How Scienda works
What lands in your inbox, and what each setting changes. Everything below is adjustable at any time from Categories and Profile.
The digest
Scienda watches PubMed for you across 24 specialties. Once a day it sends you a single email listing what is new in the categories you follow — title, journal, date, article type, a journal-quality indicator and a link to PubMed. It selects and routes; it does not host the articles, and it never contains the abstract or the full text.
A quiet day means no email, unless you ask to be told there was nothing (Empty digest emails, in Profile).
What the service does and does not do is set out in the Terms of Service, which prevail over this page.
Journal quality
Every journal is ranked within its own field — a cardiology journal is compared with cardiology journals, not with Nature. The badge you see reflects that rank:
- Elite is a short list of major general journals (NEJM, Lancet, JAMA, Nature…). They always come through, whatever your other settings.
- Unranked does not mean “poor”. It means not measured — a new journal, or one too small to rank fairly. Unranked journals are never dropped on that ground alone.
- The measure comes from OpenAlex, an open bibliographic database, and is refreshed about once a year.
Minimum quality
Per category, you set the lowest badge you are willing to receive:
- Elite
- Top 5%
- Top 10%
- Top 25%
- Top 50%
- Any ranked
Any ranked is the widest setting: every quality tier, from the top down to the bottom of the scale. Start there if you would rather see too much than miss something, then tighten.
Journals we could not rank are never included — not even at Any ranked. A ranking we do not have is one we will not imply: a handful of titles are absent from our reference measurement (too new, too small, or not matched in the source data), and we would rather leave them out than show you a quality badge that means nothing.
Article types
You receive every type by default. Unchecking one narrows a category; unchecking them all would silence it, which is why the interface will not let you.
- Basic / translational research
- Bench and translational work. Inferred from the article itself, not from a PubMed tag — there is none for it.
- Clinical trial
- Trials, including randomised ones. Matches in PubMed: Clinical Trial, Randomized Controlled Trial
- Guidelines
- Practice guidelines. Matches in PubMed: Practice Guideline
- Meta-analysis / Systematic review
- Literature syntheses with an explicit method. Matches in PubMed: Meta-Analysis, Systematic Review
- Narrative review
- Classic review articles. PubMed often tags systematic reviews as “Review” too, so this box and the one above overlap in part. Matches in PubMed: Review
- Observational study
- Cohorts, case-control, comparative and multicentre studies. Matches in PubMed: Comparative Study, Multicenter Study, Observational Study
Articles published ahead of print often reach PubMed before it has tagged them — sometimes weeks before. Rather than hide them until the tag lands, Scienda infers the type from the article itself, so recent work still reaches you under the right heading. An inference is not a certainty: an occasional article filed under the wrong type is the price of not waiting.
“My specialty's journals only”
Off by default. Turned on, a category keeps only articles published in journals of your specialty and its closest neighbours — an intensive-care journal still counts for respiratory medicine. Major general journals always come through regardless.
The switch at the top of Categories is a bulk command, not a separate setting: it writes the box on every category at once, and you can then adjust them one by one. That is why it shows as “on” only when all of your active categories are.
“Relevance to confirm”
At the bottom of the digest you may find a section with this name. It holds articles that passed your filters but that our relevance model suspects are off-topic for the category they landed in.
It is a flag, never a deletion: the articles are still there, moved to the end rather than removed. When the model is wrong — it sometimes is — you lose nothing but a scroll. An empty section simply means nothing was flagged that day.
Frequency
- daily
- — One email a day, when there is something to report.
- weekly
- — A single email on Monday, covering the whole week.
There is no delivery-time setting: no specific hour is guaranteed. Picking your own time — and time zone — is on the roadmap.
Your account
- Stopping the emails and closing your subscription are two different things. The unsubscribe link at the bottom of a digest stops the emails and leaves your account, its settings and its history untouched — you can start them again from Profile whenever you like.
- To end the contract itself, use Cancel your contract.
- Anything not answered here: [email protected].