Objective 4a: Keyword Search for Un-indexed Records (search 3)
Search 3 in the required template for your group projects is a keyword search for unindexed PubMed records (non-MEDLINE PubMed records). These are records for articles from journals not indexed by MEDLINE. These articles are represented by records in PubMed because the full-text versions have been deposited in PubMed Central.
_______________________________________________________________________
Why run this search?
Often one fourth of the articles retrieved by a keyword search are unindexed. These records can’t be retrieved by a MeSH search.
_______________________________________________________________________
Prepare for AND RUN the search:
Because this is a search for unindexed records, you will have to remove most of the limits you applied earlier. The only filters that can be applied during a search for unindexed records are language, publication date and systematic review filters. The other filters depend on indexer-applied tags/terms.
- Clear the contents of the PubMed search box, paste the keyword search (without filter terms in the documented version of search 2) into the search box.
- Click the “Search” button or hit the “Enter”/”Return” key on your keyboard.
- Remove the “Article Type” filters (the “Clinical trial, phase II, Clinical trial, phase III, Clinical trial, phase IV, and the randomized controlled trial” filters).
- Remove the “Humans” filter.
- Remove the publication date applied during search 2. That publication date range is not the focus of search 3.
- Now place your cursor in the search box. Place your cursor at the end of the search strategy. Add a space and then —
NOT MEDLINE[sb] AND random*
The NOT means “take out the following thing”. MEDLINE[sb] means “MEDLINE subset”. After you take the MEDLINE subset out of a set of results, unindexed records are all that are left.
The — AND random* — will limit retrieval to items that contain phrases like “random assignment”, “randomly assigned”, “participants were randomized”, “participants were randomised”. If you are researching a rare condition and don’t have enough randomized controlled trials for a monograph, you could add — AND trial — instead of AND random*. If you search for trials, instead of randomized controlled trials, revise the heading accordingly.
- Click the “Search” button or hit the “Enter”/”Return” key on your keyboard.
- This version of the tutorial search retrieved 3 results when the tutorial was last updated in August 2025.
_______________________________________________________________________
Document the search:
- Click the “Advanced” link under the PubMed search box to go to the “Advanced” search page.
- Find the row in the “History and Search Details” table that ends with NOT MEDLINE[sb]
- Click the “Ellipsis” icon (…) in that row and select the “Add Query” option that appears
- Scroll up to the search box and copy its contents.
- Paste the copied search under the heading for search 2.
- If you wish, note the number of results on the line below the search strategy.
Search Strategies:
1 MeSH search for indexed PubMed (MEDLINE) records for Phase II-IV and randomized controlled trials**
(((“eculizumab” [Supplementary Concept]) AND “Hemoglobinuria, Paroxysmal”[Mesh]) AND “Embolism and Thrombosis”[Mesh]) AND “therapy” [Subheading] AND (clinicaltrialphaseii[Filter] OR clinicaltrialphaseiii[Filter] OR clinicaltrialphaseiv[Filter] OR randomizedcontrolledtrial[Filter]) AND (humans[Filter]) AND (english[Filter])
3 results
2 (optional). Keyword search for indexed PubMed (MEDLINE) records for Phase II-IV and randomized controlled trials** that cannot be identified by the MeSH search.
(Eculizumab* OR 219685-50-4 OR h5G1.1* OR h5g11 OR h5G1-1 OR 5G1.1 OR Soliris OR Elizaria OR Alexion OR Acveris OR Bekemv OR Epysqli) AND (((intermittent OR paroxysmal ) AND ( haemoglobinur* OR hemoglobinur* OR haematur* OR hematur* OR haematinur* OR hematinur*)) OR PNH OR Marchiafava-Micheli [tiab]) AND (Clot OR Clots OR budd-chiari[tiab] OR embol* OR postthromb* OR thromb*) AND ((clinicaltrialphaseii[Filter] OR clinicaltrialphaseiii[Filter] OR clinicaltrialphaseiv[Filter] OR randomizedcontrolledtrial[Filter]) AND (humans[Filter]) AND (1800/1/1:2005/1/1[pdat]) AND (english[Filter]))
0 results
3 Keyword search for unindexed PubMed records for randomized controlled trials**.
(Eculizumab* OR 219685-50-4 OR h5G1.1* OR h5g11 OR h5G1-1 OR 5G1.1 OR Soliris OR Elizaria OR Alexion OR Acveris OR Bekemv OR Epysqli) AND (((intermittent OR paroxysmal ) AND ( haemoglobinur* OR hemoglobinur* OR haematur* OR hematur* OR haematinur* OR hematinur*)) OR PNH OR Marchiafava-Micheli [tiab]) AND (Clot OR Clots OR budd-chiari[tiab] OR embol* OR postthromb* OR thromb*) NOT MEDLINE[sb] AND random* AND (english[Filter])
3 results
4. Keywords search for unindexed and indexed PubMed records for meta-analyses, observational and economic studies, and guidelines:
5. (optional) EMBASE search for records for Phase II-IV and randomized controlled trials**:
6 (optional): ClinicalTrials.gov search
The following drug names were omitted from the keyword search strategies 2- 5 because they were used frequently to indicate discussion of irrelevant concepts and were causing retrieval of irrelevant results:
** If you are studying a rare disease and find few if any randomized controlled trials, search for any clinical trial instead of randomized controlled trials only and change the template headings accordingly.
_______________________________________________________________________
Using the search results
Almost all keyword searches retrieve both relevant and irrelevant results. If you are using this tutorial to guide your group project search and have a lot of search results, do not be discouraged! You just have to scan the titles and abstracts retrieved by this search for articles that meet your formulary project’s inclusion criteria. You do not need to read the full-text versions of all the review articles, editorials, and irrelevant trials that have been identified. Just the full-text versions of the relevant trials (depending on your topic either randomized controlled trials only or any clinical trial).