Why am I getting old papers among my update search results for Systematic Reviews?

It would help if you understood how the databases’ indexing system works.

I don’t know of any Information Professional (librarian/information specialist) involved in Evidence Synthesis projects and has not faced this question. The review team is curious and furious about the old papers appearing among their update search results! They think they might have screened it before and you are adding to their workload.

The first microaggression that a librarian gets on such occasions is that “Are you sure you have done it correctly?” No, sir, no ma’am, I am a professional who enjoys doing things wrongly, putting my job and, worst, my profession at risk, and not checking my work before sending out the results!

REASON: The Databases’ Indexing Process

To answer this question, you need to understand how databases such as MEDLINE and Embase index the journals and their contents.

The databases were created to save our time in browsing, allowing us to search rather than browse. Good examples are the phonebook, the table of content at the beginning of the book, or an ‘index’ at the end of the book. Imagine wanting to find a Prednisolone in a 500-page textbook on Spinal Cord Injury browsing all the pages!

What the electronic databases do is to create a record per journal paper and enter the values/date for the following data fields/metadata:

  • Bibliographic information (authors, title, journal name, year of publication, date of publication, databases entry date, volume, pages, DOI, ISSN, etc.)
  • Abstract, and
  • Some index terms (usually called controlled vocabularies such as MeSH and Emtree)

Not every record contains all this information; some don’t have abstracts, some have no titles! However, if their content matches your search, the record will appear among your search results.

This process of documenting information for journal papers in bibliographic databases is usually called ‘Indexing’ and God knows publishers kill to get their journals indexed in MEDLINE and Embase! Why? because their audience searches these databases to find information. No one has time to browse millions of journal articles! Simply, the publishers will earn more money if you find the papers and go to their journal to read the full-text report. The publishers even charge more for the indexed journals either on open access fee or subscription fee!

What it has to do with old records in update search?

Patience, my friend!

Since there are so many journals with a very diverse quality, the good databases have to peer-review each journal to decide if it is good enough to be indexed in the database. This indexing process sometimes takes 5 years because some databases need to know if the journal is stable! As someone who has indexed over 40 journals in databases, I can tell you that many journals from non-famous publishers or low- or middle-income countries usually fail on their first try.

On the other hand, the databases approach the copyright owners of the important journals that are not indexed in the database even though the journal might have ceased publication 50 years ago! For example, if MEDLINE indexes an important old journal that Embase doesn’t, Embase would try to find a way to agree with the new copyright owner of the journal to index them. The process of negotiating an agreement to index a journal or a publishers’ all journals may take years and fail several times. Sometimes there is no agreement and that’s one of the reasons we have to search both Embase and MEDLINE for systematic reviews because each of them indexes unique journals that the other one doesn’t. Another reason is that MEDLINE is based at US NLM is biased towards indexing the English-language journals that would benefit US users.

When a database starts indexing a newly added journal to their collection, they don’t usually index only the upcoming issues but also the back issues and old papers from the journal as well. Some databases only go back to when the journal started to publish high-quality papers, others index everything from volume 1, issue 1 until today.

Still, shouldn’t we limit the update search to publication dates?

No! Publication date of articles is an alternative for limiting the update search when the databases entry date or database publication date is unavailable!

Since the information professionals are aware of the changes in databases and the indexing process, they try to use database entry date when limiting the search. It means searching for the records that have entered the database recently rather than the papers that have been published recently!

Is that it? Sometimes, the librarians may limit a search to databases ‘record update date’. Why? Because some databases may change their existing records:

  • The database may change their index terms or add new index terms (MeSH/Emtree), so if you were using that index term in your search, your previous search would not find these records but your new search, which is limited to the ‘update date’ of records will!
  • The database may add notes to articles on errata/corrections/retraction. You would miss if the paper has errors or is retracted and it is in your systematic review. Using ‘update date’ rather than ‘publication date’ or ‘database entry date’ would solve this issue.

What happens if a record from my previous search gets updated?

Since not all databases provide ‘database entry date’ or ‘records update date’ as a search option and it is important for systematic reviews, some information specialists — the perfectionist type — prefers to repeat the searches in full and run a cross-search de-duplication against the previous search at the cost of missing updated records during the de-duplication process. There is no perfect way to update the searches right now :D The only solution is to search for included papers again to see any notes, comments, retraction, or erratum on the publisher’s website before publishing the review.

Bottomline

Databases may index old papers between your old and update searches and those papers will appear among your update search. Be polite when asking, keep calm and trust your librarian.

Please Follow and Sign Up for this blog for similar readings :D

--

--

--

A Proper Information Scientist/Professional with a Pinch of Career and Life Lessons

Love podcasts or audiobooks? Learn on the go with our new app.

Recommended from Medium

Hybrid Recommender System

Is it a stone or something else?

The Tableau Desktop Certification: Why You Should Care and How You Can Pass

The Netflix Prize — How Even AI Leaders Can Trip Up

How to detect and remove outliers from a dataset?

Crypto feature importance for Deep Reinforcement Learning

Volume Price Analysis

How to evaluate social scoring value?

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Farhad

Farhad

A Proper Information Scientist/Professional with a Pinch of Career and Life Lessons

More from Medium

An idiot’s guide to finding love in managed isolation

Hunger or Thirst: How to Differentiate ?

Of Action by and Change in novice student researcher: Reality check before engaging in Action…

Participatory Action Research: Participatory means collaborative efforts to empower participants. Action relates to change in real life, Research is about creating new knowledge.

Seven scientific women who deserve the spotlight