Can I Use one single database such as PubMed to identify articles for my review

In Brief

• Multiple databases should be used when you search for the pertinent resources for your systematic review
• You need to develop strategies so that you can gain access to several databases in a short period.
• PubMed is not the only option, you can access other databases like MEDLINE, embase.com etc.

Background

It is recommended to make use of several databases while searching for relevant resources for your systematic review. Nevertheless, database search is a toilsome job and eats up lot of time because syntax of search methods are specific to the database. Therefore your aim should be to go for an optimized combo of databases to perform proficient searching in your systematic reviews and whether the ongoing practices of published reviews are right or not.

Methods

Developing strategies for search

Comprehensive analysis search strategies are also developed by a search service that is managed by the librarian (Gilreath, 2019). Knowledge specialists have created an efficient method which lets them conduct searches in several databases in a far shorter time than other methods. This process of scanning literature and a realistic examination of it was presented in different review papers. In brief, the approach is an easy way to incorporate the words of the thesaurus and the phrases title / abstract into a single line search technique. And the quest is streamlined. To find out possible new terms, articles indexed by some recognized thesaurus terms but leave out the present search terms in abstract or title were removed. Fresh terms were included to the fundamental search and reviewed. Once an optimum recall has been reached, macros are used to convert the search syntax between databases, while manual modification of the terms of the thesaurus is still needed.

The review assignments embrace a bigger variety of subjects, from healing efficiency and diagnostic accurateness to morals and community health. Searches are normally created in MEDLINE in Ovid; Embase.com; the Cochrane Central Register of Controlled Trials (Primary) through the Wiley Interface; the Science Core Set Site (hereafter referred to as the Web of Science) (Premji & Ganshorn, 2020).

PubMed limits records in the “as supplied by publisher” sub-set to find references not yet indexed in MEDLINE; and Google Scholar. Generally we use the first 200 articles as ordered in Google Scholar’s significance list. If the number of references from other sources was limited, we predicted that the overall number of valid available references would be high. Throughout this scenario, Google Scholar’s impact count was reduced to 100. In addition, we searched for Scopus when the total number of hits was low, and we included CINAHL (EBSCOhost), PsycINFO (Ovid), and SportDiscus (EBSCOhost) in our search when appropriate for the subject (van Eck & Waltman, 2019).

Beginning in May 2013, at the time of retrieval, the number of records obtained from each search was registered for each archive. Upon conclusion of the search, the complete results of all the repositories used for each of the systematic reviews were imported into a special EndNote repository and stored for this study without reduplication. The researchers who requested the search received a reduplicated EndNote file from which they selected the references relevant to their systematic review for inclusion. In this research, all searches were established and carried out by W.M.B.

Finding out appropriate references for reviews that are published

In July 2016, we checked PubMed for all reviews published since 2014, where first authors were associated with Erasmus MC, Rotterdam, The Netherlands, and compared those with search registrations performed by Erasmus MC ‘s medical library (van der Valk et al., 2019). Reviews written were used if the analysis methods and findings were reported at the time of the latest change and if the websites Embase, MEDLINE, Cochrane CENTRAL, Web of Science, and Google Scholar were included in the review at least (Dunaway, 2019). We extracted the list of Final Included References from the published journal article. We did document the first author’s department. We defined specific MeSH terms relating to the most important disease and intervention discussed in the article in order to categorize the categories of patient / population and reaction. We replicated the peak MeSH word out of the MeSH tree just beneath the disease division or, in case its surgery, straight underneath the phrase MeSH therapy. We have chosen the field from a pre-determined set of wide domains that includes therapy, epidemiology, etiology, diagnosis, management, and prognosis. Lastly, we reviewed whether the reports mentioned restricting their included references to a specific study design.

We kept individual End Note library by the author name and year of publication as term of search for individual related publication. This was done to realize whether the searches we made came up with the references that were included and if it has then which database has given the citation.

In case this gave irrelevant output, so we had our search confined to only using a separate portion of the title or the name of the second author. By tracking down the evidence numbers of EndNote search outputs, we came to know about the source database of these resources. If the EndNote fails to show the reference included, then we assumed that the author has certainly applied some alternative measure to classify the references. These alternative methods may be cited references, searching older write-ups or toning famous authors. However, we left these out in our study.

Conclusion

We suggest that if you are searching for systematic reviews in biomedicine or any other subject, you should at least include a combination of Pubmed, MEDLINE ,Embase,  Web of Science, and Google Scholar. Specific subjects resources such as CINAHL and PsycINFO should be incorporated if the analysis section explicitly focuses on the primary focus of a specialized subject website, such as CINAHL for nursing and related health or PsycINFO for behavioral and psychological health sciences. Cochrane CENTRAL may likewise be–  helpful for studies where RCTs are the preferred study design. Ignoring one or more respiratory we mentioned  as the four prime databases will give you  more specific searches with reduced number of results, but it is up to the researchers to decide whether this is worth the increased risk of losing related resources. This research further emphasizes once again that looking for repositories alone is not enough to recover all relevant references, though.

Future research should continue to explore the recollection of individual queries outside the domain scope and should suggest concentrating on the most appropriate combinations of databases rather than specific databases.

References

1. Dunaway, D. (2019). Bibliometrics for Faculty Evaluation: A Statistical Comparison of h-indexes Generated Using Google Scholar and Web of Science Data. Codex: The Journal of the Louisiana Chapter of the ACRL, 5(3), 18–29. Retrieved from http://journal.acrlla.org/index.php/codex/article/view/160
2. Gilreath, C. L. (2019). Computerized literature searching: Research strategies and databases. Retrieved from https://books.google.com/books?hl=en&lr=&id=OCyNDwAAQBAJ&oi=fnd&pg=PT10&dq=Gilreath,+C.L.,+2019.+Computerized+literature+searching:+Research+strategies+and+databases.+Routledge&ots=DvV9fFt6bB&sig=_GVfo2hPtZG-W6M0CkW0ExE0x0A
3. Premji, Z., & Ganshorn, H. (2020). Can database-level MEDLINE exclusion filters in Embase and CINAHL be used to remove duplicate records without loss of relevant studies in systematic reviews? An exploratory study. Journal of the Canadian Health Libraries Association/Journal de l’Association Des Biblioth�ques de La Sant� Du Canada, 41(1), 3–15. Retrieved from https://journals.library.ualberta.ca/jchla/index.php/jchla/article/view/29437
4. van der Valk, E. S., van den Akker, E. L. T., Savas, M., Kleinendorst, L., Visser, J. A., Van Haelst, M. M., … van Rossum, E. F. C. (2019). A comprehensive diagnostic approach to detect underlying causes of obesity in adults. Obesity Reviews, 20(6), 795–804. Retrieved from https://onlinelibrary.wiley.com/doi/abs/10.1111/obr.12836
5. van Eck, N. J., & Waltman, L. (2019). Accuracy of citation data in Web of Science and Scopus. ArXiv Preprint ArXiv:1906.07011. Retrieved from https://arxiv.org/abs/1906.07011