eduKateSG · Why Science?
A precise biomedical search begins by knowing which database layer supplied the record
Read PubMed through MEDLINE, publisher and archive routes, MeSH indexing, publication types, filters, citations and full-text links without confusing discovery with appraisal.
Biomedical questions often begin in PubMed because it brings together citations and abstracts from MEDLINE, life-science journals and online books, with links that may lead to PubMed Central or publisher full text. The database is broad, but its records do not all arrive through one route or carry identical indexing.
MEDLINE is the National Library of Medicine’s curated bibliographic database and a major component of PubMed. MEDLINE records receive NLM indexing, including Medical Subject Headings, while other PubMed records can come from additional documented sources and may not yet have complete MEDLINE indexing.
Did you know? A PubMed result can display keywords, publication types or automated terms even when a human-assigned MeSH set is absent. A careful searcher checks the record’s indexing information before assuming that every subject label was reviewed in the same way.
This guide treats PubMed as a powerful discovery system and MEDLINE as a defined indexed subset. Neither a PubMed identifier nor a database filter proves that a study is well designed, current, applicable to a patient or suitable for a school claim.
Chapter index
Chapters 1–9: PubMed sources, MEDLINE and record status
- 1. PubMed is larger than MEDLINE
- 2. The PMID identifies a PubMed record
- 3. A PMCID points to PubMed Central
- 4. MEDLINE journal selection is documented
- 5. MeSH is a controlled vocabulary
- 6. Indexing can arrive after publication
- 7. Automatic term mapping expands a query
- 8. Field tags make intent explicit
- 9. Boolean logic controls combinations
Chapters 10–17: MeSH, publication types, filters and evidence limits
- 10. Publication types describe study forms
- 11. Filters can depend on indexed data
- 12. Date fields answer different questions
- 13. Similar articles are algorithmic suggestions
- 14. Citation links have coverage limits
- 15. Full-text availability is a route outward
- 16. Saved search methods need a receipt
- 17. Biomedical information skills support careers
PubMed is a citation and abstract database whose records come through several documented routes. PubMed contains MEDLINE records plus citations from other NLM-supported routes. Calling every PubMed record MEDLINE-indexed hides meaningful differences in selection and indexing. This makes biomedical searching both faster and more honest.
Try this
Evidence question: Which component supplied this citation? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Inspect the record source and indexing status. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Keep the identifier, retrieval date and source page in the note so the route can be repeated.
Evidence checkpoint
Record the evidence object for pubmed is larger than medline, its identifier or version context, and one limit that must remain visible.
MEDLINE is a curated component of PubMed, not a synonym for the whole database. A PubMed identifier, or PMID, names the database record. It is distinct from a DOI, PMCID and other identifiers that may connect to related systems. It also keeps database coverage separate from the quality of an individual study.
Try this
Evidence question: Which identifier scheme does this number use? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Store PMID, DOI and PMCID in separate fields. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. If a field is missing, write “not shown” and follow the official documentation rather than inventing a value.
Evidence checkpoint
Record the evidence object for the pmid identifies a pubmed record, its identifier or version context, and one limit that must remain visible.
MeSH headings and publication types support precise searching when their indexing status is understood. The PMCID identifies an article in the full-text archive. A PMID can exist without a PMCID, and a full-text link can also lead to a publisher. Learners can build reproducible queries without treating a filter as a truth machine.
Try this
Evidence question: Where is the complete text hosted? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Follow the labelled route and record the destination. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Before comparing two records, check that their object types, methods, versions and units really align.
Evidence checkpoint
Record the evidence object for a pmcid points to pubmed central, its identifier or version context, and one limit that must remain visible.
Full-text links are routes outward; a PubMed record is not automatically the complete article. NLM applies scientific, editorial, production and related considerations when selecting journals for MEDLINE. Selection is journal-level infrastructure, not a guarantee for each result. A careful search receipt preserves terms, dates, filters and the record source.
Try this
Evidence question: Is the journal currently indexed for MEDLINE? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Check the NLM Catalog and the record date. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Finish by naming one supported statement and one stronger statement that still needs another evidence layer.
Evidence checkpoint
Record the evidence object for medline journal selection is documented, its identifier or version context, and one limit that must remain visible.
PubMed is a citation and abstract database whose records come through several documented routes. Medical Subject Headings organise biomedical concepts with preferred terms, entry terms and hierarchy. This supports searches that are broader than one author’s wording. This makes biomedical searching both faster and more honest.
Try this
Evidence question: Which MeSH concept represents the question? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Open the MeSH record and inspect narrower and broader terms. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Keep the identifier, retrieval date and source page in the note so the route can be repeated.
Evidence checkpoint
Record the evidence object for mesh is a controlled vocabulary, its identifier or version context, and one limit that must remain visible.
MEDLINE is a curated component of PubMed, not a synonym for the whole database. New PubMed records may appear before MEDLINE indexing is complete. Searching only MeSH can therefore miss recent or non-indexed records. It also keeps database coverage separate from the quality of an individual study.
Try this
Evidence question: Could an unindexed recent paper be absent? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Combine controlled vocabulary with title-and-abstract terms. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. If a field is missing, write “not shown” and follow the official documentation rather than inventing a value.
Evidence checkpoint
Record the evidence object for indexing can arrive after publication, its identifier or version context, and one limit that must remain visible.
MeSH headings and publication types support precise searching when their indexing status is understood. PubMed can map untagged words to subjects, journals, authors and other indexes. Helpful expansion can also make a search less obvious. Learners can build reproducible queries without treating a filter as a truth machine.
Try this
Evidence question: How did PubMed interpret this phrase? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Open Search Details and save the translated query. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Before comparing two records, check that their object types, methods, versions and units really align.
Evidence checkpoint
Record the evidence object for automatic term mapping expands a query, its identifier or version context, and one limit that must remain visible.
Full-text links are routes outward; a PubMed record is not automatically the complete article. Tags such as title/abstract, author, journal and MeSH terms limit where PubMed looks. A fielded query is easier to explain but can reduce recall. A careful search receipt preserves terms, dates, filters and the record source.
Try this
Evidence question: Which field should contain this concept? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Test the untagged and fielded versions. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Finish by naming one supported statement and one stronger statement that still needs another evidence layer.
Evidence checkpoint
Record the evidence object for field tags make intent explicit, its identifier or version context, and one limit that must remain visible.
PubMed is a citation and abstract database whose records come through several documented routes. AND narrows to shared conditions, OR joins alternatives and NOT excludes records. Parentheses preserve the intended grouping of concepts. This makes biomedical searching both faster and more honest.
Try this
Evidence question: Does the query match the question’s logic? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Read it aloud as sets before running it. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Keep the identifier, retrieval date and source page in the note so the route can be repeated.
Evidence checkpoint
Record the evidence object for boolean logic controls combinations, its identifier or version context, and one limit that must remain visible.
MEDLINE is a curated component of PubMed, not a synonym for the whole database. Publication-type indexing can identify clinical trials, reviews, editorials and many other forms. Labels can be absent, updated or broader than a user expects. It also keeps database coverage separate from the quality of an individual study.
Try this
Evidence question: Which publication type is present and indexed? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Verify the record and read the methods. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. If a field is missing, write “not shown” and follow the official documentation rather than inventing a value.
Evidence checkpoint
Record the evidence object for publication types describe study forms, its identifier or version context, and one limit that must remain visible.
MeSH headings and publication types support precise searching when their indexing status is understood. Article type, age, language and other filters may rely on publication type or MeSH-related fields. Applying them can exclude recent records not yet indexed. Learners can build reproducible queries without treating a filter as a truth machine.
Try this
Evidence question: Which filter requires completed indexing? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Document the filter and run a sensitivity search without it. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Before comparing two records, check that their object types, methods, versions and units really align.
Evidence checkpoint
Record the evidence object for filters can depend on indexed data, its identifier or version context, and one limit that must remain visible.
Full-text links are routes outward; a PubMed record is not automatically the complete article. Publication date, electronic publication, entry date and indexing-related dates can differ. The chosen field changes a timeline search. A careful search receipt preserves terms, dates, filters and the record source.
Try this
Evidence question: Which date defines the inclusion window? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Name the date field in the protocol. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Finish by naming one supported statement and one stronger statement that still needs another evidence layer.
Evidence checkpoint
Record the evidence object for date fields answer different questions, its identifier or version context, and one limit that must remain visible.
CHAPTER 13 OF 17 · Appraise the paper
13. Similar articles are algorithmic suggestions
PubMed is a citation and abstract database whose records come through several documented routes. Related-record features can help widen reading, but similarity ranking is not a systematic-review method or a validity score. This makes biomedical searching both faster and more honest.
Try this
Evidence question: Which seed record produced this suggestion? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Use suggestions to find candidates, then screen them consistently. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Keep the identifier, retrieval date and source page in the note so the route can be repeated.
Evidence checkpoint
Record the evidence object for similar articles are algorithmic suggestions, its identifier or version context, and one limit that must remain visible.
MEDLINE is a curated component of PubMed, not a synonym for the whole database. Cited-by and reference connections depend on available data and matching. A low count can reflect recency or coverage rather than importance. It also keeps database coverage separate from the quality of an individual study.
Try this
Evidence question: Which service and date produced the links? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Keep citation evidence separate from study appraisal. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. If a field is missing, write “not shown” and follow the official documentation rather than inventing a value.
Evidence checkpoint
Record the evidence object for citation links have coverage limits, its identifier or version context, and one limit that must remain visible.
MeSH headings and publication types support precise searching when their indexing status is understood. PubMed may link to free archive copies or publisher pages. Access status, manuscript version and licence still need checking at the destination. Learners can build reproducible queries without treating a filter as a truth machine.
Try this
Evidence question: Which copy did you actually read? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Record host, version, licence and access date. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Before comparing two records, check that their object types, methods, versions and units really align.
Evidence checkpoint
Record the evidence object for full-text availability is a route outward, its identifier or version context, and one limit that must remain visible.
Full-text links are routes outward; a PubMed record is not automatically the complete article. A reproducible search includes the complete query, database, filters, date and number of results. Screenshots alone can omit translated terms and settings. A careful search receipt preserves terms, dates, filters and the record source.
Try this
Evidence question: Could another reader rerun this search? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Export the query and write the exact search date. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Finish by naming one supported statement and one stronger statement that still needs another evidence layer.
Evidence checkpoint
Record the evidence object for saved search methods need a receipt, its identifier or version context, and one limit that must remain visible.
CHAPTER 17 OF 17 · Appraise the paper
17. Biomedical information skills support careers
PubMed is a citation and abstract database whose records come through several documented routes. Clinicians, librarians, systematic reviewers, researchers and health communicators use PubMed. Their responsibility is to connect precise retrieval with careful appraisal. This makes biomedical searching both faster and more honest.
Try this
Evidence question: Which decision needs more than an abstract? Answer in three parts: identify the visible field or file, explain how it bears on the question, and state the boundary of that evidence. Preserve the database’s own object names and accession rather than flattening everything into the word ‘data’.
Try this: Build a PubMed search and evidence receipt for one science question. Turn the result into a compact claim-evidence-reasoning note, then ask a partner to retrace it. Keep the identifier, retrieval date and source page in the note so the route can be repeated.
Evidence checkpoint
Record the evidence object for biomedical information skills support careers, its identifier or version context, and one limit that must remain visible.
Contents · Previous chapter · Continue at the Science Learning Hub
Search precisely, then read beyond the record
PubMed makes biomedical literature discoverable at scale, while MEDLINE contributes curated journal selection and structured indexing. The distinction helps learners understand why records, filters and subject headings can differ.
Save the query translation, field tags, MeSH terms, publication types, filters, date field, PMID, DOI or PMCID, result count and search date. Then open the full article, assess methods and limitations, and check corrections, retractions and later evidence.
| Reading object | What it can support | Question to keep visible |
|---|---|---|
| PubMed record | Citation and abstract entry | Which source and indexing status apply? |
| MEDLINE indexing | Curated NLM bibliographic layer | Which MeSH and publication types were assigned? |
| Identifier | PMID, DOI or PMCID | Which system and object does it name? |
| Search receipt | Query, filters and date | Can another reader reproduce the result set? |
Sources and next reading
Substantive resource claims in this article were checked against the primary pages below. A structured record is a route into evidence; it does not by itself prove a scientific conclusion. Recheck the live record, evidence labels and release context before using it in assessed work or research.
For a classroom enquiry, save a short research receipt containing the question, search terms, selected object, exact identifier, version, access date and the evidence boundary you observed. Add the file or page that supports each statement. This small routine turns browsing into accountable science learning and gives a teacher, classmate or future analyst a route they can independently follow.
Before sharing the conclusion, ask another reader to retrace the route without verbal hints. If they reach a different version, file or interpretation, improve the receipt and narrow the claim. Reproducibility begins with this practical check. It teaches careful reading, supports fair collaboration and keeps uncertainty visible while evidence moves from an online service into notes, analysis and a final explanation.
A useful final review separates five questions. What object was found? Which organisation or repository maintains it? What exact evidence was inspected? Which statement does that evidence support? What remains uncertain? Answer each question in one sentence, then connect every answer to a visible field, file or source page. This review catches common mistakes: citing a collection when a file supplied the value, treating the newest release as the analysed release, confusing permission with scientific suitability, or turning a discovery count into a conclusion. It also gives students a calm way to improve work. A missing detail becomes a next research step instead of an invitation to guess. When the route is complete, the final explanation can stay simple because the identifiers, dates, versions, limitations and supporting evidence are already organised underneath it.
Primary sources
- PubMed about — accessed 25 October 2026.
- PubMed User Guide — accessed 25 October 2026.
- About MEDLINE — accessed 25 October 2026.
- MEDLINE policies — accessed 25 October 2026.
- NLM Medical Subject Headings — accessed 25 October 2026.