Pubmed Prompts on Errors in your Searches

When conducting searches in Pubmed, a searcher may make errors, either by omission or by mistyping. In many cases (but not all) Pubmed alerts you to the error and helps you identify it.

When are you alerted to an error?

Sometimes, it happens that a searcher omits one of a pair of parentheses or double quotes, which may result in the search done not as you’ve intended and producing different results. If you made such an error, the “first line of defense” that Pubmed offers would be a message that appears right above the search results.

For example, your intended search strategy was: “lung carcinogenesis” AND (cell* OR molecular)
Instead, you typed “lung carcinogenesis AND cell* OR molecular). In that case you will see, right above the search results: The following terms were ignored: “, )

The “second line of defense” could be found in Advanced Search. Details next to your Query in Advanced Search always display how your Query, or the exact search strategy you entered in the search box, was “translated” by Pubmed, i.e. what was actually processed and searched upon. So, when trying to get more details about the error and the exact place in the search where the error has occurred (especially in the case of complex and lengthy search strategies) you can go to Advanced Search and look at the Details for the search Query in question. A red Warnings icon with an Exclamation Mark would alert you to the problems and by clicking the caret symbol next to it you will see the unmatched quotes and parentheses in red in the search you entered.

Another example: you are searching with the keyword “breast” in the title or abstract fields of the Pubmed record but instead of breast[tiab], where [tiab] is a combined title/abstract field, you mistyped it as breast[taib]. You will see the following message above the search results Unknown field was ignored: [taib] and in the Advanced Search>Details you will see more details related to how Pubmed “translated” your search strategy, as shown below:

In some cases, when you enclose a phrase in double quotes intending to find references that include exactly same phrase, Pubmed actually does not find such references despite the fact that some of them do exist. That’s because Pubmed does not index all phrases and in this it differs from other databases, such as Embase, Scopus, etc. For example, if you type “chemoradiotherapy side effect*” you will be alerted by Pubmed that Quoted phrase not found: “chemoradiotherapy side effect*”. In fact, such references may be present in Pubmed but the search for this exact phrase will be “translated” by Pubmed as a broader query. Pubmed would parse the phrase and search on its components combining them with a Boolean Operator “AND”, behind the scenes. The details of how it was done can be found in Advanced Search>Details.

When are you not alerted to an error?

There are some instances where your error by omission or mistyping may go undetected. If your search includes the Boolean operator “OR” and you omitted it in error, Pubmed will not alert you to it. That’s because a space (in place of the “OR” that you’ve omitted) is treated by Pubmed as a Boolean Operator “AND”. So, your search will produce totally different results, but the error will go unnoticed.

Also, if you mistyped but the term you entered in error is still a legitimate term Pubmed would not detect anything as incorrect. In the first example, if you omit an asterisk in the keyword cell*, where the asterisk was used for truncation, allowing retrieval of references with words with different endings (e.g. cell, cells, cellular), Pubmed will process the search without noticing your error as the noun cell is a legitimate term.

Of course, if you made an error not by omission or mistyping but because of the lack of expertise, e.g. you incorrectly used an “AND” operator or a space instead of an “OR” between the terms in the same category or synonyms, Pubmed will not detect this error either. The takeaway for minimizing the chance of making such an error is taking a Library Pubmed class or, if you feel that the search results are not what you expected, consulting a Research Informationist at the Memorial Sloan Kettering Cancer Center Library.

New Journal: Seminars in Interventional Radiology

A new journal has been added to the Library collection, Seminars in Interventional Radiology. Seminars in Interventional Radiology is a review journal that publishes topic-specific issues in the field of radiology and related sub-specialties.

The journal provides comprehensive coverage of areas such as cardio-vascular imaging, oncologic interventional radiology, abdominal interventional radiology, ultrasound, MRI imaging, sonography, pediatric radiology, musculoskeletal radiology, metallic stents, renal intervention, angiography, neurointerventions, and CT fluoroscopy along with other areas.

Reporting Standards and Research Resource Identifiers (RRIDs)

Following reporting guidelines of all kinds has become common – if not required – practice in health research and publishing over the last decade. The EQUATOR (Enhancing the QUAlity and Transparency Of health Research) Network which promotes “transparent and accurate reporting and wider use of robust reporting guidelines”, now makes available 448 reporting guidelines in its online library. Among them is one entitled – “Unique identification of research resources in the biomedical literature: the Resource Identification Initiative (RRID)”. 

As per the two papers cited below, the “Resource Identification Initiative was launched as a pilot project to improve the reporting standards for research resources in the methods sections of papers and thereby improve identifiability and scientific reproducibility”.

What is a Research Resource Identifier (RRID)?

An RRID is essentially a “Persistent Unique Identifier” that “is designed to help researchers cite the key biological resources used to produce their scientific findings”. Key resources can include: Antibodies, Model Organisms, Cell Lines, Plasmids, and other Tools (software, databases, services). Adding this degree of detail about the resources used in their research makes it possible for others to track these items down should they be interested in replicating or building on the published work. Authors can search for existing RRIDs using the Resource Identification Portal which was created in support of the Resource Identification Initiative, and whose aim is “to promote research resource identification, discovery, and reuse”. 

Not surprising, journal publishers – including Nature and AACR – have gotten on-board and started encouraging the use of RRIDs by authors submitting manuscripts for publication in their journals. In fact, in June 2020, it was announced that AACR journals will be integrating a text-mining tool called SciScore into its journals’ submission platform to help them with better implementing this effort.

From their Press Release:

SciScore evaluates scientific manuscripts for compliance with recommendations and requirements designed to address different aspects of rigor and reproducibility in the published literature, e.g., MDAR, ARRIVE, CONSORT, and RRID standards. This tool provides a score and a supporting report to identify whether key areas of reproducibility and transparency are addressed in the manuscript.

To learn more about this new development, be sure to view the video recording of the December 17, 2020 Advancing Authorship event entitled: Meeting the Challenges of Reproducibility, hosted by the MSK Library. The first speaker was Daniel Evanko, PhD, Director of Journal Operations and Systems at the American Association for Cancer Research (AACR), who has “been heavily involved in efforts to improve the communication, transparency, and reproducibility of published scientific research for over 10 years”. 

For more information on reporting standards, be sure to Ask Us at the MSK Library!