#EvidenceSynthesis

2025-05-07

Hi #EvidenceSynthesis folks and #medlibs
You have about ten minutes to give feedback to the ESIC people

Link: qualtricsxm45bq73zl3.qualtrics

The Stage 3 reports are at evidencesynthesis.atlassian.ne

Since you don't have time to read them at this point, I will suggest some talking points:

1 The AI working group's suggestion of generic classifiers (to tag/retrieve/screen studies with a particular study design, population, setting, outcome, whatever) is one of the few ideas that are IMO both valuable and feasible

2 You can say that technical mechanisms will someday allow for an AI-led framework where human oversight is significantly reduced, yet evidence integrity remains uncompromised – but saying that doesn’t make people believe it.

3 Before we build xkcd.com/927/ in terms of literature databases, or ES data sharing standards, or ontologies -- let's reflect on the benefits of decentralization. Would the rest of the EBM world be worrying quite so much about US politics if not for the centrality of NCBI, NLM, PubMed, PMC, CDC, HICPAC, NIOSH, etc in global information networks?

2025-04-21

Someone explain to me -- I thought the whole point of considering using LLMs for #EvidenceSynthesis screening was to process large datasets quickly.
But in this paper
Ghossein, J., Hryciw, B. N., Ramsay, T., & Kyeremanteng, K. (2025). The AI Reviewer: Evaluating AI’s Role in Citation Screening for Streamlined Systematic Reviews. JMIR Formative Research, 9(1), e58366. doi.org/10.2196/58366
They had a human-labeled dataset of 1186 citations, and they only tested their LLMs on 121 of them -- all of the included studies and 9% of the excluded ones.

2025-04-04

This, but replace "writing code" with "writing sensitive and reproducible #EvidenceSynthesis searches"
And replace "a requirements document" with "a systematic review protocol"

David Chisnall (*Now with 50% more sarcasm!*)
@david_chisnall@infosec.exchange

The reason I get so annoyed about people pitching LLMs as a way to 'democratise programming' or as end-user programming tools is that they solve the wrong problem.

The hard part of programming is not writing code.  It's unambiguously expressing your problem and desired solution.  Imagine if LLMs were perfect programmers.  All you have to do is write a requirements document and they turn it into a working program.  Amazing, right?  Well, not if you've ever seen what most people write in a requirements document or seen the output when a team of good programmers works from a requirements document.
2025-04-03

What absolute perfect timing in light of the announcement that DOGE is cutting the size of ERIC significantly

Fitzgerald, S. R., Weaver, K. D., & Droog, A. (2025). Selecting a specialized education database for literature reviews and evidence synthesis projects. Research Synthesis Methods, 16(1), 30–41. doi.org/10.1017/rsm.2024.11

#EvidenceSynthesis #Education

Figure showing the overlap of Education Source, Education Database, Educator's Reference Complete, and ERIC. The first is much bigger and thus contains many more unique journals than the rest. The other three all contain more than 300 unique journals too.
Wolfgang Viechtbauerwviechtb@scholar.social
2025-03-19

1/3 In a recent paper, we described the use of location-scale models in meta-analysis to not only examine how moderators may be related to the size of the effect, but also how moderators may be related to the amount of heterogeneity among the effects: doi.org/10.1002/jrsm.1562

Such models can be fit with the metafor package in R: wviechtb.github.io/metafor/ref

#statistics #MetaAnalysis #EvidenceSynthesis #ResearchSynthesis #Rstats

Forest plot showing the results from 13 trials on the effectiveness of the BCG vaccine against tuberculosis. At the top, two normal distributions are shown for the predictive distributions of the effects for trial without and with random assignment.
Kate Nyhan is changing serverskdnyhan@social.esmarconf.org
2025-03-13

New paper out:
Rethlefsen et al (including me!) (2025). Improving peer review of systematic reviews and related review types by involving librarians and information specialists as methodological peer reviewers: a randomised controlled trial. BMJ evidence-based medicine, bmjebm-2024-113527. Advance online publication. doi.org/10.1136/bmjebm-2024-11
#medlibs #PeerReview #EvidenceSynthesis

Kate Nyhan is changing serverskdnyhan@social.esmarconf.org
2025-03-12

Can anyone explain to me the relationship between the "SHOW ME the evidence" paper published in JBI Evidence Implementation, Campbell Systematic Reviews, Clinical and Public Health Guidelines, Cochrane, and Environmental Evidence in November, and the "Evidence Synthesis Infrastructure Collaborative"?
#EvidenceSynthesis #SystematicReviews

Nathaniel D. Porterndporter@sciences.social
2025-03-11

#JobOpportunity at #VirginiaTech University #libraries. Join our growing #EvidenceSynthesis team as Evidence Synthesis Information Scientist. First review April 1. careers.pageuppeople.com/968/c

Wolfgang Viechtbauerwviechtb@scholar.social
2025-03-06

The Evidence Synthesis and Meta-Analysis in R Conference (ESMARConf) is back! It will be held June 11th to the 13th, 2025.

esmarconf.org/2025/

Recordings of the talks and workshops from previous years can be found here: esmarconf.org/recordings/ (and directly on you YouTube channel: youtube.com/@esmarconf).

#ESMARConf #MetaAnalysis #EvidenceSynthesis #RStats

Kate Nyhan is changing serverskdnyhan@social.esmarconf.org
2025-02-27

Wow, PROSPERO look very different and it sounds like it is very different.
This is a great idea IMO
#EvidenceSynthesis #medlibs

During the registration process, the PROSPERO system uses the information provided to identify any similar reviews that have previously been registered. If similar reviews are found, and the person registering the current review decides to continue with their registration, they are asked to provide the reasons why their new review is needed (although this is not mandatory). The reasons provided (including none) become part of the registration record.
Kate Nyhan is changing serverskdnyhan@social.esmarconf.org
2025-02-12

And here it is: where you can request free access to Policy Commons to search for #GreyLit for an #EvidenceSynthesis project:
docs.google.com/forms/d/e/1FAI

Kate Nyhan is changing serverskdnyhan@social.esmarconf.org
2025-02-11

OK I know that #EvidenceSynthesis is becoming a bigger part of the literature but I didn't realize the rate was climbing this steeply, or that it's already so high.
More than one in twenty articles in PLOS ONE have been systematic/scoping/rapid reviews, or their protocols, for two years and a bit? Seems surprising.

How many PLOS One articles are systematic/scoping/rapid reviews (protocols included), as a proportion of each year's records, by PubMed create-date year?
Starting in 2008 at a low level -- under half a percent -- but increasing steadily (except for 2018, when it was pretty flat for a year), until it's almost 7% for 2025 so far.
Kate Nyhan is changing serverskdnyhan@social.esmarconf.org
2025-02-05

From today's Policy Commons newsletter
#EvidenceSynthesis

Are you planning a systematic or scoping review?

Last year, we gave Paul Cairney and Claire Toomey from University of Stirling complimentary access to Policy Commons to help them with their systematic review on systems leadership. Of the texts that met their inclusion criteria, we were pleased to read that they found almost as many relevant items in Policy Commons as they did in Web of Science. If you’re planning a systematic review and would like complimentary access to Policy Commons to find grey literature, please apply using this form.
Alfredo Sánchez-TójarASanchez_Tojar@ecoevo.social
2024-10-07

Today is the day. We are hosting the workshop at Bielefeld University

"The #EvidenceSynthesis Toolkit: Perspectives from Philosophy and Biology"

Unmatchable list of speakers

More at: shorturl.at/4OP1V

Supported by DFG, NC3, JICE InChangE #systematicreview #metaanalysis

Kate Nyhan is changing serverskdnyhan@social.esmarconf.org
2024-09-26

The first gen AI study that I really want to read (maybe even do):
1 recruit participants WITH DIFFERENT LEVELS of experience in evidence synthesis and searching
2 show them chatbot output for prompts like "pretend you're a medical librarian and create a SR search strategy to use in PubMed for my research question," using whatever the state of the art guidelines are for "prompt engineering"
3A ask the participants to evaluate the chatbot output

#medlibs #EvidenceSynthesis

Kate Nyhan is changing serverskdnyhan@social.esmarconf.org
2024-09-04

For 1000 *protocols* of *living* systematic reviews, how many reviews get published, and how many continue to be updated on the original schedule? Anyone have an estimate?
#medlibs #EvidenceSynthesis

carinavrcarinavr
2024-09-03

The amazing Neal Haddaway, and colleagues, are hosting a five-day Evidence Synthesis for Social Transformations International School, 25-29th November 2024 at the University of Almeria in Spain.
It focuses on methodology training in the context of research into social transformations.
For more info: evidencesynthesisschool.github

Kate Nyhan is changing serverskdnyhan@social.esmarconf.org
2024-09-03

Have you ever seen an SR that did a sensitivity analysis based on only the OA included papers? Or even pointed out how many of the included papers are OA? Would love to read examples, don't care if they included only version of record OA or not.
#EvidenceSynthesis #SystematicReviews #OA #OpenAccess

Wolfgang Viechtbauerwviechtb@scholar.social
2024-09-03

Last week, I gave a talk about model selection in complex meta-analyses. It covers some general model considerations and gives several examples of multilevel and multivariate meta-analyses. You can find the slides for this presentation (and the code for the examples) here: wvbauer.com/doku.php/presentat

#MetaAnalysis #EvidenceSynthesis #RStats

Sergio Graziosi - has movedGraziosiSergio@social.esmarconf.org
2024-08-15

Anyone attending #GES2024, as well as anyone interested in #EvidenceSynthesis who will be in Prague on September 9, might want to attend the free "Introduction to EPPI-Reviewer" workshop.
More info and registration link are here:
globalevidencesummit.org/satel

Client Info

Server: https://mastodon.social
Version: 2025.04
Repository: https://github.com/cyevgeniy/lmst