The biomedical literature has exploded over the past 50 years. The purpose of this study was to track the growth of drug therapy literature using the online provider PubMed. We utilized the first 20 Disease [C] MeSH listed on the 2005 MeSH tree structures for our study. Drug therapy literature was searched using the disease MeSH with the ending / drug therapy or / chemically induced. Publication numbers were compiled each year from 1966 to 2003. Disease articles increased by 612% from 1966 to 2003. Drug therapy articles increased by 1,116% during the same period. As a percentage of all disease articles, drug therapy publications increased from 11% to 18% over this time period. Drug therapy review articles grew by 10,521% over the time period; drug therapy randomized clinical trials increased by 5,228%. Geriatric drug therapy articles increased at a greater rate than pediatric literature (1,210% vs. 637%). Infectious disease (14%), oncology (14%), immunologic diseases (10%), cardiovascular disease (9%), and neurologic/psychiatric diseases (8%) constitute the highest percentages of all drug therapy articles. Drug therapy literature is growing at a faster rate than the disease literature on PubMed. Drug therapy review articles alone are approaching 10,000 articles/year and are the fastest growing subsection of the drug therapy literature on PubMed.

It is impossible for any clinician to keep up with all the biomedical literature published today. Haynes et al. (1) stated the problem: “If physicians were to read everything of possible biomedical relevance, they would need to read 5,500 articles per day.” Of course, this statement was published in 1986, when there were approximately 20,000 biomedical journals published in the world. Today, there are over 40,000 journals, so a clinician now has to read over 10,000 articles per day to keep up (2). It seems clear that the biomedical literature is proliferating at an accelerated rate. But, does the entire body of literature grow uniformly? Are there subsections of the literature growing at a faster or slower rate? Obviously, there are scientifically “hot” areas that probably expand rapidly for a period of time. We were curious regarding the growth of the pharmacotherapy literature relative to the disease literature and the biomedical literature as a whole.

Bibliometrics can be defined as “the use of statistical methods in the analysis of a body of literature to reveal the historical development of subject fields and patterns of authorship, publication, and use” (3). Medical bibliometrics is often narrowly thought of as the use of citation analysis or impact factors to assess publication patterns (4-6). On the contrary, medical bibliometric methods have been used to evaluate the geography of biomedical publications (7-12) and the publication patterns of individual and therapeutic classes of drugs (13-15).

The purpose of this study was to utilize bibliometric methods to track the growth of drug therapy literature using the online provider PubMed.

METHODS

All searches were performed during the months of February and March 2005 using the online provider PubMed. We searched the first 20 Disease [C] Medical Subject Headings (MeSH®) listed on the 2005 MeSH tree structures [C01] through [C20], excluding [C21] through [C23] (Table 1). Literature was searched using the search strategies listed in Table 2. Search strategy 1 was used to search drug therapy-related articles; search strategy 2 was used for disease articles; and search strategy 3 was used to search for drug-induced disease articles. Searches were further classified using the limit fields “publication type” and “subsets” and publication dates. Additional qualifiers included English language and human selection on the PubMed limit search option. Publication numbers were compiled each year from 1966 to 2003. Data were converted to logarithms for graphical representation. We did not include 2004 data because some were still being indexed into the Medlars system in February and March 2005.

RESULTS

Figure 1 represents the growth of disease articles, drug therapy articles, and drug-induced disease articles from 1966 to 2003. Percentage increase over this time was greater for the subset of disease articles that are drug therapy articles (1,116%) than for all disease articles (612%) (see Table 3). The drug-induced disease literature appears to be a slower-growing subsection of the literature compared to, for example, review articles, randomized clinical trials, and acquired immunodeficiency syndrome (AIDS) or cancer drug literature. As a percentage of all disease articles, the subset of drug therapy has increased from 11% in 1966 to 18% in 2003. Figure 2 maps the growth of the cancer, AIDS, and complementary and alternative medicine drug therapy literature, along with the specialty practice areas of pediatric and geriatric drug literature. The AIDS drug literature follows the expected time course consistent with the identification of the disease and its growth (3,226%).