Examining corrected papers in the “Primary Health Care” category in the Web of Science database

Authors

Abstract

Aim The aim was to conduct detailed analyses, as very little is known about the articles corrected in the “primary health care” category in the Web of Science (WoS) database.
Material and Methods ‘Primary health care’ was entered as a category in the WoS search interface; the time period was not selected, and the document types ‘corrected’ and ‘correction, addition’ were selected. The correction grades achieved were analyzed into two groups: major and minor. The locations of the corrections in the article were determined, and subgroup analyses were provided.
Results In the primary healthcare category, corrections were detected in 1,104 out of 106,593 articles. When we examine the correction sources, we find that there are 610 articles by the author, 250 articles by the publisher, six editors, six authors and publishers, four publishers and editors, and three authors and editors. According to the nature of the corrections, 20 (2%) were determined as major corrections, 860 (78%) as minor corrections, and 220 (20.00%) as unknown; no note. An average correction rate of 1.18 per accessible article was determined.
Discussion New strategies should be examined to improve the correction of errors in articles more effectively and efficiently. It should be kept in mind that readers are a control mechanism, and they should be made aware that they are at least as responsible as the editor, publisher, and authors.

Keywords

Introduction

Scientific article editing is a change that may have the potential to affect the general image of the journal, publisher, or author. However, the information based on scientific evidence is of a quality that will be a part of the intellectual foundation, but does not require the article to be removed entirely from the scientific community [1].
The increasing number of publications can lead to scientific misconduct and unintentional errors. These errors may include various elements such as typos, omissions in author information, incorrect labeling of figures, and reference errors. According to the rules of scientific ethics, authors and editors are required to correct errors clearly and promptly as soon as they become aware of them. Publishing incorrect data, especially in fields such as health sciences, can disrupt the progress of science in the right direction [2, 3, 4].
Corrections and retractions have been available on PubMed and Web of Science (WoS) platforms from the past to the present. PubMed began recording corrections and retractions in 1987, while the WoS database shows that corrections have been published since at least 1901. This is actually an expected situation. The publication of corrections predates the invention of the printing press [5]. Correction has been an option for editors and writers throughout the history of science. However, what is noteworthy is that while there has been a steady increase in the number of publications covered by WoS, the correction rate has remained stable since the 1950s [6]. When the fields of articles with WoS-based correction are examined, the most common ones are physics (n = 38,899, since 1905), chemistry (n = 31,463, 1911), engineering (n = 22,521, 1927), biochemistry and molecular biology (n = 21,155, 1930), and general internal medicine (n = 16,443, 1901) areas [6]. The literature review revealed that, on average, there was more than one correction in every 100 articles [7, 8, 9, 10]. It is predicted that the increase in the total number of articles will also lead to an increase in the number of corrections in the coming years.
Although they are part of scientific knowledge, little is known about primary health care articles with corrections. Therefore, this study will specifically try to answer the following questions, and these will be its aims:
1. What is the seriousness of the corrections made to the articles?
2. What are the reasons for the correction of articles in journals?
3. What are the features of correction articles?
4. What is the accessibility of correction articles?

Materials and Methods

Researchers scanned the Web of Science database for corrected articles in the “Primary Health Care” category on October 1, 2024. In addition, all correction articles in the system where the authors did not set a year limitation on the subject were examined. ‘Primary health care’ was entered as a category in the Web of Science search interface; the time period was not selected, and the document types were selected as “correction”, “correction, and addition”. The results are saved to a file. Duplicate articles were excluded from the study. The analysis continued with the correction notes obtained.
The researchers conducted an independent analysis of several correction sets to categorize the correction attributes and determine which contained major or minor corrections. After the analysis, the categorized areas were compared and discussed, and a consensus was reached in terms of common categories. During the data collection phase, one of the authors categorized the remaining revisions as agreed upon. In case of any doubt in the categorization of a correction, the author who collected the data tried to eliminate this situation by consulting the other author. The data recorded after the literature review for each corrected study are as follows: who made the correction, number of corrections, place of correction (title, author information, abstract, introduction, methods, result, discussion, conclusion, keywords, acknowledgment, table/ figure, references/attributions, funding, manner of publication, conflicts of interest), severity of corrections (this is qualitatively categorized as “minor” and “major”).
If the corrections in the article were small in number, did not significantly change the interpretation and conclusion of the research, or did not change the general quality of the article, they were categorized as ‘minor corrections’. These are considered title errors, typos, errors in author information, and minor numerical errors in tables and figures. Errors where there are multiple errors, where corrections lead to changes in the interpretation of data or study results, or where corrections change the general nature of the research are classified as ‘major correction’ [1].
Ethical Approval
Our research involves examining articles published in the publicly accessible Web of Science database. It has been verbally stated that ethical approval from our institution’s ethics committee is not required.

Results

Characteristics of Corrected Articles in the Primary Health Care category
In the primary health care category, 1104 out of 106,593 articles were marked as corrections (rate in the category = 1.035%). Four of these articles were considered duplicates and were excluded from the study. The correction notes for 220 articles could not be found despite research. The final analysis of the study was conducted using 880 articles out of a total of 1,100, for which correction grades were determined. When the number of corrections is examined, as shown in Table 1, 70.82% of the articles corrected in the primary health care category in the WoS database contain only one correction. While the rate of articles with more than one correction was 9.18%, 20.00% were found to have no correction notes. An average correction rate of 1.18 was determined per accessible article (Table 1).
When the sources of the corrections in the articles were examined, the source author was found in 610 articles, and the publisher was found as the source in 250 articles. In addition, there are six editors, six authoırs and publishers, four publishers and editors, and three authors and editors corrections (Table 1).
Severity of Corrections Made to Articles in the Primary Health Care category
According to the results of our research, there were a total of 1043 corrections in 880 articles. It was determined that most of the corrections made to articles in the primary health care category in the WoS database were minor (Figure 1).
When we look at the characteristics of the corrections, the minor corrections are the highest headings in themselves: author information (author names, e-mail, links) (n = 316, 32%), errors in tables and figures (n = 217, 22%), and errors in the introduction (n = 153, 15.5%) are the mistakes made. Other minor corrections are reference (n = 67), result (n = 51), abstract (n = 45), title (n = 30), method (n = 20), acknowledgement (n = 20), conclusion (n = 19), conflict of interest (n = 19), discussion (n = 18), funding (n = 17), manner of publication (n = 10), and keywords (n = 1) errors. Major corrections were the highest headings in themselves: table and figure (n = 10, 25%), result (n = 8, 20%), manner of publication (n = 6, 15%), discussion (n = 5, 12.5%), method (n = 4, 10%). Other major corrections were made to the introduction (n = 2), conclusion (n = 2), references/ citations (n = 2), and funding (n = 1).
When we examine the distribution of total corrections by quality, the three highest correction location numbers are author information 316, table/figure 227, and introduction 155. The lowest are keywords 1, manner of publication 16, and funding 18. The distribution of total corrections according to their characteristics is shown in Table 2.
When the frequencies of the corrections are examined according to their nature, 20 (2%) were determined as major corrections, 860 (78%) as minor corrections, and 220 (20.00%) as unknown, with no note (Figure 1). Corrections that change the overall integrity of the article or affect its interpretations or conclusions are classified as major. In our research, 20 articles were classified as major by the researchers. When we look at the sources of the major corrections, 7 (35%) were from the publisher, and 13 (65%) were from the author. Detailed information and the basis of major corrections are explained in Supplementary Tables S1 and S2.

Discussion

This research revealed a series of results after examining the correction notes of articles corrected in the “Primary Health Care” category in the Web of Science database. A total of 1104 articles that met the research criteria were analyzed. The total number of corrections in these articles was determined as 1043 after the analysis of 880 accessible articles (rate in category = 0.82%). In our research, 70.82% of all articles had only one correction, while 9.18% had more than one correction. Those for which no correction notes were found included a rate of 20.00%. In addition to the low level of increase in correction statistics, literature research has also reported on the seriousness of corrections. In a pilot study investigating randomized controlled trials corrections by Royle and Waugh (2004), articles deemed worthy of review were analyzed. A major error is defined as an error that is expected to affect the analysis and results. One hundred articles published in four general medical journals were reviewed. The collected data was analyzed from the perspectives of an information specialist and an experienced referee/public health consultant. Although the corrections were not at a level that affected the results of the articles, they were considered noteworthy to ensure accurate data that would prevent confusion among evaluators and readers [11]. In another study conducted by Hauptman et al. in the USA in 2014, correction notices of 20 English medical journals were evaluated. A total of 557 articles were examined in the research, and it was determined that 24.2% of them contained at least one major error that significantly affected the article [8]. In another study conducted in Canada, 40 systematic reviews were examined. In total, 26 errors were detected in 127 articles. Among these errors, 38 were identified. When classified according to severity, six errors were found to be major, and 20 errors were minor [12]. In another study examining neurosurgery articles published in the last 30 years, 441 articles were examined, and the error rate was found to be 0.81% [13]. With the importance placed on the publication of research based on reliable medical evidence in the information age, correction policies are expected to be transparent and accessible to all segments, clearly visible to authors and the general public. Much of the reporting and associated controversy regarding corrections is based on medical journals. Journals in the field of medicine are among the most cited and researched journals. WoS and PubMed databases are used in most articles [14, 15, 16, 17].
Upon reviewing the literature, a bibliometric study was conducted by examining seven emergency medicine journals between 2000 and 2020 for publication errors. The results of this research included articles containing 251 corrections. Among the 18250 relevant publications, the error rate of the articles was 1.3% [7]. In our study, 1104 out of 106,593 articles in the primary health care category (a rate of in the category = 1.035%) were corrected. In our study, the increase in the number of correction articles over the years is visually shown (Figure 2). In this article, written in 2024, 13 articles from 2024 were not included in the visual analysis in order to avoid incorrect interpretations. The trend line of corrections in the field of emergency medicine between 2000 and 2020 showed that the number of correction notifications was 10-15 per year, with an average of 11.9, and did not indicate any trend [7]. The corrected sections of scientific articles were recorded by us. The most common errors in minor corrections were errors in author information (32%), tables and figures (22%), and errors in the introduction (15.5%). In major corrections, errors in tables and figures (25%), results (20%), and the manner of publication (15%) came to the fore.
Analyses regarding corrections can be made in any scientific field, on any subject, or throughout a journal. In a study examining corrections and errata regarding COVID-19, the most common corrections were author information (76 records), tables (29 records), results and findings (28 records), introduction (25 records), figures (25 records), and unknown (17 records) were observed to be related to [2]. The study, which examined corrections made in a single journal, was published in PLOS One and involved 649 corrections. This study was similar to the methodology of our research. By error location, the largest number of corrections (39.3%) indicated a situation related to author notification. Author information was followed by figures, funding, and tables, respectively [18]. In the correction analysis conducted in the field of emergency medicine, the areas with the most errors were the author information (40.5%), followed by the tables and methodology section [7]. Our findings, as well as the studies we compared, similarly revealed that frequent corrections were made to author information, tables, and figures. It shows that researchers need to pay more attention to these sections in stages, such as submission, revision, and proofreading, in order to avoid making mistakes.
When examining the sources of corrections in our research, the authors made the majority (55.45%), while 22.72% were made by the publishing houses. A small percentage of corrections came from the editor or other sources. Of the 20 major corrections, 65% are from the author, and 35% are from the publishing house. In our study, when the unknown 20 percent was included, the remaining 78% were marked as minor errors, and 2% were marked as major errors. In a study conducted in the field of emergency medicine, authors were the primary responsible party for errors requiring correction, with a rate of 93.2%. Apart from this, only a small portion of the errors, 7.2%, were classified as major. 62.9% consisted of minor errors that had no impact on the interpretation or conclusion of the manuscript [7]. Although we observe from the correction notes that minor errors are the majority, it is essential to identify major errors that could alter the results and interpretations of the article, thereby preventing the spread of misinformation in the scientific world.

Limitations

This study had some limitations, the most important limitation being that a single database was used. Additional articles may have been found in other databases to be included in the study. However, the WoS database is a platform that is frequently used safely in bibliometric studies. Although a single database was used to provide a more detailed coverage of a category, this deficiency was attempted to be mitigated by compiling a large number of articles from a comprehensive and diverse pool with considerable effort. The second limitation is that there may be differences between researchers’ evaluations regarding the determination of major and minor errors or the source of correction. The two researchers aimed to minimize differences between observers by frequently reaching consensus.

Conclusion

With the advent of online journal publishers and scholarly social media platforms, new strategies must be examined to track and correct errors more effectively and efficiently. An article containing incorrect information can spread further if it is cited. For these reasons, it should be kept in mind that readers are a control mechanism, and they should be made aware that they are at least as responsible as the editor, publisher, and writers. Additionally, authors and publishers must collaborate more effectively to produce high-quality scientific articles.

References

1. Wager E, Williams P. Why and how do journals retract articles? An analysis of Medline retractions 1988–2008. J Med Ethics. 2011;37(9):567-70. doi:10.1136/ jme.2010.040964.
   Article PubMed Google Scholar
2. Moradi S, Abdi S. Pandemic publication: correction and erratum in COVID-19 publications. Scientometrics. 2021;126(2):1849-57. doi:10.1007/s11192-020- 03787-w.
   Article PubMed Google Scholar
3. Graf C, Wager E, Bowman A, et al. Best Practice Guidelines on Publication Ethics: a publisher’s perspective. Int J Clin Pract Suppl. 2007;61(152):1-26. doi:10.1111/j.1742-1241.2006.01230.x.
   Article PubMed Google Scholar
4. Molckovsky A, Vickers MM, Tang PA. Characterization of published errors in high-impact oncology journals. Curr Oncol. 2011;18(1):26-32. doi:10.3747/ co.v18i1.707.
   Article PubMed Google Scholar
5. Beal P. A dictionary of English manuscript terminology: 1450 to 2000. Oxford: Oxford University Press; 2008.p.452-7.
   PubMed Google Scholar
6. Fanelli D. Why growing retractions are (mostly) a good sign. PLoS Medicine. 2013;10(12):e1001563. doi:10.1371/journal.pmed.1001563.
   Article PubMed Google Scholar
7. Vural S, Kaya HB, Coskun F. A bibliometric study on the publication errors in emergency medicine journals from 2000 to 2020. Am J Emerg Med. 2022;60:140-4. doi:10.1016/j.ajem.2022.08.001.
   Article PubMed Google Scholar
8. Hauptman PJ, Armbrecht ES, Chibnall JT, et al. Errata in medical publications. Am J Med. 2014;127(8):79-785. doi:10.1016/j.amjmed.2014.03.012.
   Article PubMed Google Scholar
9. Castillo M, Northam M, Halm KE. Postpublication errors in imaging-related journals. AJNR Am J Neuroradiol. 2012;33(8):1447-8. doi:10.3174/ajnr.A3026.
   Article PubMed Google Scholar
10. McEachern JD, Leswick DA, Stoneham GW, et al. Radiological errors in the Canadian Journal of Emergency Medicine. CJEM. 2014;16(5):361-9. doi:10.2310/8000.2013.131183.
    Article PubMed Google Scholar
11. Royle P, Waugh N. Should systematic reviews include searches for published errata? Health Info Libr J. 2004;21(1):14-20. doi:10.1111/j.1471- 1842.2004.00459.x.
    Article PubMed Google Scholar
12. Farrah K, Rabb D. Errata for trial publications are not uncommon, are frequently not trivial, and can be challenging to access: a retrospective review. J Med Libr Assoc. 2019;107(2):187.doi:10.5195/jmla.2019.629.
    Article PubMed Google Scholar
13. Akhaddar A. Error publication (published erratum) in neurosurgical journals worldwide using PubMed during the last 30 years. Childs Nerv Syst. 2021;37(2):637-43. doi:10.1007/s00381-020-04824-y.
    Article PubMed Google Scholar
14. Fang FC, Steen RG, Casadevall A. Misconduct accounts for the majority of retracted scientific publications. Proc Natl Acad Sci U S A. 2012;109(42):17028-33. doi:10.1073/pnas.1212247109.
    Article PubMed Google Scholar
15. Peterson GM. Characteristics of retracted open access biomedical literature: a bibliographic analysis. JASIST. 2013;64(12):2428-36. doi:10.1002/asi.22944.
    Article PubMed Google Scholar
16. Samp JC, Schumock GT, Pickard AS. Retracted publications in the drug literature. Pharmacotherapy. 2012;32(7):586-95. doi:10.1002/j.1875- 9114.2012.01100.x.
    Article PubMed Google Scholar
17. Steen RG. Retractions in the scientific literature: is the incidence of research fraud increasing? J Med Ethics. 2011;37(4):249-53. doi:10.1136/ jme.2011.043463.
    Article PubMed Google Scholar
18. Strothmann M. No bad publicity: the bibliometric advantage of corrected articles in PLOS ONE. Science & Technology Libraries. 2018;37(4):394-408. doi: 10.1080/0194262X.2018.1503990.
    Article PubMed Google Scholar

Declarations

Additional Information

Publisher’s Note
Bayrakol MP remains neutral with regard to jurisdictional and institutional claims.

Rights and Permissions

Creative Commons License

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0). To view a copy of the license, visit https://creativecommons.org/licenses/by-nc/4.0/

View full license details →

About This Article