Mapping the literature of cytotechnology

Abstract

The major purpose of this study was to identify and assess indexing coverage of core journals in cytotechnology. It was part of a larger project sponsored by the Nursing and Allied Health Resources Section of the Medical Library Association to map the literature of allied health. Three representative journals in cytotechnology were selected and subjected to citation analysis to determine what journals, other publication types, and years were cited and how often. Bradford's Law of Scattering was applied to the resulting list of cited journals to identify core titles in the discipline, and five indexes were searched to assess coverage of these core titles. Results indicated that the cytotechnology journal literature had a small core but wide dispersion: one third of the 21,021 journal citations appeared in only 3 titles; another third appeared in an additional 26 titles; the remaining third were scattered in 1,069 different titles. Science Citation Index Expanded rated highest in indexing coverage of the core titles, followed by MEDLINE, EMBASE/Excerpta Medica, HealthSTAR, and Cumulative Index to Nursing and Allied Health Literature (CINAHL). The study's results also showed that journals were the predominantly cited format and that citing authors relied strongly on more recent literature.

INTRODUCTION

Cytotechnologists are medical laboratory technologists who specialize in cytology, the study of cell structure, function, and pathology. Much of their work involves preparing, processing, and microscopically examining cell specimens to detect abnormalities that might indicate malignant disease, changes in hormonal status, inflammation, or infection. They may also assist with determining and implementing the appropriate procedures for collecting the specimens [1–4].

As a field of employment, cytotechnology dates back to the 1930s; by the 1950s, it had become a distinct allied health discipline. Its emergence and development as a profession was a direct result of the work of famed cytologist and anatomist George N. Papanicolaou (1883–1962), often called “the father of modern cytology” (5, 6). In the 1920s, Papanicolaou began to publish material on the cytologic method for hormonal evaluation and, in 1928, suggested that this method might be of value in diagnosing cancer of the uterine cervix [7–9]. He subsequently helped organize an extensive study of development of the vaginal smear for early diagnosis of reproductive tract cancer in women. The first paper reporting results of this study appeared in 1941 [10–12].

As interest in the “Pap test” increased and as cytodiagnostic techniques became more widely applied, the need for technical assistants to microscopically examine and screen cytologic specimens also grew. These first technical assistants—many trained by Papanicolaou himself—were the product of apprentice-type programs, which consisted mainly of intensive microscopic study and review of available samples as they were received in the few established cytology laboratories in the late 1930s and early 1940s. Formal lectures were rare, because knowledge of cytodiagnostic criteria for each organ system was still developing [13].

By 1947, several structured programs offered lectures and microscopy training to both physicians and prospective cytotechnologists. The Board of Registry of the American Society of Clinical Pathologists offered the first cytology technician examination in 1957 [14]. Five years later, the first set of uniform national standards for cytotechnology training programs were approved and have been revised and updated four times since then [15].

Current cytotechnology programs are sponsored by universities, hospitals, medical schools, laboratories, or a combination of these facilities. Program length may vary from one to four years, depending on whether the cytotechnology training is continuous or integrated into undergraduate programs that culminate in a baccalaureate degree [16]. All cytotechnology programs in the United States are collaboratively reviewed for approval and accreditation by the Cytotechnology Programs Review Committee of the American Society of Cytopathology. In 1998, there were sixty such active programs (17, 18).

Cytotechnologists presently number around 10,000 [19]. They are employed by hospitals and private commercial laboratories and in research, educational, managerial, or other health care industry positions. Their field is multidisciplinary, incorporating aspects of anatomy, histology, pathology, and physiology, as well as biochemistry, molecular biology, and computerization [20].

This study is part of a larger project sponsored by the Nursing and Allied Health Resources Section of the Medical Library Association to map the literature of allied health (21, 22). The purpose of this study is to identify the core journals in the discipline of cytotechnology through citation analysis and subsequently to assess indexing coverage of these core titles. A secondary purpose is to determine the relative frequency of cited format types and publication years. Searches of several computerized indexes did not turn up any previous similar studies of the cytotechnology, cytopathology, or clinical pathology literature.

The results of this study will hopefully prove beneficial to librarians responsible for maintaining collections that support cytotechnology departments and programs, as well as to librarians providing reference service and information instruction to cytotechnology practitioners, students, and researchers. In addition, the study's results may have implications for journal coverage given to the core literature of cytotechnology by index and database producers. Finally, the study may prove useful to cytotechnologists interested in knowing where their profession's most widely read and cited literature is being published and how to access it most efficiently.

METHODOLOGY

This study has followed the common methodology described in the project overview [23]. Minor variations from the methodology are noted when appropriate.

To select source journals for the study, the “Brandon/Hill Selected List of Books and Journals in Allied Health” [24] was consulted first but did not prove useful because the cytotechnology discipline was not specifically covered, and the nine titles listed under medical technology seemed either too general (e.g., Clinical Laboratory Sciences) or more relevant to other medical laboratory specialties (e.g., Clinical Chemistry). An informal e-mail survey was subsequently sent to the twenty-three cytotechnology program directors who had e-mail addresses included in the Health Professions Education Directory [25], asking them to supply the names of journals they considered to be core in the cytotechnology field. Based on the thirteen responses received to the twenty messages that were successfully delivered, the following three titles were selected as source journals for this study: Acta Cytologica, Cancer Cytopathology, and Diagnostic Cytopathology. The second title was new as of 1997, when it became a separately published, but continuously numbered, subsection of the journal Cancer (making up one volume of the latter per year). Even though it was new and could not be analyzed for a three-year time period as specified in the project methodology, it was nonetheless included in the study because so many of the survey respondents named it as an important title.

Acta Cytologica and Diagnostic Cytopathology were each analyzed for the same three-year time period, 1995 through 1997. Because it was new, Cancer Cytopathology could only be analyzed for one year, 1997. Source articles used for analysis were all research reports, letters, commentaries, and editorials published in these three journals during the specified years that had references at the end, excluding classic article reprints. Reference lists were obtained by searching Science Citation Index Expanded* and, when necessary, by consulting the actual articles.

A fourth title, also new as of 1997—the ASCT Journal of Cytotechnology, published by the American Society for Cytotechnology—was named by a number of survey respondents as potentially noteworthy and was initially included as a source journal in the study. However, it was subsequently excluded for two reasons: (1) the number of citations in it (18) and citations to it (0) were not significant and did not change the study's citation analysis results one way or the other; and (2) it was not covered in any of the major indexing services (even as late as the end of 1998) and thus could not be subjected to the study's indexing analysis.

The following data elements from each cited reference were entered into an Excel spreadsheet: source journal, source year, cited year, and cited journal. If a reference was to something other than a journal article, then the “cited journal” column was filled in as either “book” or “miscellaneous.” The “miscellaneous” category included conference proceedings, government publications, newspaper articles, audiovisuals, software, dissertations, personal communications, and any other material not readily identifiable as either a book or journal.

The completed spreadsheet was first sorted and analyzed by format to identify predominance of formats cited. As noted above, format categories used were journal articles, books, and miscellaneous. The spreadsheet was then sorted and analyzed according to citation year and grouped into multiyear segments. In addition, the cited data was broken down by format type.

Next, the spreadsheet was sorted by cited journal and the number of citations per title was counted. Each title, along with its corresponding number of citations, was subsequently entered into a second spreadsheet, sorted in rank order according to the number of cited references and then divided into three equal zones using the bibliometric principle known as “Bradford's Law of Scattering.” According to Bradford's Law, Zone 1 journals are the titles cited most often and represent the core literature of a field. Zone 2 journals are cited frequently but not as frequently as Zone 1 titles and represent the next most productive body of literature. Zone 3 journals are cited least and represent the least relevant literature. The theoretical construct behind Bradford's Law is that, for any specialty, a relatively small number of journals can be expected to account for a disproportionate amount of literature [26]. Thus, the Zone 1 and 2 groupings can be expected to contain a relatively small number of titles but a relatively large number of citations; with Zone 3, just the opposite is true.

If a journal had changed titles, citations to any previous titles were combined under the most recent one. Citations to the journal Cancer Cytopathology, one of the study's three source journals, were combined under the title Cancer because the separate subsection title was only a year old when the citation study was conducted and: (1) the number of cited references to it (12) was minimal, and (2) authors and indexers did not yet appear to be citing the subsection consistently as a title separate from Cancer, thus making it impossible to tell from the data as entered into the first spreadsheet whether a reference to Cancer was actually a reference to Cancer Cytopathology.

Finally, the journal titles in Zones 1 and 2 were checked for indexing coverage in five databases: Cumulative Index to Nursing and Allied Health Literature (CINAHL), EMBASE/Excerpta Medica, HealthSTAR, MEDLINE, and Science Citation Index Expanded. Even though the project methodology called for checking only one publication year for indexing coverage, the same three years used in the source journal analysis (1995 through 1997) were checked because the additional work involved was minimal, and the searching was done late enough in 1998 (the last quarter) that indexing by the various services was assumed to be complete for all three years. The results of each title search in each database were entered into a third spreadsheet and compared with the total number of articles published in each journal from 1995 through 1997. The percentage of articles indexed from each title in each database was calculated, and indexing coverage was then rated for each journal and database on a scale of 5 to 0 (high to low): 5 (95%–100% of journal articles included), 4 (75%–94%), 3 (50%–74%), 2 (25%–49%), 1 (1%–24%), and 0 (< 1%). A total indexing coverage score was also calculated by adding the individual scores for each journal in each database.

Because it was not feasible to count manually the total number of articles published in all Zone 1 and 2 titles from 1995 through 1997, figures used for total article counts were the numbers of articles indexed by the database that said it provided full coverage of the title. When more than one index claimed to provide full coverage—for instance, both Science Citation Index Expanded and MEDLINE each said they indexed all Zone 1 and 2 titles cover-to-cover—the index that appeared to provide the most complete coverage of the title was used to establish total article counts. In all cases, that index was Science Citation Index Expanded.

Both EMBASE/Excerpta Medica and Science Citation Index Expanded included citations to individual meeting abstracts in four Zone 1 and 2 titles. However, these citations were eventually eliminated from the study because including them would not have improved the coverage scores for any of the titles in either index but would have significantly and disproportionately reduced three of the titles' scores for MEDLINE, which does not index meeting abstracts.

RESULTS

A total of 1,426 articles and 23,565 cited references from the three source journals were analyzed: 790 articles with 13,295 citations from Acta Cytologica; 56 articles with 1,404 citations from Cancer Cytopathology; and 580 articles with 8,866 citations in Diagnostic Cytopathology.

Table 1 shows that 89.2% (21,021) of all the cited references were to journal articles and 9.6% (2,267) were to books. The remaining 1.2% (277) cited miscellaneous formats.

Table 1.

Cited format types and frequency of citations

Table 2 shows citation formats broken down by publication period. For all formats, the 1990 to 1997 period made up the greatest percentage of citations: 45.3% of all journal citations, 47.6% of all book citations, and 61.7% of all other citations. Further analysis of all citations from the 1990s showed that most were published between 1990 and 1994: 85.4% of all journal citations, 88.6% of all book citations, and 63.7% of all other citations. Nearly 85% of all citations to all formats were published between 1980 and 1997.

Table 2.

Cited format types by publication periods

A total of 1,098 different journal titles were cited. Table 3 shows title distribution by zone. Only three titles (0.3%) fell into Zone 1, but they accounted for over a third (36.8%) of all journal citations. An additional 26 titles (2.4%) fell into Zone 2 and accounted for almost another third (30.1%) of all journal references. The rest of the citations (33.1%) were scattered among the remaining 1,069 journals (97.3%). Only one reference appeared in 448 titles; two appeared in 177 titles.

Table 3.

Distribution by zone of cited journals and references

The three Zone 1 titles—Acta Cytologica, Cancer, and Diagnostic Cytopathology—paralleled the three titles selected as source journals for the study. These three titles along with the twenty-six in Zone 2 are listed in Table 4 in descending order according to the number of cited references to each. At least half of the twenty-nine journals were clinical pathology or cytology titles; the remainder were other clinical specialty titles or general medicine titles.

Table 4.

Distribution and indexing coverage of cited journals in Zones 1 and 2

Table 4 also shows indexing coverage scores for each Zone 1 and 2 journal in the five databases searched. Science Citation Index Expanded provided the most extensive indexing coverage of Zone 1 titles (all titles covered, 100.0% of all articles) followed by EMBASE/Excerpta Medica (all titles covered, 93.0% of all articles), then MEDLINE (all titles covered, 92.7% of all articles), and HealthSTAR (all titles covered, 59.7% of all articles). CINAHL did not provide any coverage of Zone 1 titles. For Zone 2 titles, Science Citation Index Expanded again provided the most extensive coverage (all titles covered, 100.0% of all articles), followed by MEDLINE (all titles covered, 82.1% of all articles), then EMBASE/Excerpta Medica (all but one title covered, 70.3% of all articles), and HealthSTAR (all titles covered, 50.5% of all articles). CINAHL provided very limited coverage of Zone 2 titles (five titles covered, 3.6% of all articles). The total indexing coverage scores given in Table 4 indicate the relative indexing coverage for all Zone 1 and 2 titles in each of the databases searched.

DISCUSSION

Not surprisingly, most of the cited references analyzed in the study (nearly 90%) were to journal articles, indicating that journal literature was of primary importance in the cytotechnology field. This finding was consistent with other disciplines in the health sciences. Similarly, most of all analyzed references (nearly 85%) were to materials published since 1980, indicating a strong reliance on more recent literature. Again, this finding was typical of the health sciences disciplines in general. Consistent with Bradford's Law of Scattering, the study's cited journal titles showed a wide dispersion among a small core, with only about 3% of the titles accounting for two-thirds of all the citations.

Of the five indexes consulted, Science Citation Index Expanded provided the most complete coverage. It indexed all issues of all twenty-nine titles published during the period studied, apparently from cover-to-cover, including individual meeting abstracts. In addition, it was the only index consulted to cite Cancer Cytopathology separately from Cancer, facilitating browsing access to the subsection's contents. MEDLINE ranked second in overall indexing coverage (all titles covered, 82.9% of all articles). It indexed all full-length articles and research reports as well as most letters, editorials, and commentaries. EMBASE/Excerpta Medica ranked third (all but one title covered, 71.9% of all articles), followed by HealthSTAR (all titles covered, 51.1% of all articles). CINAHL ranked last, providing only minimal overall coverage (five titles covered, 3.6% of all articles).

For searchers needing comprehensive coverage of the core cytotechnology, cytopathology, and clinical cytology literature, Science Citation Index Expanded appears to be the index of first choice, although it is probably the least available and most expensive of the databases consulted and for many searchers not even an option. For searchers needing full coverage of the substantive core journal literature in the field, the more readily accessible MEDLINE is probably the database of choice, and librarians should feel very comfortable using it and suggesting it to users. For searchers needing access to an adequate portion of core literature in the field that is less research or clinically oriented, HealthSTAR seems a reasonable alternative to MEDLINE and a better choice than CINAHL. While comparing what articles are indexed by the databases providing partial coverage may have provided additional insight into each database's relative usefulness, CINAHL's ranking seems unlikely to have improved because such a small number of core titles were covered (and none from Zone 1).

CINAHL's poor showing was somewhat surprising, considering that coverage of the allied health literature is emphasized in both its name and marketing and that medical or laboratory technology is one of the allied health disciplines it says is covered in the index. A review of CINAHL's journal list does indeed show that eleven titles are covered in this area, but none are specific to cytotechnology and only one, Archives of Pathology and Laboratory Medicine, is from the group of twenty-nine core journals identified in this study [27]. As noted earlier, however, all of these core titles are basically clinical medicine titles, not one of CINAHL's areas of concentration. While acknowledging that different databases understandably have different scopes and missions, extending coverage to some of the less clinically oriented material in these journals seems both appropriate and compatible with CINAHL's focus. For example, including selected articles from some or all these titles about cytotechnology or cytopathology education, training, standards, and trends or medical laboratory administration, automation, regulation, and quality control would not only significantly improve CINAHL's coverage of the cytotechnology and laboratory technology field but also enhance its uniqueness as “the allied health index.” In addition, extending coverage to the relatively new ASCT Journal of Cytotechnology seems indicated. Although excluded from the study for reasons noted earlier, its contents and target audience appear to be well suited to CINAHL's scope.

CONCLUSION

The results of this study showed that journal literature was of primary importance to the cytotechnology field and that the cytotechnology journal literature consisted of a small core with wide dispersion. The results also indicated a strong reliance on recent literature. All of these findings were consistent with the health sciences disciplines in general.

Analysis of indexing coverage of the core cytotechnology journals showed that Science Citation Index Expanded provided the most extensive coverage and CINAHL provided the least. MEDLINE ranked second to Science Citation Index Expanded, providing full (but not complete) coverage, and EMBASE/Excerpta Medica was third, providing adequate coverage. HealthSTAR provided partial coverage, rating somewhat surprisingly ahead of CINAHL.

These results hopefully will be beneficial to librarians working in institutions with cytotechnology programs and departments to build and maintain more relevant collections and to guide users to appropriate indexing sources. In addition, the study's results may have implications for journal coverage given to the core cytotechnology literature by indexing services. Finally, the findings may also prove useful to practicing and prospective cytotechnologists interested in knowing where the most productive literature is and how best to access it.