Methodology
How research is found, checked and summarised
Last reviewed .
Where we look
An automated job runs once a day and queries official, public research databases through their published interfaces:
- PubMed (US National Library of Medicine) for peer-reviewed journal articles;
- Europe PMC for preprints, which are labelled as not yet peer reviewed;
- ClinicalTrials.gov for new trial registrations and newly posted results.
Editors also check guidelines, regulators (such as the US FDA and the European Medicines Agency) and ALS/MND organisations by hand. Search terms are "amyotrophic lateral sclerosis", "motor neuron disease" and "motor neurone disease", with extra priority for C9orf72, flail arm and lower motor neuron topics.
Last completed research check: .
How items are screened
- Duplicates are removed using DOI, PubMed ID, PubMed Central ID, trial ID, web address and title.
- Case reports, editorials, letters, errata and retracted papers are excluded.
- Guidelines, systematic reviews, randomised trials and large studies are prioritised. Animal and cell studies can be included but are always labelled as such.
- Previously published sources are re-checked against PubMed for retractions and expressions of concern.
Not every search result is published. Many records are not relevant, are too preliminary, or repeat known findings. On days with nothing that qualifies, nothing is posted.
How summaries are written and reviewed
A draft is created for each candidate. Drafts may be written with the help of an AI language model, which is given only the paper's public record (title, abstract and registry details). Every draft is read and checked against the original by a human editor before it can be published. Editors correct, rewrite or reject drafts.
Summaries are written in our own words for a general audience. Each one states the study type, separates what was found from what it might mean, includes a "what this does not show" note, and links to the original.
Dates and labels
Each post shows the date the study or announcement was published and, separately, the date it was added to this website. Evidence labels show where a claim comes from: peer-reviewed journal, preprint, trial registry, guideline, official regulator update, or company claim.
Corrections
If a source is corrected, retracted or superseded, the post is marked clearly and its editorial history is kept. To report an error, please use the contact page.
Limitations
- Only open, public records are searched. Paywalled full texts are not copied; summaries may rely on abstracts when the full paper is not openly available, and this can miss detail.
- Conference abstracts and non-English papers are mostly not covered.
- The automated check depends on external services; if one is unavailable, that run reports it, and the "last research check" time reflects only completed runs.
- This is a small, independent project, not a clinical service or an official body.