Assessment of bibliographic databases performance in information retrieval for occupational and environmental toxicology

Our plain-English summary of Assessment of bibliographic databases performance in information retrieval for occupational and environmental toxicology, by J.F. Gehanno, C. Paris, B. Thirion, et al (Occupational and Environmental Medicine, 1998).

The study

• Authors: J.F. Gehanno, C. Paris, B. Thirion, J.F. Caillard
• Journal: Occupational and Environmental Medicine
• Published: 1998;55(8):562-6
• Cited by: 39 articles, as counted by NIH iCite
• Read it: PubMed · Publisher · Free full text

Abstract

Objective: To determine the efficiency of the major bibliographic databases by assessing the percentage of references among the total literature available that can be retrieved from each database. We also evaluated the best database combinations to carry out an exhaustive search.

Methods: BIOSIS, EMBASE, MEDLINE, NIOSH-TIC, and TOXLINE were searched on two topics: allergy to latex and asbestos and mesothelioma, in the title, abstract, or keywords (textwords). This search was performed for the years 1994 and 1995. All the records were classified by journal and author's name and were verified for each record whether or not it was indexed in each database. Statistical analysis was performed with chi 2 test.

Results: 777 articles in 510 issues were found. The efficiency of each database (percentage of articles recovered) and of combinations varied between 11% and 63% for one database and between 42% and 86% for a combination of two databases. The reasons why these differences exist between databases, and within a database, between two different subjects or two different years are reported.

Conclusion: Firstly, it is not advisable to assert that a bibliography is complete when only one database is searched. Secondly, the efficiency of the databases may be quite different. Finally, it is suggested that the best way to be as exhaustive as possible is to search two or more databases-for example, in EMBASE and TOXLINE, or to a lesser extent EMBASE and MEDLINE. This seems to be the best compromise solution between time consumed for searching and efficiency.

In plain English

This was not a health study. It measured how well five bibliographic databases (BIOSIS, EMBASE, MEDLINE, NIOSH-TIC and TOXLINE) found published journal articles. The authors searched each database on two topics, asbestos and mesothelioma (search A) and allergy to latex (search B), for articles from 1994 and 1995. They pooled all 777 records found (315 for search A, 462 for search B) and treated that pool as the total literature, while noting that this underestimated what had really been published. Single databases retrieved between 11% (NIOSH-TIC) and 63% (MEDLINE) of the records for a topic. For search A, MEDLINE, EMBASE and TOXLINE gave the best results, with no statistically significant difference among them. Two databases combined retrieved between 42% and 86%; the top figure came from EMBASE plus TOXLINE on search A. Three databases retrieved up to 96%. The authors concluded that a bibliography should not be called complete when only one database is searched. They suggested TOXLINE plus EMBASE as the ideal pair, cautioning that only two topics and two years were studied.

We wrote this plain-English summary from the full article with the help of AI, and a second AI model checked each sentence against the article. It is not the authors' wording; the abstract and the article are their own account.

What the study did and found

• Design: Comparison of how many articles each bibliographic database, alone and in combination, retrieved (information-retrieval study, not a health study)
• Who was studied: Journal articles published in 1994 and 1995, retrieved from five databases (BIOSIS, EMBASE, MEDLINE, NIOSH-TIC, TOXLINE); grey literature and books excluded
• Exposure: No health exposure was studied. Asbestos and mesothelioma was one of two search topics (search A). The other was allergy to latex (search B).
• Main results:
  • All five databases combined found 777 records: 315 for search A (asbestos and mesothelioma) and 462 for search B (latex allergy).
  • Single databases retrieved between 11% (NIOSH-TIC) and 63% (MEDLINE) of the total records for one of the two topics.
  • For search A (asbestos and mesothelioma), the best single databases were MEDLINE, EMBASE or TOXLINE, with no significant difference among them.
  • Combinations of two databases retrieved between 42% and 86% of records. The 86% figure was EMBASE plus TOXLINE on search A.
  • Combinations of three databases retrieved between 68% and 96% of records. The 96% figure was EMBASE plus MEDLINE plus BIOSIS on search A.
  • MEDLINE retrieved 63% (199) of search A records but only 37% (173) of search B records. The authors attributed this to a journal issue on allergy that MEDLINE did not index.
  • For search A, NIOSH-TIC retrieved 25% of records for 1994 and 16% for 1995. The authors attributed the difference to its indexing of a 1994 special issue of Annals of Occupational Hygiene on a chrysotile asbestos workshop.
• The authors' conclusion: The authors concluded that a bibliography should not be called complete when only one database is searched. They judged TOXLINE plus EMBASE the ideal combination of two databases.
• Limitations the authors state:
  • Treating the 777 records found as the whole literature underestimated the real amount published.
  • Only two topics and two years were covered, so it is hard to say whether the findings apply to other topics.
  • Grey literature was excluded from the results.

Topics

Medical terms in this study, explained in our Lexicon: NIOSH, Mesothelioma.

Related studies

• EV@LUTIL: An open access database on occupational exposures to asbestos and man-made mineral fibres — E. Orlowski, S. Audignon-Durand, et al. (2015)
• Particle toxicology: from coal mining to nanotechnology — P.J. Borm (2002)
• Toxicology and epidemiology: improving the science with a framework for combining toxicological and epidemiological evidence to establish causal inference — H.O. Adami, S.C. Berry, et al. (2011)
• A short history of the toxicology of inhaled particles — K. Donaldson, A. Seaton (2012)
• Evaluating chemical risks: results of a survey of the British Toxicology Society — P. Slovic, T. Malmfors, et al. (1997)
• Proceedings of the 2017 National Toxicology Program Satellite Symposium — S.A. Elmore, F. Aeffner, et al. (2017)

Browse more research in Occupational and Environmental Medicine, or return to medical research.