Living evidence syntheses for [LC] therapeutics: combining rigorous protocols to build efficiency while maintaining rigour, 2026, Atkins, Glasziou+

[SNT Gatchaman]

Living evidence syntheses for long COVID therapeutics: combining rigorous protocols to build efficiency while maintaining rigour

Spoiler: Authors

Atkins, Tiffany; Glasziou, Paul; Chakraborty, Samantha; Turner, Tari; Rada, Gabriel; Byambasuren, Oyungerel

BACKGROUND
Given the rapidly changing evidence, the creation and maintenance of a living systematic review database of therapeutics for long COVID is an ideal and necessary approach considering the rapidly changing evidence that continues to be identified. This paper describes methods and results of a collaboration between three teams who produced a living literature review on long COVID therapeutics—Australian Living Evidence Collaboration (ALEC), Bond University, and Epistemonikos COVID-19 L.OVE (Living Overview of Evidence) database.

METHODS
We took a collaborative and iterative approach to analyse the commonalities and differences between each project and develop an agreed comprehensive, collective approach. A plan for ongoing (monthly) updates and dissemination was built.

RESULTS
Despite minor differences, there was also a clear alignment of goals between the three teams. Differences in search strategy, search methods and screening criteria were identified, investigated, and resolved. A comparison of overlaps helped establish a common collaborative approach. A combined library of 218 randomised controlled trials and 56 systematic reviews was created which led to the optimised search strategy. The combined 218 RCT library covered 20 different treatment categories of which 14 were pharmacological and 6 were non-pharmacological. Further refinements and collaborations led to a transformed initial database library of 102 randomised controlled trials as of June 2024 before the team commenced monthly updates.

CONCLUSIONS
Despite initial differences, a comprehensive search strategy based on the collaboration of the three teams was developed. Ongoing monthly updates were initiated and are now planned for well into the future to make continual and rapid updates to the library of evidence surrounding therapeutics for long COVID. Where global public health is concerned, it is valuable to review and refine processes in the early stages, so that they can be reliable. We recommend open collaboration to achieve the goal of creating accessible, efficient, and reliable evidence syntheses.

Web | DOI | PDF | Systematic Reviews | Open Access

 

[SNT Gatchaman]

Last week we had Exploring duplication in reviews of Long COVID: 2020–2023. Now we seem to have publications that try to combine those duplicated efforts.

"Must have Moar Evidence-Based Medicine"

The Library of studies is currently not publicly available but plans are underway to make it available.

 

[Trish]

This looks like a Cochrane type of exercise. I notice Glasziou is involved, which doesn't bode well.

 

[rvallee]

Given the rapidly changing evidence

The what now? There is no such thing, in fact there is yet no evidence of anything, and it's been 6 years. There have been plenty of baseless claims, but that's a different issue entirely, and far more problematic. What planet are these people on? This is a total navel-gazing experiment. They essentially describe what they are doing, and don't even seem to consider problematic the fact that this entire approach has yielded nothing of value for anyone. I have no idea who this is for or what purpose it could possibly serve.

The selection of trials is extreme cherry-picking, there have been far more than the 56 they selected, and it's most notable that over half (31) of those were for physical/exercise therapy, with 15 more in the dubious category, treatments that have no chance of ever being useful.

Our case study illustrates the complexity in trying to develop and maintain a database of all trials in a specific condition.

They call this evidence-based medicine. These people are experts in the field and they can't get through the tutorial level. This is all super basic stuff that has long been solved and they're still struggling with this like some drunk trying to put pants on their head. They pretty much keep identifying that this whole approach is worthless, but they can never accept it, hence why it is worthless in the first place.