Databases

Scope: provide the database name(s); add to records where a database rather than a website or interface is discussed

Appraisal of: Younger P, Boddy K. When is a search not a search? A comparison of searching the AMED complementary health database via EBSCOhost, OVID and DIALOG. Health Info Libr J 2009:26:126-135.

Reviewer(s): 
Short description: 

The aim of this study was to investigate whether searching different interfaces DIALOG DataStar; EBSCOhost and OVID SP to the AMED allied health and complementary medicine database produced the same results when using identical search terms.

Using a simple search on Echinacea and the common cold, DIALOG returned 29 hits, OVID 14 and Ebsco 8. Single word searches also demonstrated a similar variation in the number of records retrieved, as did searches on other simple case studies. The authors conclude that depending on the interface, the number of hits retrieved from the same database with the same simple search can vary dramatically.

Limitations stated by the author(s): 

No field limits, wildcards or truncation was applied to the searches. With manipulation the results would be different.
They did not investigate the relevance of the results. 

Limitations stated by the reviewer(s): 
No additional limitations detected by the reviewers.

Appraisal of: Katchamart, W., Faulkner A, Feldman B,Tomlinson G, Bombardier C.. PubMed had a higher sensitivity than Ovid-MEDLINE in the search for systematic reviews. J Clin Epidemiol 2011;64: 805-807

Reviewer(s): 
Short description: 

The aim of this study was to compare the performance of Ovid-MEDLINE vs. PubMed for identifying randomized controlled trials of methotrexate (MTX) in patients with rheumatoid arthritis (RA). They measured the performance of each strategies for Ovid-MEDLINE and PubMed for a systematic review of MTX in RA. They found that PubMed retrieved more citations overall than Ovid-MEDLINE. Of 20 citations that met eligibility criteria for the review, Ovid-MEDLINE retrieved 17 (85%) and PubMed 18 (90%). The precision was 0.881% (NNR: 114) for Ovid-MEDLINE vs. 0.884% (NNR:113)  for PubMed.

Limitations stated by the author(s): 

The study was based on the results of searches conducted for one systematic review of one specific topic. The generalisability of the results needs to be tested in other reviews.

Limitations stated by the reviewer(s): 
The authors did not include the Ovid ‘MEDLINE In-Process & Other Non-Indexed Citations’.
Study Type: 
Single study

Appraisal of: Irvin, E. Does it matter which version of MEDLINE you search?

Reviewer(s): 
Short description: 

Based on the conference abstract: The aim of the study is to determine if there is a difference in the yield of PubMed and OVID MEDLINE and to identify which has the highest yield. The researcher selected ten reviews by the Back, Palliative Care, and Supportive Care (PaPaS) Cochrane Review Group and then ascertained how many of the included studies were contained in PubMed and OVID Medline. In the 5 physical interventions reviews, PubMed contained 57% (43/76) and OVID MEDLINE 49% (37/76). The results of the other 5 reviews are not presented in the abstract. The author explain the higher number of references retrieved in PubMed was due to  the fact that PubMed indexes citations from journals that precede the date those same journals are selected for Ovid indexing. 

Limitations stated by the reviewer(s): 
This is a retrospective analysis, as such it is unable to take account of search strategy implementation and the potential for a search interface to affect the number of articles retrieved. They only checked if the references were indexed in the different interfaces, not the differences from carrying out actual search strategies.
Study Type: 
Single study

Appraisal of: Craven J, Jefferies J, Kendrick J, Nicholls D, Boynton J, Frankish R. A comparison of searching the Cochrane library databases via CRD, Ovid and Wiley: implications for systematic searching and information services. Health Info Libr J. 2014

Reviewer(s): 
Short description: 

The different interfaces for searching Cochrane Library databases are compared in this paper. Firstly CENTRAL and CDSR are compared via Ovid and Wiley and then DARE, NHS EED and HTA databases are compared via Ovid, Wiley and the CRD website. The authors explore search syntax comparisons including proximity operators, MeSH headings, search term order and fields searched. The authors found that there were differences in the use of MeSH headings, searching for free text and using proximity operators. They also suggest a “best match” search syntax to be used across the different interfaces. 

Limitations stated by the author(s): 

Firstly only certain aspects of interface functionality were examined and compared. Secondly only a small number of terms were used to explore each of the search functions. Thirdly the relevance or usefulness of additional records found via one interface but not another was not explored. Lastly, each difference in search syntax was examined individually and not in a combination of search terms/ search string as would happen in the real world.

A follow-up study is planned: to apply the findings to more real-life by comparing full search strategies.

Limitations stated by the reviewer(s): 
The demise of DARE and NHS EED will impact on the relevance of this paper. Cannot find any numbers on how many unique references there were in the different interfaces. No analysis on recall and precision.
Study Type: 
Single study

Appraisal of: Allison MM. Comparison of CINAHL® via EBSCOhost®, Ovid®, and ProQuest®. J Electron Res Med Lib. 2006;3:31-50.

Reviewer(s): 
Short description: 

This papers investigates the differences in three search interfaces – EBSCO, Ovid and ProQuest  for searching CINAHL database. The interfaces were explored for unique search features and records display. In addition, five basic searches that could be considered equivalent across vendor platforms, and thus comparable, were done to ascertain if doing the same search in different interfaces produced the same results.

The results indicated that most interfaces retrieved the same results, the exception being the subject-keyword search which returned the most results in EBSCO (also retrieving 3 / 8,33 % unique results) and the least in ProQuest (where the search also retrieved 4 / 11,11 % unique results). This variation is explained by the different subject options in the three platforms. 

Limitations stated by the reviewer(s): 
No clear conclusion. No investigation of recall and precision.
Study Type: 
Single study

Appraisal of: Wood H, Arber, M, Glanville J. Information resources used in published systematic reviews of economic evaluations [poster].

Reviewer(s): 
Short description: 

This study examined the search methods of recent systematic reviews (SRs) of economic evaluations (EEs) to evaluate whether the sources searched to identify studies reflected current recommendations. A pragmatic search was conducted in Ovid MEDLINE to identify English SRs of EEs published since January 2013. For reviews that met inclusion criteria, two reviewers extracted data on: which general medical and specialised economic databases were searched, the HTA sources searched, and any additional sources or search techniques employed. Results were compared against NICE's economic search requirements for single technology appraisals and the resources recommended in the costs and economic evaluation chapter of SuRe Info.  42 reviews were selected for inclusion. 5 of these reviews searched all of the resources required for NICE submissions, while 9 searched at least 4 of the 6 resource types recommended by SuRe Info. None of the studies searched all 6 types of the search resources recommended by SuRe Info. These results suggest that the majority of SRs of EEs do not meet the current recommendations for identifying economic evidence. 

Limitations stated by the author(s): 

Since a pragmatic search was used to identify articles in the sample, it is possible that some relevant systematic reviews were not included in the analysis.

Because of frequent mistakes and ambiguity in the reporting of database and interface names in the included reviews, the authors had to make some assumptions about which resources were searched. It is possible that the use of both NHS EED and the HTA Database was overestimated as a result of reviews reporting only that "The Cochrane Library", which contains both databases, was searched. 

Limitations stated by the reviewer(s): 
No additional limitations detected by the reviewer.
Study Type: 
Single study

Appraisal of: Golder S, Wright K, Rodgers M. Failure or success of search strategies to identify adverse effects of medical devices: a feasibility study using a systematic review. Syst Rev 2014;3:113.

Short description: 

A systematic review of the safety of recombinant human bone morphogenetic protein-2 (rhBMP-2) for spinal fusion was used as a case study to investigate whether adverse effects terms were prevalent in the bibliographic records of the included studies. The aim of the study was to assess whether the development of a search filter for medical devices would be feasible.

Out of the 82 publications (49 studies) included in the systematic review, 51 were indexed in MEDLINE and 55 in Embase. The results showed that 94% of the MEDLINE records and 95% of the Embase records contained at least one adverse effects related term. However, the records contained a wide range of adverse effects terms, and no single term was able to retrieve the majority of the included publications. In addition, the most successful search terms were different from search terms shown to be most successful when searching for adverse drug effects. The study concluded that the development of adverse effects search filters for medical devices is feasible, but that such filters should be developed using a large set of relevant records in order to identify the optimal combination of search terms.

Limitations stated by the author(s): 

The main limitation to this study is that because only one systematic review was used as a case study, the generalisability of the study results is limited. This case study is also of a medical device with a pharmaceutical component, while a case study of a different kind of a medical device might have given different results.

In addition, unlike many other HTAs and systematic review authors, the study authors were able to obtain unpublished data directly from the manufacturer. They also included an unusually high number of conference abstracts and multiple publications for the same study in their review. This might have influenced the findings of this study.

Limitations stated by the reviewer(s): 
No additional limitations detected by the reviewers.
Study Type: 
Single study

Comments from the authors:

Appraisal of: Kahn TJ, Ninomiya H. Changing vocabularies: A guide to help bioethics searchers find relevant literature in National Library of Medicine databases using the Medical Subject Headings (MeSH) indexing vocabulary. Kennedy Inst Ethics J 2003; 13

Short description: 

The aim of the publications is to present a vocabulary on bioethics as part of the Medical Subject Headings (MeSH) catalog provided by the National Library of Medicine (NLM), Washington (DC). The purpose of the authors is to help researchers to find publications on bioethics and related topics in MEDLINE/PubMed, LOCATORplus and other NLM’s Gateway system databases.

The publications provide information on the subset of keywords in bioethics as well as a list of the Bioethics Thesaurus Keywords and MeSH equivalents. Information is presented on how to use the subset “bioethics”, how to find publications by entering MeSH Headings and free-text queries in MEDLINE/PubMed, and when to use the MeSH Subheading “Ethics”. As well guidance on how to access and use the LOCATORplus database and limiting the retrieval results to bioethics-related publications, searching with MeSH Headings and free text queries is given. 

The 2003 publication has been extended in 2005.

Limitations stated by the author(s): 

No limitations are stated.

Limitations stated by the reviewer(s): 
Some modifications were undertaken since 2005 in MEDLINE/PubMed. NLM Gateway now features the databases Meeting Abstracts and Health Services Research Projects in Progress (HSRProj) only.

Appraisal of: Fangerau HC. Finding European bioethical literature: An evaluation of the leading abstracting and indexing services. J Med Ethics 2004; 30 (3): 299–303.

Short description: 

The aim of the article is to ascertain indexing of medical ethics publications in commonly used bibliographic databases. The purpose of the author is to introduce a set of databases which provides the highest possible quantity of bioethical journal publications to researchers as well as the added value of each database considered.

The article provides information on the methods of collecting the existing Medical Ethics journals, the searching for coverage of these by bibliographic abstract and indexing services, the evaluation results, and finally the conclusions for doing literature searches on medical ethics journal publications.

The search for Medical Ethics journals yielded in 290 periodicals. Thereof 173 journals (60%) were identified in at least one of the considered databases; the remaining journals were not indexed. The highest coverage rates were found in Current Contents (23%) and Medline (22%). A maximum quantity of Medical Ethics journals (45%) was obtained when combining the 10 databases with the highest coverage: Current Contents, MEDLINE, Research Alert, Social Sciences Citation Index, EMBASE, AgeLine, CINAHL, E-psyche, Sociological Abstracts, and Family Index.

A tendency to index more North American than European journals was observed (=bias). EMBASE seems to cover the highest number of European journals.

However, the author recommends that “a medical ethics literature search has to be carried out in several databases in order to reach an adequate collection of literature.”

Limitations stated by the author(s): 
  • The analyzed set of journals is a representative sample, not the whole range of journals possibly comprising articles on medical ethics.
  • Current Contents cites literature from the last five years.
Limitations stated by the reviewer(s): 
No national / supranational services (as European databases) were included in the study. Due to the high turnover in existing relevant journals as well as indexing services (e.g. Current Contents now is part of the ISI Web of Science) the presented results couldn’t be reproduced in detail today. Nevertheless the main results and the recommendations are still valid.
Study Type: 
Single study

Appraisal of: National Reference Center for Bioethics Literature, National Information Resource on Ethics and Human Genetics (Ed). Bioethics searcher’s guide to online information resources. Washington, DC: Kennedy Institute of Ethics, Georgetown Universi

Short description: 

The aim of the article is to present selected information resources available online and relevant to bioethics. The purpose of the authors is to give bioethics researchers an idea and guidance on the literature databases at 1) The Joseph and Rose Kennedy Institute of Ethics (KIE), Georgetown University, Washington, DC, 2) the National Library of Medicine (NLM), Washington, DC, as well as a short introduction to 3) BELIT (German Reference Centre for Ethics in the Life Sciences, DRZE) and 4) Global Ethics Observatory (UNESCO).

The article provides information on searching the KIE databases ETHXWeb and GenETHX, besides the Syllabus Exchange Database and the International Bioethics Organizations Database. In addition, some information on the National Reference Center for Bioethics Literature Classification Scheme is introduced as well as on some features to find bioethics information in the NLM Databases PubMed, NLM Catalog, and the Bioethics Portal. 

Limitations stated by the author(s): 

No limitations were stated.

Limitations stated by the reviewer(s): 
No limitations. As the features in the databases have been modified in the last years the screen shots are not always up-to-date.
Syndicate content