We chanced upon a number of errors in a PubMed entry (PMID: 24727622) of the abstract of an article published in your journal a decade ago. This prompted us to think how PubMed entries are rectified and whether it may be important to publish an erratum in a forthcoming issue of the journal when the original source on the journal’s website has no error.

The abstract as published on PubMed [1] mentioned that the study was conducted at “6 medical colleges, 7 non-teaching government hospital, and 2 corporate hospitals”. That makes it 15 sites though the total is mentioned as nine. In the PubMed entry, the study period is between August 2072 and March 2073 which is glaringly impossible.

Figure 1: Screenshot of the abstract on PubMed (PMID 24727622) as on May 16, 2024.

This is what prompted us to check the journal’s website and we found the study period was actually between August 2012 and March 2013. The entry on the journal’s website [2], also says the study was conducted at 9 sites — 6 medical colleges, 1 non-teaching government hospital and 2 corporate hospitals.

Even though the entry has been on PubMed for 10 years now, the errors have not been reported or corrected. We checked how a reader could report an error or a mistake in PubMed and get it corrected.

PubMed is a repository of citations of biomedical and life science literature. The National Library of Medicine (NLM), like any other human-run institute, is prone to human or mechanical errors.

Instead of reporting errors in PubMed citations to NLM which controls PubMed, readers who spot any errors are now directed to report them directly to the publishers [3]. The team that submits the XML citation data to PubMed will then be able to fix the errors.

With the PubMed Data Management (PMDM) system released in 2016, correction of citation data is now the responsibility of PubMed data providers and not the NLM Data Review team [4].

Publishers can use their Secure File Transfer Protocol (SFTP) accounts to upload corrected files securely and in a confidential manner. Electronic submission of corrections directly by the publishers now allows the corrected version to be available to the public within 24 hours of uploading the correctly formatted XML file.

NLM’s Errata policy has not changed in the recent past. Journals are expected to publish errata to correct errors appearing in the original article. In line with the International Association of Scientific, Technical & Medical Publishers (STM) guidelines, an original article is never changed [5]. Published articles remain extant, exact, and unaltered to the maximum extent possible. The publisher needs to alert readers to a correction in the publishing record without changing the publishing history.

Since these are not errors in the original article but errors in the PubMed entry for this article, we wonder if publication of an erratum is required.

The PMDM system shortens and simplifies the process of correcting citation errors on PubMed. This is, however, not a citation error but an error in the Abstract. As per Google Scholar database, the article has already been cited at least 53 times. Does the error in PubMed entry of the abstract of a citation need to be handled differently? Should you add an erratum? We would love to watch how IJME handles the error.

Natasha Das (corresponding author — [email protected], https://orcid.org/0000-0003-4418-2013), Cofounder and Director, Hazlo Consultancy Pvt Ltd, Delhi, INDIA; Saurendra Das ([email protected], https://orcid.org/0000-0002-1803-903X), HOD, Department of Clinical Research, IMS-SUM Medical College and Hospital, Bhubaneswar, INDIA

Editor’s Note: We thank the authors for bringing these errors to our attention. We have corrected the abstract on PubMed as of May 30, 2024.