Evolution of the Journal of Neuro-Ophthalmology and the Clinical Ophthalmology Literature: A 20-Year Retrospective

Evidence-based medicine relies on publication of studies that form the basis of the evidence. The studies vary by design including case reports and case series that propose clinical associations, cross-sectional or longitudinal observational studies with larger sample sizes that demonstrate statistical associations, and randomized control trials (RCTs) that compare interventions in matched groups to provide gold standard experimental evidence and meta-analyses, which distill findings of multiple studies. The concept of level of evidence is often applied to generate a hierarchy of study types based on the type of question being asked, with not all study types being applicable to all study questions, both due to the nature of the question (e.g., an RCT is not the best study for a question of prevalence or incidence) and/or feasibility (e.g., RCT may not be practical for study of a rare, slowly progressing disease). The Oxford Center for Evidence-Based Medicine offers excellent resources on this topic (1-3). There has been increasing recognition of the importance of study design and quality of reporting including development of consensus guidelines for reporting of certain study types (e.g., STROBE and CONSORT). Critical reading of the literature has become a focus at all levels of medical education.

Editorial Evolution of the Journal of Neuro-Ophthalmology and the Clinical Ophthalmology Literature: A 20-Year Retrospective Kimberly D. Blankshain, MD, Charlotte E. Joslin, OD, PhD, Stacy L. Pineles, MD, Heather E. Moss, MD, PhD E vidence-based medicine relies on publication of studies that form the basis of the evidence. The studies vary by design including case reports and case series that propose clinical associations, crosssectional or longitudinal observational studies with larger sample sizes that demonstrate statistical associations, and randomized control trials (RCTs) that compare interventions in matched groups to provide gold standard experimental evidence and meta-analyses, which distill findings of multiple studies. The concept of level of evidence is often applied to generate a hierarchy of study types based on the type of question being asked, with not all study types being applicable to all study questions, both due to the nature of the question (e.g., an RCT is not the best study for a question of prevalence or incidence) and/or feasibility (e.g., RCT may not be practical for study of a rare, slowly progressing disease). The Oxford Center for Evidence-Based Medicine offers excellent resources on this topic (1-3). There has been increasing recognition of the importance of study design and quality of reporting including development of consensus guidelines for reporting of certain study types (e.g., STROBE and CONSORT). Critical reading of the literature has become a focus at all levels of medical education. A review of clinical publications in 2 leading ophthalmology journals showed an increase in prospective articles as a proportion of total publications and stable publication of case reports between 1980 and 2000 (4). We sought to make a similar comparison between 1995 and 2015 in the Journal of Neuro-Ophthalmology (JNO) as well as 4 widely read general ophthalmology journals. We chose this interval because it spans publication of guidelines for specific study types, presumably bringing with it increased awareness of study design. For general ophthalmology journals, we selected Ophthalmology, the American Journal of Ophthalmology, JAMA Ophthalmology/Archives of Ophthalmology, and the British Journal of Ophthalmology for study based on review of impact factor rankings for ophthalmology journals, excluding subspecialty and review journals. For these 4 journals and JNO, articles were identified using Pubmed searches for each journal with completion date in 1995 or 2015. Among the 5 journals studied, 1,164 articles published in 1995 and 1,767 articles published in 2015 were screened. Study designs of the articles were categorized as case report, case series, case-control, cohort, cross-sectional, RCTs, and interventional non-RCTs by a single rater and verified by a second rater based on abstract review (Table 1). Figure 1 compares the proportion of study types in ophthalmology journals with JNO in 1995 and 2015. Figure 2 illustrates the change in proportion of study types in 2015 compared with 1995. In the 20 years between 1995 and 2015, JNO had annual publication growth exceeding that of the general ophthalmology journals. Increasing impact factor for JNO during that period suggests that quality increased along with quantity. Thus, JNO is serving an increasing role in the dissemination of neuroophthalmic clinical research. This is particularly important to balance the relatively poor representation of Department of Ophthalmology (KB), University of Cincinnati, Cincinnati, Ohio; Departments of Ophthalmology and Visual Sciences (CEJ) and Epidemiology (CEJ), University of Illinois at Chicago, Chicago, Illinois; Department of Ophthalmology (SLP), Stein Eye Institute, University of California, Los Angeles, California; Department of Ophthalmology (HEM), Byers Eye Institute, Stanford University, Palo Alto, California; and Department of Neurology and Neurological Sciences (HEM), Stanford University, Palo Alto, California. Unrestricted grant from Research to Prevent Blindness to the UIC Department of Ophthalmology and the Stanford Department of Ophthalmology. NIH K23 EY 024345 (H.E.M.) and NIH P30 EY026877 (H.E.M.) The authors report no conflicts of interest. Address correspondence to Heather E. Moss, MD, PhD, Spencer Center for Vision Research at Stanford, 2370 Watson Court, Suite 200, MC 5353 Palo Alto, CA; E-mail: [e-mail] Blankshain et al: J Neuro-Ophthalmol 2020; 40: 141-143 141 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Editorial TABLE 1. Definition of clinical study designs Case Report: n = 1 Case series: n . 1, descriptive, cross-sectional, or longitudinal Cross-sectional study: comparison of metrics between 2 or more groups of subjects or comparison of 2 or more metrics within a group of subjects at a single time point Case-control study: Analytical study of a group of subjects identified based on outcome Cohort: Analytical study of a group of subjects identified at baseline with longitudinal data Nonrandomized interventional study: prospective experimental study comparing treatment where allocation is not randomized Randomized controlled study: prospective experimental study comparing treatment where allocation is randomized neuro-ophthalmology in general ophthalmology journals, where only 3.3% of articles in 2009 represented neuroophthalmology (5), and some of whom lack neuroophthalmic representation on their editorial boards. In both general ophthalmology journals and JNO, there has been an increase in the number and proportion of randomized controlled trials, the gold standard for clinical research, between 1995 and 2015. There was a simultaneous decrease in the proportion of nonrandomized interventional studies suggesting a shift from lower to higher quality studies, which accompanies overall increasing level of evidence in ophthalmology publications (6,7). Other analytical study designs have also increased in both number and proportion in general ophthalmology journals, which suggests continuation of growth in this area previously demonstrated between 1980 and 1999 (4). Publication of articles reflects the combined biases of researchers, funding agencies, and journal editors, with the biases of the latter 2 groups likely impacting the first group. This growth may have resulted in increased awareness of study design on the part of all of these constituencies, with issuance of reporting guidelines for studies (e.g., STROBE and CONSORT) potentially being a cause or result of increased awareness. The most dramatic change in both general ophthalmology journals and JNO from 1995 to 2015 was a decrease in descriptive studies in general and case reports specifically. Case reports were virtually eliminated from general ophthalmology journals during the period of study, in contrast to relative stability in the publication of case reports at the end of the 20th century (4). Although JNO trends generally parallel those of general ophthalmology journals, there are a few points of divergence that reinforce the needs of our subspecialty. First, in contrast to general ophthalmology journals, JNO has a more pronounced increase in case-control studies and a small decrease in cohort studies. This highlights the type of study designs more appropriate to rare neuro-ophthalmic diseases. Second, although the proportional decrease in case report publication was similar in JNO as the general journals, there remained a robust representation of case reports in JNO in 2015, in contrast to elimination from general journals. Although these do not provide the higher levels of evidence characterizing other study designs, they continue to play an important role in hypothesis generation across the widely dispersed neuro-ophthalmic community and thus form an important part of our literature. Over 20 years, JNO has shown disproportionate growth in clinical research articles while showing similar trends in decreasing publication of descriptive studies and increasing FIG. 1. Study designs of articles with living human subjects in general ophthalmology journals and Journal of Neuro-Ophthalmology in 1995 and 2015. 142 Blankshain et al: J Neuro-Ophthalmol 2020; 40: 141-143 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Editorial FIG. 2. Change in proportion of study designs of articles with living human subjects between 1995 and 2015 among articles published in general ophthalmology journals and Journal of Neuro-Ophthalmology. publication of analytical studies compared with general ophthalmology journals. It has maintained case report representation in the literature and favored growth in case-control studies, appropriate to the size and diversity of our subspecialty, as well as the rare nature of the diseases we treat. There remain many opportunities to critically evaluate trends in the neuro-ophthalmic literature including examination of study design choice and reporting quality as has been performed by other authors for specific study types and topics in ophthalmology (7-9) and for the medical literature more broadly (10). These are also important features of the literature because it pertains to serving as the evidence basis for medical practice. 4. 5. 6. 7. 8. REFERENCES 1. The Oxford levels of evidence 2. Available at: https://www. cebm.net/index.aspx?o=5653. Accessed July 13, 2019. 2. The 2011 oxford CEBM levels of evidence (introductory document). Available at: https://www.cebm.net/index.aspx? o=5653. Accessed July 13, 2019. 3. Explanation of the 2011 oxford centre for evidence-based medicine (OCEBM) levels of evidence (background document). Blankshain et al: J Neuro-Ophthalmol 2020; 40: 141-143 9. 10. Available at: https://www.cebm.net/index.aspx?o=5653. Accessed July 13, 2019. Ang A, Tong L, Bhan A. Analysis of publication trends in two internationally renowned ophthalmology journals. Br J Ophthalmol. 2001;85:1497-1498. Kumar A, Cheeseman R, Durnian JM. Subspecialization of the ophthalmic literature: a review of the publishing trends of the top general, clinical ophthalmic journals. Ophthalmology. 2011;118:1211-1214. Bojikian KD, Gupta D, Dettori JM, Dettori NJ, Dettori JR, Chang P, Slabaugh MA. Evidence in ophthalmology: are we doing better? Ophthalmology. 2015;122:2584-2586. Lai TY, Leung GM, Wong VW, Lam RF, Cheng AC, Lam DS. How evidence-based are publications in clinical ophthalmic journals? Invest Ophthalmol Vis Sci. 2006;47:1831-1838. Fung AE, Palanki R, Bakri SJ, Depperschmidt E, Gibson A. Applying the CONSORT and STROBE statements to evaluate the reporting quality of neovascular age-related macular degeneration studies. Ophthalmology. 2009;116:286-296. Siddiqui MA, Azuara-Blanco A, Burr J. The quality of reporting of diagnostic accuracy studies published in ophthalmic journals. Br J Ophthalmol. 2005;89:261-265. Chavalarias D, Wallach JD, Li AH, Ioannidis JP. Evolution of reporting P values in the biomedical literature. JAMA. 2016;315:1141-1148. 143 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited.