Different types of life cycles with regard to citation history of top-cited articles in ophthalmology. The abscissa (x axis) shows the years by decade, and the ordinate (y axis) shows the number of times that an article was cited for 10-year cumulative data. The open circles illustrate the year of publication of the cited article with the reference number. A-E, Type 1 life cycle. F-H, Type 2 life cycle. I-N, Type 3 life cycle. O, Type 4 life cycle. P, Type 5 life cycle.
Ohba N, Nakao K. The 101 Most Frequently Cited Articles in Ophthalmology Journals From 1850 to 1949. Arch Ophthalmol. 2010;128(12):1610-1617. doi:10.1001/archophthalmol.2010.308
Copyright 2010 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2010
We screened 32 ophthalmology journals that had published articles during the period from 1850 through 1949 to identify top-cited articles in the field of ophthalmology (hereafter referred to as citation classics) using the online database Science Citation Index Expanded (Thompson Reuters, Chicago, Illinois). The 101 most frequently cited articles were published in 16 journals. Archives of Ophthalmology had the most top-cited articles (n = 31), followed by American Journal of Ophthalmology (n = 24) and Albrecht von Graefe's Archiv für Ophthalmologie (n = 9). These articles originated from 14 countries, with the United States publishing the majority (n = 58). Most of the citation classics are clinical studies on topics such as rubella cataract, retinopathy of prematurity, keratoconjunctivitis sicca, sympathetic ophthalmia, and the first report of eponymous diseases (eg, Leber hereditary optic neuropathy, Duane retraction syndrome, and Stargardt disease). A considerable number of these articles were ignored initially and for several decades after publication, but, like the classic fairy tale Sleeping Beauty, they have been rediscovered. Our study provides a historical perspective on the classic papers in the literature that are still influential in ophthalmology.
The citation of published articles provides an indication of the relevance of the cited work to later literature. The number of citations an article receives after its publication reflects its influence on the scientific community. Analysis of the number of times an article is cited may allow for the identification of a seminal advance in a specialty and may provide a historical perspective on the field's scientific progress. Citation classics in a given specialty are defined as those articles that have been most frequently cited in journals dedicated to the specialty and related fields. An earlier study identified and described citation classics in ophthalmology from 1975 to 2006.1 We report herein an analysis of classic articles that were published in ophthalmology journals during the century from 1850 to 1949 and have received the most frequent citations up to the recent decades.
In September 2009, we accessed the online database Science Citation Index Expanded (Thompson Reuters, Chicago, Illinois),2 which provides citation data for its source journals, and searched for articles that were published in ophthalmology journals between 1850 and 1949, using the Cited Reference Search option of the database. In the query box for the year of publication, we entered “1850-1949” to limit publication during the century. In the query box for the journal source, we used a Boolean “OR” logic search strategy with a set of suffix terms, including ophth*, ophthal*, ophthalmol*, ophtal*, oftal*, auge*, ocul*, ottal*, eye*, and physiological optics. In this way, we identified 7316 articles that were published in a total of 32 ophthalmology journals between 1850 and 1949 and that received at least 1 citation (Table 1). With the use of this search strategy, we retrieved data on the number of times each article was cited, the first author's last name and first-name initials, the abbreviated journal title, and the article publication year, volume, and first page number. Citation format errors in the retrieved information were not rare, and they included simple page number inversions, incorrect volume number but correct page number and year, and partial misspelling of author's last name, but we were generous in our citation tallies, including these relatively accurate citations as long as there was only 1 minor error per citing string.
The records were compiled in an Excel spreadsheet (Microsoft 2003, Microsoft, Redmond, Washington). The 101 most-cited articles were identified, and their original texts were reviewed to characterize the article title, authors, affiliation and country, journal source, publication year, language, type of research methodology, study topic, and ocular tissue studied. Furthermore, information about the publications that cited each classic article was obtained to investigate the subject category of the articles that cited the relevant classics and the year-by-year citation history to define the relationship between year since publication and citations per year.
Table 1 provides a summary of the journal titles and publication years that we screened. Of 32 journals, 19 started publication in the second half of the 19th century, with the oldest one in 1838 (Annales d’Oculistique), and 13 started publication in the first half of the 20th century. The articles were written in 6 different languages: English (n = 20), German (n = 7). French (n = 2), Japanese (n = 1), Polish (n = 1), and Russian (n = 1).
We identified a total of 7316 articles that have received at least 1 citation. The number of articles cited was led by American Journal of Ophthalmology (third series), followed by Archives of Ophthalmology, Klinische Monatsblätter für Augenheilkunde, Albrecht von Graefe's Archiv für Ophthalmologie, British Journal of Ophthalmology, and Transactions of the Ophthalmological Society of the United Kingdom. These journals also had a larger number of articles that were cited 50 times or more (Table 1). The number of articles receiving at least 1 citation increased steadily from 1850 to 1948 (ie, from 82 articles in the 1850s to 2070 articles in the 1940s).
With regard to number of times cited, the top-cited article was the one by Gregg,3 whose first report of rubella cataract was published in 1941 in the Transactions of the Ophthalmological Society of Australia and has since been cited 1160 times; this number is greater than that of the top-cited article in the previous study from 1975 to 2006.1 As shown in Table 2, the remaining 4 of the top 5 articles cited are as follows: Osterberg4 in 1935 on the human retinal photoreceptor cell topography (cited 417 times); Takayasu5 in 1908 (cited 372 times), which was the first to describe Takayasu arteritis; Talbot and Marshall6 in 1941 on the neural mechanisms of visual localization and discrimination (cited 366 times); and Michaelson and Campbell7 in 1948 on the development of retinal vascular system (cited 365 times). With regard to citation rank, the article that received 100 citations ranked 59th, the article that received 75 citations ranked 121st, the article that received 50 citations ranked 241st, and the article that received 25 citations ranked 734th. The top 101 were defined as citation classics3- 103 because there were 4 joint positions at the 98th rank with 80 citations.
Table 3 provides data on the journals in which the citation classics were published. The citation classics were published in 16 journals. Archives of Ophthalmology had the most top-cited articles (n = 31), followed by American Journal of Ophthalmology ( n = 24 ), Albrecht von Graefe's Archiv für Ophthalmologie (n = 9), British Journal of Ophthalmology (n = 8), and Acta Ophthalmologica (n = 7). Up to the mid 1910s, before the founding of British Journal of Ophthalmology, American Journal of Ophthalmology (third series), and Acta Ophthalmologica, European journals (including Albrecht von Graefe's Archiv für Ophthalmologie and Transactions of the Ophthal mological Society of the United Kingdom) contributed predominantly to the citation classics.
Table 4 shows the countries of origin of the citation classics. More than half of the citation classics originated in the United States, followed by the United Kingdom, Germany, Sweden, and Switzerland. These articles were written in English (n = 84), German (n = 14), French (n = 2), and Japanese (n = 1).
Of the 101 citation classics, 77 were written by a single author, 18 by 2 authors, and 6 by 3 authors (Table 2). Ten authors contributed to 2 or more articles as the first author. Cogan authored 6 articles,17,55,63,80,94,99 Davis,19,35,65 Friedenwald,14,40,76 and Holmes12,13,37 each authored 3 articles, and Callender,16,46 Duane,27,59 Kinsey,72,90 Leber,28,53 Mandelbaum,54,96 and W. C. Owens23,93 each authored 2 articles (Table 2). None of the citation classics were the product of amulti-institutional or international collaboration.
The specialty categories of the 101 citation classics (Table 2) were anatomy/physiology/biochemistry (n = 25), pediatric ophthalmology/genetic ophthalmology (n = 25), pathology/microbiology/pharmacology (n = 14), general medical and surgical ophthalmology (n = 14), neuro-ophthalmology (n = 10), ocular oncology (n = 7), and systemic ophthalmology (n = 6). Seventy-one citation classics were observational studies of new clinical disease, diagnostic and therapeutic innovations, or histopathologic studies. Thirty articles reported results of basic research, of which 22 were on the physiological optics or visual psychophysics in humans and 8 were on the ocular anatomy and physiology in animals. The main topics included studies on retina, choroid, and vitreous tissue (n = 28), visual function (n = 15), ocular adnexa (n = 11), the cornea (n = 11), the uvea (n = 10), and the optic nerve (n = 7).
The subject categories of journals that cited the citation classics ranged from ophthalmology to many other specialties (Table 2). Of the 101 citation classics, 74 were cited most by articles in ophthalmology journals; these articles with predominant citations by ophthalmology journals dealt with a diverse range of clinical topics, such as tonometry,14 age-related macular degeneration,58 use of cocaine in ophthalmic surgery,68 pigmentary glaucoma,78 aqueous veins,86 herpetic keratitis,89 and gonioscopy.92 The remaining 28 articles were cited most frequently in other medical journals besides ophthalmology, including clinical neurology or neuroscience (n = 10), general medicine (n = 8), pediatrics (n = 2), and surgery (n = 2). Fifteen articles3,5,6,12,13,20,21,37,39,44,62,73,75,93,98 have received citations primarily in speciality journals other than ophthalmology. For instance, the article by Gregg3 on rubella cataract has received the most citations in pediatrics, public health, environmental health, occupational health, and obstetrics journals; the 2 articles by Holmes12,13 on the visual disturbances due to perforating gunshot head injuries have received many citations in neuroscience, clinical neurology, and behavioral science journals; the article by Warkany and Schraffenberger20 on the experimental ocular malformations due to maternal vitamin A deficiency has received most citations in articles concerning developmental biology, nutrition, and general medicine; and the article by Tay44 on the cherry red spot in the macula has been cited primarily in journals dedicated to clinical neurology, medicine, and pediatrics.
The Science Citation Index Expanded online database provided the data on the annual citations that each article has received since publication. Analysis of 10-year cumulative citation data revealed that the relationship between the decade since publication and citations per decade is fitted by 1 of 5 types of life-cycle curve (Figure). Type 1 is characterized by a rapid increase in citations to a peak in the first or second decade after publication, followed by a decrease in citations in the subsequent decades (Figure, A-E), and this type was identified in 17 citation classics. Type 2 is characterized by a gradual increase in citations to a peak in the 3 or more decades after publication, followed by a gradual decrease in citations in the subsequent decades, revealing an inverted u-shape relationship (Figure, F-H); the type 2 life-cycle curve was identified in 36 articles, 11 of which were published before 1930, including Batten's case reports of cerebral degeneration with macular changes,84 Bailliart's assessment of retinal artery blood pressure,100 and Elschnig's work on sympathetic ophthalmia.101 A type 3 life-cycle curve means that an article takes several decades after publication to get started in the citation history, followed by a marked increase in the latest decades (Figure, I-N); the type 3 life-cycle curve is identified in 30 articles. It is remarkable that 15 of the 30 articles with the type 3 life-cycle curve were published before 1920, including Leber's description of hereditary optic neuropathy,53 Treacher Collins' description of eyelid anomaly,73 Schirmer's study of tear production and secretion,15 von Hippel's description of retinal angioma,33 Tay's description of familial infantile idiocy,44 Duane's retraction syndrome,27 de Rötth's introduction of amniotic membrane transplantation for conjunctival repair,47 and Sorsby et al's description of fundus dystrophy characterized by late-onset progressive macular degeneration.38 The type 4 life-cycle curve has a nearly constant citation rate from publication up to the present, and this type was identified in 10 articles (Figure, O); the type 4 life-cycle curve is seen in Takayasu's case report5 of unique retinal vascular disease called Takayasu retinitis, and Heerfordt's uveoparotid fever.26 The type 5 life-cycle curve has marked citations in the early decades after publication, followed by a trough of few or no citations, which is followed by a revival in citations (Figure, P); this type was identified in 7 articles, including reports of retinopathy of prematurity by Terry9 and Owens and Owens,23 and Lindau's description of a syndrome characterized by angiomatosis of the retina and hemangioblastoma of the cerebellum.61
These citation classics provide a historical perspective of the progress in ophthalmology from 1850 to 1949. It is remarkable that the majority of these old articles are still cited and continue to have an influence in the field of ophthalmology. Clinical observations were predominant with case reports of new disease, innovations of visual functional examinations, and improvement of medical and surgical management of ophthalmic disease. Basic studies consisted of classical physiology, biochemistry, and pharmacology as well as light microscopic anatomy and histopathology. Notably, simple methodologies have made a profound impact on a wide range of clinical topics, including rubella cataract,3 keratoconjunctivitis sicca,8 retinopathy of prematurity,9,23,93 visual disturbances associated with cerebral damage,12,13,37 diabetic retinopathy,31 senile disciform macular degeneration,58 and pigmentary glaucoma.78 The articles that first described a genetic disease, such as Usher syndrome,24 Stargardt disease,30 and Leber hereditary optic neuropathy,28 were a remarkably high proportion of the citation classics, which may be attributed to the development of molecular genetics in the 1990s and 2000s.
A comparison of the current citation classics with the previously reported citation classics in ophthalmology from 1975 to 20061 demonstrates that there was a shift of the main contribution from European countries to the United States and points to the outstanding progress in research technologies, including refined epidemiology, molecular biology, immunology, laser technology, intraocular lens implantation, vitreoretinal surgery, randomized controlled trial–based research, and translational research. Concerning authorship, 77 of the 101 current citation classics were by a single author, whereas 14 of the 100 articles in the modern citation classics1 were product of multicenter studies with as many as 50 or more authors, and the remaining 86 articles were by a mean of 3 authors. The changing trend in authorship may be attributed to the advancement of technologies that have necessitated the collaboration of researchers of different background.
The precise tempo of the aging process of the citation history of publication may vary with the category of journals and the study topic. Among the 5 types of citation life cycles that we identified herein, the type 3, characterized by many decades' latency after publication followed by marked citations in recent years, provides an intriguing story that reminds us of the classic fairy tale Sleeping Beauty: an article that goes unnoticed (sleeps) for a long time and then, almost suddenly, attracts a lot of attention (is awaked by a prince). A representative instance is the report in 1940 of amniotic membrane transplantation for ocular surface disorders by de Rötth,47 which was barely recognized until the late 1990s (Figure). It is notable that related articles by contemporaries Sorsby et al104 and Lavery105 had a similar citation life cycle, although with fewer citations. To the best of our knowledge, this is the first paper to use the metaphor of Sleeping Beauty to analyze citation patterns in the medical literature. The sleeping beauties are not just an exotic whim in physics and psychology.106,107
There are limitations in the current citation classics. The online database Science Citation Index Expanded is updated on an ongoing basis,2 and the citation scores may vary with the date of retrieval. For instance, the analysis of the Archives of Ophthalmology 's most frequently cited articles in 1986108 revealed that Davis65 and Forrest32 had then received 90 and 101 citations, respectively, whereas, in the current analysis, their citations are 95 and 137, respectively. In the sociology of science, obliteration by incorporation occurs when, at some stage in the development of a science, certain ideas become so accepted that their contributors are no longer cited (ie, original seminal work is absorbed into current knowledge and is no longer explicitly cited).109 Not a few articles must have been excluded from the list of the current citation classics. To take a few examples, von Graefe's 1853 paper on iridectomy for acute glaucoma110 and Horner's 1864 paper on the Horner syndrome111 have received fewer citations than the criteria of the current citation classics. Furthermore, herein, we have been concerned only with articles in ophthalmology journals. It is likely that we have overlooked influential articles that were published in other medical journals and received frequent citations (eg, Behçet's 1937 article on Behçet disease that was published in Dermatologische Wochenschrift112 and has received more than 850 citations in a wide range of specialty journals). Despite these limitations, the current citation classics provide a historical perspective of the scientific advances of ophthalmology from 1850 to 1949 that are still influential in the 21st century, and they offer researchers hints about how to produce an article that can be cited frequently over several decades.
Correspondence: Norio Ohba, MD, Orthoptics and Visual Science, Faculty of Medical Welfare, Aichishukutoku University, Asahigaoka 109-3, Minami-sakaemachi Owariasahi-shi, Aichi 488-0046, Japan.
Submitted for Publication: October 21, 2009; final revision received February 14, 2010; accepted February 16, 2010.
Financial Disclosure: None reported.