sum, 11 trials with moderate risk of bias were included. The living systematic review resulted in an evidence map with articles presented according to categories of symptoms. Methodological challenges were as follows: creating a flexible and rapid workflow while maintaining a robust and transparent methodology; keeping abreast of a growing plethora of COVID-19-specialized registers and tertiary databases; and new and evolving terminology, e.g., challenges in creating a comprehensive search strategy for the “Post” aspect of COVID-19. Conclusions As expected, not many published trials on Post COVID-19 were found up to June 2022; the research on long-term symptoms was probably still in pre-publication or the preprint phase. However, because this was the first living systematic review at SBU, useful lessons were made for future “living” projects. Patient, public, and/or healthcare consumer involvement: no.