Abstract
This makes it impossible to read and keep pace with the sheer volume of literature available on a particular topic, let alone an entire discipline. 2,3 There is also an increasing reliance on systematic reviews as being representative of what is arguably regarded as being the state of the art relative to a particular topic. Unfortunately, these publications may not be up to date, and there is a high probability that they may promote misconceptions caused by erroneous reporting of misinterpreted or improperly cited material. 1 –4 They do, however, offer the reader the allure of being a concise resource providing information and analysis in the area of interest. It is not surprising therefore, that there is an ever-increasing reliance on the use of databases and search engines as a preferred mechanism for clinicians and researchers to cull the literature and find relevant articles that are suited to their needs. There are several options for doing so, ranging from Google to customized journal listservs, blogs, and various proprietary and open access web sites. All of these resources rely on the use of key words and subject headings to refine the search and retrieve the relevant literature for the particular query.
The United States National Library of Medicine (NLM) produces PubMed and also is responsible for producing and publishing Medical Subject Headings (MeSH), which form the basis for cataloging and categorizing the scientific medical literature. MeSH grew out of the recognition as early as 1947 that the increasing complexity of scientific literature necessitated increasingly sophisticated approaches to organization and access. 5 Subject cataloging and the subject indexing of journal articles were acknowledged to be the essence of bibliographic control, and a set of subject headings were to be created for subject description of books and for indexing of journal articles. The first edition of Medical Subject Headings (MeSH) was published in 1960. MeSH contained 4,300 descriptors and was designed to be used for both indexing and cataloging. MeSH also included 67 topical subheadings. 5 The second edition was published in 1963 and contained 5,700 descriptors without any topical subheadings being included. The Preface to the 1963 edition noted that subheadings presented “overlapping meaning” and the arrival of MEDLARS with coordinated retrieval obviated the need for subheadings. Topical subheadings returned to MeSH in 1966 with 42 subheadings being included. The 2010 edition of MeSH contains >25,000 subject headings in an 11-level hierarchy with 83 subheadings. MeSH is revised and updated on an annual basis. 5
PubMed currently cites >20,000,000 articles. 2 This alone is a staggering statistic, particularly when one considers the fact that each entry must be cataloged, and be capable of being retrieved. Modern bibliographic indexing is an automated process that uses software that scans for specific text and phrases, replacing tedious manual searches of text passages. It is assumed that this process is as accurate as it is efficient. We also presume that searches based on MeSH terms will yield appropriate results. Unfortunately, the disciplines represented in this Journal aren't specifically indexed. In fact, a search using the term “phototherapy” or “laser therapy” will yield articles on psoralen and UVA therapy for cutaneous disorders (PUVA), ionizing radiation, and photodynamic therapy (PDT), rather than photobiomodulation.
Consider the following example brought to my attention by a colleague recently. He wanted to investigate the potential use of low level laser therapy (LLLT) on rosacea. A recent editorial sidebar had apparently suggested that “laser” and other light-based therapies had been tried, with variable or lackluster results. He went to the PubMed web site and conducted a search on “rosacea lllt” using the following search terms and strategy: “rosacea”[MeSH Terms] OR “rosacea”[All Fields]) AND (“laser therapy, low-level”[MeSH Terms] OR (“laser”[All Fields] AND “therapy”[All Fields] AND “low-level”[All Fields]) OR “low-level laser therapy”[All Fields] OR “lllt”[All Fields]. 6 The search retrieved 17 articles. All of the retrieved articles covered topics related to rosacea and its treatment. Several of the articles discussed the use of pulsed dye lasers, near infrared (NIR) devices and intense pulsed light (IPL) devices. However, none of the articles mentioned LLLT. 6
It is indeed unfortunate that such a glaring example of inaccurate bibliographic indexing exists, and still more troubling that photobiomodulation hasn't achieved consideration as being “worthy” of a separate MeSH heading, despite a growing scholarly literature. This lack makes it more difficult for those interested in learning more about these topics to do so. It also misinforms the scientific and lay communities by equating all forms of light-based therapy with each other, irrespective of the nature of the technologies being used. Other ramifications include direct and indirect effects on impact factors, as these are interdependent on the frequency of citation, which in turn depends upon the ability to retrieve, view, and understand published literature.
It is also disconcerting that this is not the only example of inaccuracies in our current system. Have you ever conducted a search, either on your own, or with the assistance of a librarian, only to find that several of the retrieved articles bear no relationship to the searched terms? This is becoming a more frequent experience in my own literature searches, on various topics, including lasers. Perhaps this is inevitable noise caused by the sheer volume and complexity of the literature at hand, and the fact that errors are inherent in any system. Perhaps it is caused by errors in indexing or errors in search engine retrieval protocols. It does make the review of search results more cumbersome, regardless of the reason.
We have made tremendous progress from the days of having to manually search large and dusty tomes in a library reference section. However, there is still room for improvement. You might find that what you seek might not be easily found. Perhaps one day soon, photobiomodulation, laser therapy, and phototherapy will achieve separate MeSH term status. We can only hope.