How AnteSum works, exactly

The front page makes confident claims. This page backs them up: what we cover, what we fetch, what the AI does and doesn't do. The recipes themselves — the prompts, the rules, the tuning — stay ours.

What we cover

Each specialty edition monitors a hand-curated list of peer-reviewed journals: 61 in anaesthesia and pain, 52 in orthopaedics. Both are revised as journals launch, merge or fold.

Anaesthesia spans the major general and subspecialty titles — British Journal of Anaesthesia, Anesthesiology, Anaesthesia, Anesthesia & Analgesia, European Journal of Anaesthesiology, Regional Anesthesia & Pain Medicine — alongside the main pain, obstetric, paediatric, cardiothoracic and regional journals. Orthopaedics spans The Bone & Joint Journal, JBJS, Clinical Orthopaedics and Related Research, AJSM, Spine, The Journal of Arthroplasty and the main trauma, hand, foot & ankle, shoulder & elbow and spine journals.

Want the full list, or think we're missing something you rely on? Email [email protected].

What we fetch

We check every journal on the list weekly and take each newly published journal article.

In: original research, reviews, guidelines, editorials, case reports and correspondence — online-first or in-issue, each labelled by section so you can tell a trial from an editorial at a glance.

Out: front matter (contents pages, boards, indexes) and anything not published as an article.

Never: preprints. Peer-reviewed journals only.

So when the front page says "the new papers", the exact claim is: every journal article these titles publish, minus front matter. Our coverage depends on publishers' own publication records — very good for these journals, but not something we can independently guarantee.

Timing

Papers reach AnteSum from the point they first appear online, often weeks or months before print — that's why they reach you early. A paper belongs permanently to the month AnteSum first saw it; publisher dates are shown for reference but never move it. Every monthly archive stays exactly as you saw it, with no duplicates.

How papers are classified

Two steps, one AI:

• Section (Reviews, Original Articles, Guidelines, Editorials, Case Reports, Correspondence) — determined without AI, from how the article is already categorised at source.
• Subspecialty (obstetric, regional, paediatric…) — assigned by AI, within guardrails we set.

Classification happens once, when a paper is first seen, and doesn't silently change. Like any classifier it occasionally mislabels; we correct deliberately, never by re-running old months.

What the AI reads, and writes

Each subspecialty's monthly digest is written by a frontier AI model and refreshed weekly as new papers land. Its input is the month's papers — titles, journal names, section labels. It does not read full text.

Where we hold an abstract, the model sees only what the study set out to do — never its results or conclusions, which are removed before it reads anything. What a study found stays yours to read, and abstract availability never affects a paper's inclusion or prominence.

Is there human review?

No. AnteSum is built and run by a practising doctor, but each weekly update is generated and published automatically — no one signs off each edition first. That's exactly why the digest is constrained the way it is: it may only describe what's been published, never report results or draw clinical conclusions. Spot an error or a mislabelled paper? Email [email protected] and we'll fix it.

What AnteSum does not claim

• It doesn't assess study quality, risk of bias or strength of evidence. Inclusion isn't endorsement.
• It doesn't give clinical advice. Nothing here should change your practice on its own — read the paper, apply your judgement.
• It isn't a systematic review or a complete record of the literature.
• It doesn't replace reading. It exists to make sure the papers worth your time actually reach you.