How do I run a repeatable search to fill missing papers?

Run a repeatable search by splitting your topic into concept blocks, running field-tagged queries in PubMed, saving the search, repeating the same blocks in Europe PMC, and logging each query, source, date, and result count so the set can be rerun.

1. Build the concept blocks. Split your topic into its main concepts and list synonyms for each. In PubMed, attach field tags so the search engine knows which record field to match; the PubMed User Guide documents field tags, phrase searching, filters, and Advanced Search. Wrap multi-word phrases in quotation marks to keep them together. Expected result: a query string you can copy exactly into a log.
2. Run, tune, and save the PubMed search. Combine the blocks, then use filters to control recall (how many records you retrieve) and precision (how many are relevant). If the set is too large, tighten phrases or add filters; if it is too small, add synonyms. Save the search so the identical strategy can be rerun later and reported. Expected result: a saved strategy plus a baseline result count.
3. Repeat the concept blocks in Europe PMC. Europe PMC supports keyword and fielded searches across life-science literature, grants, preprints, citations, references, and links to available full text. This step can surface preprints, grant records, and citation links that a PubMed-only search may not return. Expected result: a second count and a list of records unique to this source.
4. Extend from seed papers in Google Scholar. Take two or three clearly relevant papers as seeds. Use Google Scholar's related-articles and cited-by links to move outward from each seed, and use quotation marks around titles for title searches; Google Scholar's help pages document these features along with date sorting and alerts. Expected result: candidate papers that cite or are related to your seeds.
5. Log every step. Record each query string, the source, the date run, and the result count. This log lets you rerun the strategy, show which sources added which papers, and describe the search in a methods note.

What to expect
The saved PubMed search is the backbone you can rerun; Europe PMC and Google Scholar widen the set. Compare the logged counts across runs to see whether a rerun returns the same records or new ones, and note any differences in the methods write-up.