pubmed-database

Direct PubMed and NCBI E-utilities search workflows for biomedical literature, intended for tasks that need MEDLINE and life-sciences sources rather than general web search. It guides query construction by splitting a research question into concepts and combining them with Boolean operators, then refining precision with PubMed field tags such as [ti] (title), [ab] (abstract), [tiab] (title or abstract), [au] (author), [ta] (journal abbreviation), [mh] (MeSH term), [majr] (major MeSH topic), [pt] (publication type), [dp] (date of publication), and [la] (language). It explains how to prefer controlled-vocabulary MeSH terms when a concept has a stable term, pair them with free-text synonyms for newer or variable terminology, and attach subheadings before the field tag, for example diabetes mellitus, type 2/drug therapy[mh]. Filters cover publication types (clinical trial, meta-analysis, randomized controlled trial, review, systematic review, guideline), date ranges, and availability (free full text, hasabstract). For repeatable, API-backed work it documents the E-utilities pipeline — esearch to return PMIDs, esummary for lightweight metadata, efetch for abstracts or full records, and elink for related articles — with a Python example that reads an email and API key from environment variables, calls raise_for_status(), and sleeps to respect rate limits; for large batches it recommends the NCBI history server (usehistory, WebEnv, query_key) over long PMID lists in URLs. It emphasizes search discipline suited to systematic reviews: logging the exact search string, database, date searched, filters, and result count in a reproducible table, plus a review checklist covering valid field tags, MeSH-plus-synonym pairing, explicit date ranges, environment-loaded API keys, HTTP error handling, and rate-limit compliance. Use it for MeSH queries, PMID lookup, citation retrieval, and systematic-review search passes.